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

Hello, and welcome to the Edgewise Therapeutics Update Call. [Operator Instructions] As a reminder, this conference is being recorded today. If you have any objections, please disconnect at this time. I would now like to pass the call over to Michael Nofi, Edgewise CFO. Please proceed.

Thank you, and good morning. Welcome to the Edgewise Therapeutics conference call to discuss our top line data of EDGE-7500 from the Phase II CIRRUS-HCM Part D 12-week trial in individuals with obstructive and nonobstructive hypertrophic cardiomyopathy. This morning, we issued a press release, which outlines these results. You can access the press release as well as the slides that we will be presenting today by going to the Investors and News section of our website at www.edgewisetx.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. Behrad Derakhshan, Chief Operating Officer; Dr. Aylin Tugcu, Senior Vice President, Cardiovascular Clinical Development and Medical Affairs; Dr. Michael Ayers, Director of the HCM Center of Excellence at the University of Virginia and EDGE-7500 Clinical Program Lead; and Dr. Chris Duton, Senior Vice President, Cardiovascular Clinical Research and Operations. At the special guests, we have Dr. Anjali Owens joining us, a CIRRUS-HCM Investigator and Medical Director at the Center of Inherited Cardiac Disease, and Associate Professor of Medicine at the University of Pennsylvania. Before we begin, I would like to remind you that some of the statements made during the call today 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 any undue reliance on these forward-looking statements, and Edgewise disclaims any obligation to update these statements. I will now turn the call over to Kevin.

Thanks, Michael. Thank you all for joining today. Edgewise is a leading muscle disease biopharmaceutical company developing novel therapeutics for muscular dystrophies and serious cardiac conditions. The company's deep expertise in muscle physiology is driving a new generation of novel therapeutics. Sevasemten is an orally administered first-in-class skeletal myosin inhibitor now in Phase III studies for Becker, which will read out in the fourth quarter and in Phase II studies in Duchenne muscular dystrophy. In our cardiovascular pipeline, we have EDG-7500, which is a novel cardiac sarcomeric modulator for the treatment of symptomatic hypertrophic cardiomyopathy currently in Phase II, which will be the subject of our disclosure today. And as a second novel molecule, EDG-15400 is a novel cardiac sarcomeric modulator for the treatment of heart failure, currently in Phase I, which we're moving into Phase II in the third quarter. We've really had a company-changing event as of June 1. I'd like to describe it a bit for you. So Edgewise Therapeutics entered into a definitive agreement under which Servier, an independent international pharmaceutical group governed by our foundation, will acquire Sevasemten and Edgewise's muscular dystrophy platform, which will conclude all IP, know-how and additional key personnel. These milestone payments of up to $1.1 billion and a cash infusion of $1.55 billion make for a very important event for Edgewise and the ability to independently move forward and become a true cardiovascular company. We think Servier is an excellent partner with the global scale, patient commitment and commercial reach to maximize the full potential of the drug for the treatment of patients with Becker and Duchenne. We expect this event to close sometime in the third quarter, subject to regulatory clearance. With that, I'd like to pass it over to Michael Ayers to discuss the unmet medical need in Hypertrophic Cardiomyopathy.

Thank you, Kevin. Hypertrophic cardiomyopathy is the most common inherited cardiomyopathy affecting around 1 in 300 people. It is underdiagnosed, misdiagnosed and even patients who are correctly labeled suffer from significant delays in diagnosis before the shortness of breath, chest tightness, fatigue, passing out or ultimately recognized as HCM. There are 2 subtypes, obstructive and nonobstructed, and we'll get into that on one of the upcoming slides. But suffice it to say, current treatments have significant limitations for both subsets in obstructive patients, but particularly nonobstructive hypertrophic cardiomyopathy, an area breth of any real approved evidence-based therapies. As we zoom out and look at the heart models on the next slide, you'll notice the healthy heart in the center, where all of the segments are around 1 centimeter thick. Now the first thing I'm going to point out on the heart of the left and the heart on the right, the obstructed and nonobstructed heart is actually a similarity at a cellular level. Both of these hearts are virtually indistinguishable on a cellular level, both have abnormalities in how they contract and both, more importantly, have abnormalities in how they relax. So now let's zoom out and think a little bit about the differences. The difference in an obstructive and nonobstructive heart is not the heart muscle itself, it's actually a location problem. The heart on the left will suffer from a blockage of blood flow when the thicken heart muscle is positioned in just the wrong spot. And now in order to achieve 120 millimeters of mercury of systolic blood pressure, the ventricle has to create 170 millimeters of mercury or more of pressure. There's an uncoupling of that work and now the heart is essentially working harder than it needs to. The heart on the right doesn't suffer from the location issue, but still has those same cellular problems, abnormal contraction and more importantly, just like the heart on the left, abnormal relaxation. Segueing out of that and beginning to take a 30,000-foot view as a state of modern therapies, I'll be quite candid. The obstructive population on a population level has some relatively good options. Modern therapies make patients feel better, they can exercise further. Their blood work looks better. But on an individual level, we're going to see a much less rosy picture. Only 2 out of every 3 patients in modern trials in obstructive hypertrophic cardiomyopathy will achieve full gradient relief or full symptom relief. And only 1 out of every 3 patients will achieve both full gradient relief and full symptom relief. Now there's really 2 hypotheses in the field why that might be. The first hypothesis is that these drugs reduce obstruction by reducing systolic function. That is how they work. And while that might alleviate obstruction, that might have some undue consequences when it comes to exerting yourself and your ability to muster up some of that cardiac reserve. And it certainly has some more pragmatic implications in introducing an echo-dependent dosing strategy. Now the second hypothesis as to why there's so much residual symptomatology despite effective gradient relief is that these drugs are actually mild or modest at most in improving diastolic function. Moreover, the way in which they improve diastolic function comes at direct systolic liability. We have coupled the diastolic improvement with systolic cost. And that's why when you zoom out and you look at both disease states, the real unmet need here is a drug that can alleviate obstruction without introducing systolic cost and potentially improve diastolic function directly. We now can see kind of a granular build of that unmet need, the residual symptoms, the fact that there's no approved therapies whatsoever that are evidence-based and nonobstruction. And on our top 2 bullets, we have that echo-dependent strategy, that test nisnance for providers who are having to migrate through the REMS and for patients who are having to return for echoes on weekly to monthly cadences. And that's not to mention the elephant in the room, the black box warning for systolic heart failure, which is directly tied and linked to the way these drugs work. It's in that light that 7500 steps into play with a novel mechanism of action. This drug reduces obstruction, not by diminishing overall systolic performance, but by reducing early systolic pressure generation. It also directly changes myocardial mechanics in a way that improves diastolic performance and hopefully introduces us to an era that I think you might think is quite possible of an echo independent dosing strategy moving forward that not only divorces itself from echoes, but also begins to address that large unmet need of diastolic dysfunction. And with that, I'm going to turn it over to Dr. Aylin, who can hopefully convey some of what makes us so excited this morning.

Thank you, Michael. CIRRUS-HCM Part D is our 12-week open-label goal-directed dose escalation study of EDG-7500 in both obstructive and nonobstructive HCM. The primary endpoint is safety. We enrolled 53 patients, 20 with obstructive and 33 with nonobstructive HCM. All patients started at a 25-milligram dose and were up titrated every 2 to 4 weeks with the goal of optimizing efficacy and tolerability. The escalation criteria were subtype specific. For oHCM, dose titration was guided by left ventricular allow tract gradients resting below 30, Valsalva below 50-millimeter mercury. For nonobstructive HCM, titration was guided by NT-proBNP targeting below 200 picograms per milliliter or greater than 50% reduction from baseline. After week 12, patients transition into a long-term extension on their optimized stable dose with physician flexibility to use up to 200 milligram. Next slide, please. Here are the baseline characteristics for our 53 participants. This population is directly comparable to what was enrolled in pivotal CMI trials. For our obstructive cohort, mean aged about 59, 45% female, 70% NYJ Class II, 30% NYJ Class III, atriibrillation history of 10%. Baseline KCCQ scores in the low 60s, confirming meaningful symptom burden. Left ventricular ejection fraction was preserved at 66.6%, resting LVOT gradient was 40-millimeter mercury, Valsalva 85-millimeter mercury. The nonobstructive cohort was younger at mean age 49, with longer disease duration, higher NT-proBNP at 781 picograms per milliliter and higher atrial fibrillation history at 21%. KCCQ scores were comparable across both subtypes, around 61 to 66 confirming similar symptom burden. The populations are representative, symptomatic and directly benchmarkable against published CMI data. Next slide, please. We screened 107 individuals, 53 were enrolled, a screen failure rate of 50.5%, consistent with pivotal CMI trials. Now let me walk you through the dose distributions because the 2 subtypes landed in different places, and that's a direct result of different escalation criteria we use. For obstructive HCM, where titration was guided by LVOT gradient, 65% of patients reached 100 milligram or higher by week 12, with 35% on 100 milligram and 30% on 150 milligram. About 1/3 of patients met the gradient target at lower doses, so the protocol held them there. For nonobstructive HCM where titration was guided by NT-proBNP, the distribution shifts to the right. 84% reached 100 milligram or higher with a full 50% at 150 milligram. The NT-proBNP target required more dose to achieve and the escalation design allowed patients to get there. So the difference in dose distributions between the 2 subtypes is driven by the escalation criteria, left ventricular outflow track gradient for obstructive HCM, NT-proBNP for nonobstructive HCM, not by differences in tolerability. I'd also note that our Valsalva gradient target for obstructive HCM was below 50-millimeter mercury, whereas pivotal CMI trials used below 30-millimeter mercury. Despite that, the efficacy results are robust across every domain, and we see that as a meaningful optimization opportunity as we move into dose selection for our Phase III trial. With that, I will turn it over to Dr. Anjali Owens to walk us through the left ventricular ejection fraction and efficacy data of Part D.

Thank you so much, Aylin, and good morning to everyone. It's really my pleasure to be here today to present the top line results of the 12-week Part D data. I'm going to present the results in 3 sections, starting with safety by a deep dive into all ways of assessing systolic function, then we'll focus on efficacy in the obstructive population followed by efficacy in the nonobstructive population. We have not observed any detrimental effect on systolic function with 7500 to date. Here, the first slide shows data from our Part A single-dose study and you can see on the left, resting gradients, which are responsive to Cmax versus Ctrough with the maximal gradient response observed at Cmax. But what you'll observe on EF on the right-sided graph is no sensitivity to the concentration of drug, a very flat line for EF despite the concentration and when it is taken. Next slide. Next, I have core read echo data in the obstructive group and the nonobstructive group. You can see there on the left, looking at LVEF at baseline and over 12 weeks of treatment. And on the right, the absolute change from baseline. These are extremely flat lines with very little variability, and we saw no participants with a drop in LVEF to less than 50%. Next slide. Here's another way of looking at the LVEF change. This is looking at LVEF absolute change from baseline versus EDG-7500 plasma concentration. I have dots in gray that represent the healthy volunteer placebo group, and you can see the range of LVEF on the left for them and then the obstructive group and the nonobstructive group in the triangles and the squares. And again, what you can see here is no relationship between LVEF and plasma concentration. There is no decline in LVEF as concentration goes higher. There is also no unexpected drops at low concentration. Next slide. This is another way of looking at systolic function. This is global longitudinal strain complementary to left ventricular ejection fraction. And you'll see the absolute change in GLS from baseline on the left. You got your healthy volunteer placebo group, again in gray and their range, followed by the obstructive and nonobstructive groups, looking at a far range of concentration. Again, you see no change, no decline in global longitudinal strain despite the concentration of 7500, very flat, very reassuring that we're not affecting systolic function with this drug. Next slide. Yet another way of looking at systolic function. This is global circumferential strain. Again, another way of looking at how the heart is functioning from a contractile state, healthy volunteers, placebo in gray, followed by the obstructive and nonobstructive group. And again, a very flat line right near 0, showing that despite even very high concentrations of 7500, we are not seeing a decline in systolic function as measured by global circumferential strain. Next slide. Here, we have all of the patients to date that have been dosed with 7500. So you have the single and multiple ascending dose population, Parts A and B, obstructive HCM, Part C, nonobstructive HCM and all of our Part D participants versus the EDG Plasma Concentration. And what you can see here is a very, very flat line. We are not seeing declines in ejection fraction that are dose dependent nor are we seeing outliers with unpredictable drops in ejection fraction. This to me is very good data that's accumulating over time, supporting that we may be able to move away from dosing based on echo because we don't have to worry that the ejection fraction is going to drop at higher dose levels with 7500. Next slide. So to summarize that data across all studies, we now have greater than 240 individuals who've been exposed to 7,500 with more than 700 echos now performed assessed by the core lab and over 420 of those echoes were in patients who have HCM. We've seen, one, no clinically significant changes in LVEF; two, no reductions in LVEF to less than 50% and very importantly, no heart failure events, no hospitalizations due to drop in LVEF. And this is with doses spanning from a very low dose at 25 milligrams all the way up to a high dose of 300 milligrams. This observed lack of systolic liability does support moving toward an echo-independent dosing strategy. This would uncouple dosing from safety echos and allow us to dose based on symptomatic improvement. And finally, this preservation of global circumferential strain, which is a very highly sensitive marker of systolic mechanics, really separates the mechanism of action of this drug from CMIs and the published data on CMIs. Next slide. Next, let's move to efficacy. We're going to start with the obstructive population where we saw robust gradient and feel and function improvements, again, with no impact detrimentally on LVEF. I'll start with the gradient data pictured on the left, resting data in the blue line, Valsalva data in the light green. This is baseline over 12 weeks of dosing. And at baseline gradients were 40 dropped to 18, which represents over 50% reduction versus baseline. Valsalva gradients were quite high in this population with a baseline of 85 dropping to 40 after 12 weeks, which again represents greater than 50% reduction. And you can see that in aggregate, this cohort reached our levels of obstruction that are below the threshold that we think is important. So below 50 for Valsalva and below 30 for resting. Overall, in this population, we saw 90% demonstrated clinically meaningful improvement in their gradients, either at rest or with Valsalva. Next slide. What happened to our biomarkers. So we know from studies of other drugs that modulate contractility that when you decrease the gradient, you typically see improvements in NT-proBNP and troponin. And that's exactly what we saw here after 12 weeks in the obstructive population. So you can see on the left, about a 60% reduction in NT-proBNP baseline to 12 weeks. And on the right, about a 40% decrease in troponin baseline to week 12. That's summarized with 74% of our obstructive patients who achieved either a normal level of NT-proBNP or a reduction of greater than 50% from baseline. And I'll just call your attention to the time line. We are seeing those drops quite early at the 4-week visit. Next slide. What about how patients feel and markers of disease state and stated quality of life. So for that, we're looking at the KCCQ score. And this is the KCCQ-OSS pictured in the graph on the left, where you can see from baseline to week 2, a pretty strong improvement even within the first 2 weeks that continues to improve. At week 12, we're up to a 24-point improvement in the OSS score. The CSS score also improved by about 20 points from baseline to week 12. Again, these are open-label studies, and we know that from other work, placebo accounts for at least 5- to 7-point increase. So we'll take this with a grain of salt, but very favorable data. And in comparison to an open-label COLLIGO study of mavacamten, this is real-world data for mava, about an 11-point increase by week 12. Next slide. What about a responder analysis? How do we get a sense of how many patients are going to improve, which is really what we want to know when we move into Phase III and ultimately, if the drug is approved. So this responder analysis looked at the percent of obstructive patients who achieved what we think of as a clinically meaningful change in the KCCQ-OSS defined as a greater than or equal to 5-point change. And we saw that 85% of patients in this cohort achieved that with 75% achieving greater than or equal to 10 points. And I'll just call your attention to the last bar where there's a very large improvement of over 20 points in over 30%, almost 40% of this cohort. So robust clinical improvement. Next slide. Let's look at that in another way. This is looking at field by NYHA functional class, where one is asymptomatic. And you can see here that 55% of patients are over half achieved an asymptomatic status by 12 weeks of treatment with 70% improving by at least 1 class. Next slide. This is an analysis that looks at the transition from 3 to 2 or 1 or 2 up to 1 and also includes patients who may have stayed unchanged. Now it is quite hard to go from Class III with the symptoms with very minimal exertion to asymptomatic. And so I just want to call your attention to 50% of the patients who went from Class III to asymptomatic at Class I. Next slide. What do we think is maybe the cause of this improvement in symptoms, what's happening at the level of the myocardium? And we'll get into more of this at the end of the presentation today with some high frame echo data. But what we know from transthoracic echo is that we're seeing rapid improvements in the mean E-prime lateral velocity. And that is a signal that tells us that diastolic function is likely the driver of improvement here is that we are affecting lucotrophy, ventricular relaxation and improvement in diastolic function. And you can see the gains that were here by 12 weeks. Next. Okay. Let's move into nonobstructive data, and I will show you here early observations at 12 weeks. And I'll just caveat this by saying that in my clinical experience, it often takes longer to affect patients with nonobstructive HCM than it does obstructive HCM where you don't have that gradient release and after load reduction as an early marker of benefit. You can move to the first slide. So let's start with biomarkers since we don't have a gradient in nonobstructive HCM, we focus our attention really on biomarkers, how patients are feeling and then evidence of diastolic function. So let's start with the biomarkers. And you can see here, baseline biomarkers were in the 700s for the NT-proBNP, and we saw a 65% reduction in NT-proBNP levels from baseline to week 12, and we saw a 33% reduction in troponin from baseline to week 12. And we don't know what will happen in the long-term extension, but I would submit to you that these curves to me look like they are still moving down and have not yet plateaued. So I am not sure that we've reached the maximal effect that we will ultimately see in long-term extension. We did see that 88% of these patients achieved either a normal level of NT-proBNP or a reduction of greater than 50% from baseline. Next slide. Let's look at KCCQ. And for simplicity's sake, we stuck with the OSS, which was also shown for the obstructive population. And we saw a 13-point improvement in KCCQ-OSS from baseline to week 12. And for the CSS, we saw a very similar 12-point improvement. Now again, if you take a look at that curve, it is still rising. I do not think we've yet seen the plateau at week 12, which is not surprising in a nonobstructive population, and we'll get additional data from long-term extension to see when that plateau will occur. This is a similar responder analysis. And again, this helps us to understand what percent of patients do we think will be helping with a therapy such as 7500. And we saw in this analysis that 73% of patients had what we would consider to be clinically meaningful levels of KCCQ improvement defined as greater than or equal to 5-point improvement. And again, I'll call your attention to that very large improvement group, which is pretty impressive over just 12 weeks. Next slide. This improvement, to put it into context, this is, of course, not a head-to-head study. These were 2 separate studies, one of aficamten, which was open-label Cohort 4 and 7500 on the left, again, open label. And we saw a 73% of patients with clinical improvement at week 12 compared to about 56% with aficamten from the REDWOOD study. Next slide. Let's look at KCCQ improvement as it relates to NYHA improvement. Again, these are our markers of how patients feel on a therapy. And we found that 52% of patients or just over half achieved Class 1 by 12 weeks. And again, Class I is elusive and really means that you've become asymptomatic. And we saw a larger percent, 64%, who had at least 1 class improvement. And on the next slide, I'll walk you through what those changes were. And you can see here that 58% of patients went from Class II to Class I and a smaller percent went from Class III to Class I, which is much harder to do. Next slide. What do we think is happening at the level of the heart and driving these improvements. Again, it takes us back to diastolic function in this disease in nonobstructive HCM, we know the heart is stiff. We know it is noncompliant and that drives up filling pressures that drives up the NT-proBNP and drives symptoms, particularly exertional symptoms. So what we've seen in the echo data is an improvement in the E-prime lateral of 1.6 centimeters per second that starts as early as week 4 out through week 12, and this represents a 37% improvement from baseline. Next, we'll transition back over to Dr. Ayers to talk about some preliminary data from our high frame rate echo study.

Thank you, Dr. Owens. It's crystal clear as always. We appreciate your expertise. We're very motivated to continue to marry our preclinical data to clinical data that helps with this differentiation of mechanism that makes it clear. One of those studies is a high frame rate echo study, which is going to help us understand how clinically this drug has these novel lucotropic or diastolic benefits as well as some novel benefits on right ventricular function and the way in which it's alleviating obstruction. As a teaser to some of that data to come, some of that high fidelity data set, what we have here are the E/e prime ratios for our obstructive and nonobstructive cohorts. E/e prime is considered one of our best, if not the best, stands for how high the pressure is inside the ventricle at the end of diastole. And what we see here are remarkably robust decreases in E/e prime in both the obstructed and non-obstructed cohorts that are not only directionally consistent, but consistent in magnitude of reduction. As a comparator, if you were to look at the E/e prime literature for, say, odyssey, you're going to see around a 1 to 1.3 point reduction in E/e prime. So these are significantly more improved than what's been published prior. We think that this is reflective of the tissue level myocardial relaxation that we are embarking upon our myocardium with this drug. With that, I'm going to segue into the safety data. This has been a generally very well-tolerated drug. To move through that data on a granular level, we'll see that no participants in our trial have experienced a left ventricular ejection fraction of less than 50% or heart failure due to drops in left ventricular ejection fraction. When we start moving through our treatment-emergent adverse events, or TEAEs, the most common that we have seen is in obstruction, we had 3 patients with fatigue. And in non-obstruction, we had 3 patients with an upper respiratory tract infection, the cost of doing business in the winter and 3 patients with a rash. We had 2 participants with new onset atrial fibrillation, both were in our obstructive cohort and both were ultimately deemed unrelated to study drug given very high background risk of atrial fibrillation in those 2 individuals. We had 5 participants with serious treatment-emergent adverse events, all of which were considered unrelated to drug that we won't go into detail on this slide. With that, I'm going to turn it back to the boss.

Thank you, Michael. So what we hope to have demonstrated here to you all is that EDG-7500 is a novel mechanism with a differentiated safety profile relative to standard of care, which is the CMIs. There are a couple of things about the CMIs, I think I should state that describe their mode of action. They are complete inhibitors of cardiac contraction. So at some point, you can completely eliminate cardiac contraction. We are a modulator of cardiac contraction. We only can inhibit less than 50%, and we are a partial inhibitor. That means that we can never drive the ejection fraction down in the dangerous levels. We think that this partial inhibition is reflected in the mechanism of action is that we are always poised to reengage the active filament and when a patient exercises or a patient has some level of activity, we have a greater level of cardiac reserve, which allows us to actually come back and fully perform functions. This translates in the clinic into better effects on feel and function measures and deep responses in NT-proBNP, like we've shown in this presentation. A third piece of data that we showed preclinically is that we have stronger effects on the diastolic portion of cardiac activity. I think this data clearly shows direct lusitropic effects based on the high frame rate data and the core data for echocardiographic data. Finally, I think the lack of systolic liability supports the potential of having an echo-independent dosing paradigm without any risk of drug-induced systolic heart failure. We have provided an initial synopsis of a draft Phase III protocol that does not include echo-based titration. The agency has directly told us that the trial design was reasonable for this class of agent. And that they would like to see additional Phase II data to support our desire to have a trial where we do not need to have multiple echos to actually get patients to the appropriate dose and do not have a drug mechanism that drives drug-induced systolic heart failure. And we're on track for Phase III initiation in the fourth quarter of 2026. Next slide. So why is this really important? The CMIs, because of the black box warning and the need for ejection fraction monitoring require multiple echos and highly trained, sophisticated physicians to monitor the drug. And in fact, that reality caps the number of physicians that can prescribe this drug. And that the real opportunity in hypertrophic cardiomyopathy is to move our drug, 7500 into the community cardiologists who do not want to do multiple echos. This would expand utilization of the drug and expand the market outside of the center of excellences dramatically. We think this is the real opportunity for our drug mechanism and the profile we've shown. So finally, I'd like to finish up that given the Servier deal, we are in a really strong position with a runway into the early 2030s. We currently have pre-deal $500 million on the balance sheet, no debt and 105 million shares outstanding. Next slide. So here are the milestones for the next year or 1.5 years. We anticipate having a Phase III initiation of 7500 in obstructive and nonobstructive HCM later in the year, probably fourth quarter. We intend to -- now that we have the final data set for Phase II, have an end of Phase II meeting with the agency early next quarter. We should have design feedback by the fourth quarter. And we would expect to be able to provide 48-week long-term extension data of 7500 in the first half of '27. We have another really exciting drug for the treatment of HFpEF, which is 15400. We will initiate and dose our first patient in the first quarter or the third quarter of 2026. We anticipate being able to release the first cut of data sometime in the first half of '27 and initiate a Phase III in the second half of '27. We have the finances to be able to do all of that and more based on our Servier deal. Next slide. With that, I'd like to thank you all for your attention and happy to take any questions.

[Operator Instructions] Our first question comes from Jenny Gonzalez-Armenta with Leerink Partners.

It's Jenny on for Joe. Can you just walk us through why the 2 new onset AF events were deemed unrelated to Edgewise EDG-7500 and whether there are any apparent relationships to dose exposure or timing of dose escalation? And was that assessment investigator determined or independently reviewed?

Michael, do you want to take that?

Yes. So when we communicate with our PIs, what we do is we allow them to use their clinical judgment and relatedness. In these particular instances, the patients had Afib risk calculators that were performed as part of standard of care that were both high to very high. And given the fact that they had very robust improvements in treatment, the PIs thought that the overall background disease rate was more likely driving this than our drug. That said, our second question is an excellent one. We've done extensive work looking at dose exposure relationships, temporal relationships, et cetera, and we've been unable to find any convincing connection between our drug and risk of atrial fibrillation. We're pretty happy that we've put a lot of this narrative to bed and are ready to move forward in a randomized fashion in Phase III.

And maybe just as a follow-up, if we can ask the clinician, like what are her thoughts on this rate and how she would see this if the Phase III confirms efficacy and maintains the profile, how she would view that in clinical practice?

I think, Anjali, how would you -- looking at this profile, how would you think about how competitive this molecule would be if these were the same results after Phase III, I guess?

Yes, happy to answer that. AFib is part of this disease. Up to 30% of patients with HCM have Afib. We see it more in the highly obstructed patients who've been obstructed for a long time. So it's certainly part of this disease. And I think the way to answer the question is the way that you answer any question, a randomized placebo-controlled trial. And I think that's what we need. We saw rates of Afib in all of the CMI trials to date, and you really need to answer the question with a randomized trial where you have a placebo arm.

Our next question comes from Tessa Romero with JPMorgan.

Congratulations, Kevin and team on these data. I think the most critical question here from us is what do these results here mean for the design of your Phase III program? Even if it's just initial thoughts, can you provide some perspectives on how you think about the right endpoint or endpoints in treatment duration here for each of these populations to optimize the chance of success? And Dr. Owens, perhaps you can just share how you think about this as well.

Just -- I'll just take this on the high level on the clinical trial design. I think based on our mechanistic evaluation, we think that we would have a strong effect on things like peak VO2. From a design aspect, I think it's been clear from the EMA and from the FDA that peak VO2 is an important measure or at least a CPET life measure, coupled with a KCCQ. So I think that's pretty clear of what the primary endpoints ultimately will be. There are some nuances to those primary endpoints. I think we're still under discussion with the agency on the trial designs that include 2 different trials for each population or combined study. So we have not come to a conclusion on that. So I think it's pretty much all we can provide today. I do believe ultimately that we will not have an echo-based titration. And that's all of the information we've gotten back from the agency that, that will be a clear differentiator for our drug. I know, Anjali, do you want to provide some thoughts on the Phase III design at a high level?

Sure. Yes, I think there is a drug that we have that has not shown any safety events with systolic function. So that really opens up the possibility of designing a trial in a way that's very distinct from what we've seen for CMI trials. I think that's a good thing in general to move beyond that and really focus on symptom benefit and function benefit. And I agree, the endpoints are similar to what we've seen in other trials that should be considered, but with the broad possibility of changing dosing strategies, et cetera. With regard to the time line for endpoints, I think that's a harder discussion, and there may be some nuance between what you see in an obstructive population and what you see in a nonobstructive population in terms of time that it takes to realize the full benefit of the drug. And so I think you might have to split the difference a little and perhaps look at the nonobstructive patients a bit longer.

Yes. And I think just to add to that, with the open-label extension, we'll be looking at patients beyond this 12-week study and that data will be important in making a decision of the duration of the trial and are there significant differences between the 2 populations.

Our next question comes from Yasmeen Rahimi with Piper Sandler.

Congrats on the outstanding data. One of the key objectives of the CIRRUS study was also to really explore dose titration, which you really did very well. Give us a glimpse of additional work that's needed? Or do you think this is the titration that you would propose in the Phase III study? And if that's -- or if there's a couple more dose PK/PD analysis that needs to be completed?

I think there's still additional analysis that we clearly need to do. And that's partly the open-label extension. But as you noted or as we noted, we had stopping rules for increasing dose that included gradient, which we would not use in Phase III. So we have -- are now analyzing and looking at the data of feel and function relative to the gradient response. As you can see in our initial single-dose data, there is a significant difference between a Cmax measure of gradient and a Ctrough measure of gradient. And what we've noticed preliminarily is that the gradient is not as good a predictor for feel and function with this mechanism likely because of PK. But the NT-proBNP continues to be a very strong correlate with feel and function measures. So I think that all combined, we don't have that fully crunched yet, but I think there's still much to be said about where we would start in dosing, where we would end up and the duration of those doses.

And then maybe one last question. Given with the strong balance sheet, would you consider also running a head-to-head against a beta blocker or even a CMI trials.

Certainly under consideration, but there are a number of -- Anjali certainly could opine on her view of the beta blocker versus novel mechanism approach. I think in some respects, that's a second trial as opposed to the initial trial. But it's still under the discussion and possible. So maybe that's the most complete answer I can give you.

Our next question comes from Joshua Yan with Raymond James.

This is Josh on for Marty. We just had a quick question on specifically the KCCQ. We were just wondering if you can maybe describe to us like how KCCQ responses improve over time specifically and as well as like for the dosing in Phase III, do you -- I see that I think 50% of patients are on 150 mgs in NHCM portion of the trial. Do you expect to push doses even higher? -- congratulations again on the data.

Aylin, let me speak to maybe what's known in nonobstructive HCM and the duration. And maybe Anjali can weigh in on that as well.

Yes. But we have seen in prior CMI trials with nonobstructive HCM, then the trials have been longer in duration that it takes longer effect to get a KCCQ to plateau. What we have seen in our trial is that at week 12, the KCCQ was still in an uptrend, and we have not seen a plateau effect. So the KCCQ seems to improve over longer treatment duration. And I think that's consistent with prior CMI trials as well for the nonobstructive HCM. And I'll turn it to Anjali, who can also chime in here.

Yes. I think the real concept here is to titrate the dose in a way that is commensurate with the half-life of the drug, its mechanism of action, get a patient to a stable dose for some period of time prior to your primary endpoint, right? So that's really what we think about when we're designing the trial in terms of timing and dose. And I do think that if you have no liability in terms of systolic function and no other adverse event that has cropped up yet as a limiting feature, then you do want to maximize the dose so that you get maximal benefit on your feel and function endpoints. And if that takes 8 weeks or 12 weeks, then you probably want to have a patient at a stable dose, whether it's 150, maybe there are patients who will need 200, we'll see in the open-label long-term extension if that's the case. get them to that stable dose and then give it a few weeks, give it probably 3 months so that they then have the primary endpoint at a point where they can maximize benefit.

Our next question comes from Adithya with Evercore.

This is Adi on for Cory. I had a question for the doc on the titration limits set, especially for the HCM patients, the NT-proBNP target. Could you maybe describe to us details about this target set? And if these -- how would this play out in the clinic in terms of tactical aspects?

Yes, Anjali, I think he's asking in regards to how you might utilize an NT-proBNP in a Phase III and the pros and cons of that measure.

Yes. So it's certainly a marker of stretch. And we know from the obstructive CMI trials that the response on NT-proBNP early in dosing was a very strong predictor of the ultimate improvement in peak VO2, for example. Again, this is the obstructive population. So I think that's the ballpark that you're aiming for is an early and robust decrease at least by 50%. And where we come up with that number, it's a little bit made up, and it's a little bit looking retrospectively at CMI data, but you want to get at least 50% down from baseline, I would say. And that's where these targets have sort of come from loosely. I am not a strong proponent of using that in terms of Phase III titration or anything like that. I think Phase II is the time to really explore what those changes are, what you can expect from a dose, a certain dose of your drug and then make your best guess in terms of titration and dosing to get maximal benefit. But the reason we do it is that at least in the obstructive population, it's a predictor of what you're going to get for peak VO2 change. The nonobstructive group is a bit harder, and we don't really understand yet what the best targets are in terms of surrogates that will tell us that we'll get an impact on peak VO2. That is still a little bit of an open question, and I think it probably is because that the nonobstructive group is much more heterogeneous. And there are other factors at play that impact change in peak VO2 for that group.

Got it. So just from a clinician perspective, if we do think to titrate based on feel and function, what sort of markers would you use if proBNP is still not cemented?

In the real world, we do use NT-proBNP, and we use just a basic interview that we conduct regularly when patients come in to see how they're feeling relative to baseline. Again, these are very simple questions. Are you better than you were? Are you the same? Are you worse? And then we couple that with biomarkers. And there are some programs in HCM centers that utilize NT-proBNP. There are others that do not. There's recent data from the HCMR study showing that NT-proBNP can be an important prognostic marker in this disease. And so I think you'll find over time that more of us in the clinical world, real world are using NT-proBNP. So we use it in conjunction with the interview with our patient and what their heart looks like to get an overall sense of whether they're moving in the right direction. And if the NT-proBNP remains high, for example, once they've been optimized on a CMI in the real world commercial drug, and they still have symptoms, then we may be targeting other things. We can put them on the treadmill and see if they're still obstructed and go up on the dose. We may add something like an MRA or an SGLT2 inhibitor. We may address their obesity with the GLP-1. So it kind of gives us a ballpark sense of where the patient is in the trajectory.

If I may just add to that, and I'm going to turn it to Michael as well. I think with our Phase II data, we have a good understanding what our target dose is for both obstructive and nonobstructive patient population. What is really wonderful about 7500 is we do not have to up-titrate against ejection fraction, which gives us a lot of flexibility. And with that target dose, we have an additional option to up-titrate based on symptom or based on how the patient feels and functions. So that is very different than what has been done until now where ejection fraction was a liability or something to uptitrate against, which will not be the case in 7500. Michael, do you have anything else to add?

In some ways, I'm taking my edge-wise hat off and putting my clinician hat on when I answer this. What our Phase II data has done is it has established a good understanding of a dose and efficacy relationship that we can use to design a successful Phase III trial. And it's done so without revealing some dose safety concern that should limit our titration scheme in Phase III, which leads us to be really optimistic headed into Phase III. But the clinician component of this is that ultimately, what we hope to provide exiting Phase III is a situation where you as a clinician get to use your favorite marker question study to titrate to your patient in the room. We want to remove this echo-dependent strategy that has led to medications not being given to the right patients in the right time course, right? We're not utilizing the current class of medications like we could. You're being tied to these other arbitrary rules of how to go up and down. And it could just be the doctor in the room with the patient deciding how they're going to use this drug to effectively make their patient feel.

Our next question comes from Laura Chico with Wedbush.

I have 2. On the first one, this relates actually to the extension study. Do you have a sense as to how many patients or what proportion are actually increasing towards the 200-milligram dose in the extension trial? And if you could just elaborate a little bit more on what those drivers would be to titrate higher in the extension study, that might be helpful. And then second, I guess, I'm sorry if I missed this. For Phase III, I'm trying to understand what's the specific echo schedule you are proposing to FDA for your design? I guess I'm assuming REMS placement would be a review issue, but I'd love to know if I'm wrong there.

So we haven't given a number of patients. I would say there are a number of patients that have gone up to 200. And I think it's been a combination of giving the physician flexibility. So some have moved up based on NT-proBNP. Others have moved for feel and function. If they were Class II, they would get additional drug and look to see if you could drive the Class I in the open-label extension. But that's all we can provide today on that note. What was the second question?

REMS.

REMS. Yes. So what we have proposed to the agency is no echoes during the titration, 0. And the agency directly in minutes told us that design looked reasonable, pending additional Phase II data.

Our next question comes from Debjit Chattopadhyay with Guggenheim.

Can you hear me?

Yes. Hi debjit.

Congrats on the data. I have a couple of follow-ups here. On the nonobstructive side, given the impact on diastolic function, why is the KCCQ somewhat lagging the CMIs? And then for the physicians on the call, once approved, how would you use 7,500 over the other CMIs? Or is there sort of a subset of patients who are automatic candidates for 7500 as opposed to a CMI?

That's completely false that we are lagging CMIs. I don't know where you get that data. Please provide it to me.

We're just looking at the ROC ACCQ numbers, right, in the teens versus given your prior Phase Part Ab data where it was much higher.

Well, that was an end of 2. I don't know what to say. Perhaps that's a level of interpretation. Anjali, when you're interpreting some of this KCCQ, can you walk us through how you think about it from a -- from a clinician standpoint?

Sure. The KCCQ is granular in a way that perhaps is different than what we do as clinicians in the real world. We don't use the KCCQ at our center. We follow about 2,000 patients with HCM. And so we're really benchmarking how they are relative to baseline. And that roughly correlates with the KCCQ score and an improvement of greater than 5 points is what's been shown to be clinically significant and a meaningful improvement. So when you look at improvements in the teens, that is significant improvement from a clinician standpoint and usually map back to a patient that says they feel much better.

Got it. And in terms of prioritizing 3,500 over the CMIs?

Sure. So I can give you our experience here. We follow about 300 or so patients on commercially available CMI, and we have a wait list of probably 20 patients right now waiting to get in to start a CMI. And the reason they're waiting is because of the backlog in echos and visits that are required for the REMS program. So we have reached a point at a large center of excellence with over 2,000 patients where it's difficult to start additional patients. So there is a burden from a REMS standpoint for the echo lab, for the clinicians, for the pharmacy team. If there was a drug that did not require safety echoes to be performed and could be dosed based on clinical judgment alone, that would really open up the avenues to provide the drug to an increasing number of patients. If you have similar efficacy to a CMI without the safety constraints, that would really open up your ability to prescribe.

[Operator Instructions] Our next question comes from Kripa Devarakonda from Truist Securities.

Congratulations on the data. For KCCQ, just a follow-up question. Did the CFS trajectory follow the same pattern as the OSS data that you showed? And just based on -- there's been concern from the KCCQ values because of the lack of a placebo arm. Dr. Owens, can you just help us understand a little bit more? I know you talked about it a little bit, but how you interpret these KCCQ values? And if there's anything we need to take away from the delta that you see between the nHCM and the oHCM patients?

Anjali?

Yes, sure. I'm happy to. So nHCM, in my experience, just takes longer, a little longer for them to feel the same benefit, again, because the acute relief of obstruction and the after load that goes with it is often an impetus for a more rapid change in symptoms. So I think it's a little easier to target the obstructive group because you've got that target of a gradient when it comes down, feel better. It takes a little bit longer if you're looking at diastolic function and remodeling to improve filling pressures with the nonobstructive group. That would be my shalt on that. I take KCCQ with a grain of salt always. And in particular, we know that there's a placebo effect in all studies with HCM that we've looked at, open-label and even our Phase III studies, we see a pretty strong effect even in the placebo groups for KCCQ. So I think you do have to take that with a grain of salt. That being said, the best we can do is look at other open-label design cohorts of CMI, and that's what we provided in this data set, not as a direct comparison, of course, 2 different drugs, 2 different trials, but as a ballpark framework for how to interpret the KCCQ changes. And I think if you look at the data that we showed, it's favorable.

Our next question comes from Lander Egana-Gorrono from H.C. Wainwright.

So based on the data, were you able to identify which efficacy endpoint, either biomarker or echo data correlates best with symptom improvement in KCCQ and NYHA?

I think that's -- I'll take that. I think that's still under evaluation and would be also subject to looking at longer-term data to see how that evolves over time since the trial will be at least 24 weeks. So I think that's part of the discussion.

Awesome. Awesome. And maybe one more. Just even if there were no ejection fraction reductions below 50%, did any of the patients experience a significant decline in ejection fraction that remained above 50%?

Well, no. So what we showed was all of the data that we have, and it's all within the range of the placebo group from the normal healthy volunteers, which is plus or minus 10%, which is the variability in the measure of the ejection fraction. So no, we did not see any patients outside of that placebo range.

We have shown you other markers that measure systolic function such as global longitudinal strain and global circumancial strain, which are more sensitive markers of systolic reduction. And as you have seen, we have not seen any dose exposure relationship with those markers as well, which is different than what has been shown with the CMIs.

The dose exposure curves are not only flat, but there's minimal excursion. And so as you spend some time looking through those, I would focus on both of those features.

Our last question comes from Paul Choi with Goldman Sachs.

Congratulations on all the progress. My question for Dr. Owens is if you take the totality of the data here in both the oHCM and nHCM subgroups, particularly the gradient changes as well as the biomarker changes, if you were to project to the best of your ability, what the peak VO2 changes might be in a randomized population for the next study, how would you guess this might potentially compare versus the data we have on hand from the pivotal studies for the CMI inhibitors?

That's a tough one, but I'll do my -- nothing like predicting a peak VO2. We can't do that on a good day, by the way, even on patients not on CMI. So I would say, let's separate the populations in the obstructive group, I think it's easier to predict that if you have a robust effect on gradient and NT-proBNP, that you will have a robust response in peak VO2. So what extent that will be may depend on the comparator arm and whether or not you've included patients on beta blockers, what the doses of those AV nodal blockers are. We know that can affect your exercisability, your heart rate reserve, et cetera. And with the mechanism of action that's different than CMI and may preserve that cardiac reserve, I think we may see a real benefit, in particular, in patients who are not on beta blockers. So that's one question. For the nonobstructive group, I think it's a little bit harder to predict from what we know on CMI, we're looking at a 0.4 to 0.6 change in peak VO2 despite a very robust reduction in NT-proBNP. And it may be that we need to look at a slightly different assessment of CPET, something that's a submax marker in order to really understand and maybe that it's harder to move the needle on peak VO2 in nonobstructive HCM. It may be that you need a subset of patients who are poised to respond. And again, that goes back to background therapy, whether you're suppressing their heart rate response, whether you're suppressing their ability to have cardiac reserve. And with the mechanism of action that, one, targets lusitropy; two, does not prevent cardiac reserve, I am optimistic that we'll see favorable changes in peak VO2.

I think that's the end of the questions. So really, I'd like to thank everybody involved in this program. Certainly, Dr. Owens, really, thank you very much for providing your insights today. Certainly, I would like to thank all the employees of Edgewise who participated and made this happen. I'd like to really thank all the clinical trial investigators that execute on this trial and then most importantly, the patient community who participate in these trials. And then ultimately, in the end, the shareholders who have our continued confidence and support. So it's been a really exciting time to be here at Edgewise. I believe we're on the cusp of demonstrating the potential of novel therapeutics that could fundamentally improve the lives of patients who need more effective therapies, and thank you for joining the call.

Thank you for attending the Edgewise Therapeutics Update call. This concludes today's conference call. You may now disconnect.