All right, good afternoon, and thank you for joining Guggenheim's 2026 healthcare innovations conference. I'm Debjit, one of the therapeutic analysts here, and my privilege to welcome my next presenting company, Edgewise Therapeutics. Joining from Edgewise is the company's Chief Operating Officer, Behrad Derakhshan. I don't know if I got that last name.
You got it. You got it.
Well, thank you for your time, Behrad-
Thank you.
- and if you want to give a very quick intro, and then we can hit the Q&A.
Sure. So very exciting year for Edgewise this year. It's, it's, you know, a remarkable year, as we head into the phase III readout in our Becker program with sevasemten. It's particularly remarkable for individuals with Becker who have largely been ignored by this industry, and we have the potential to launch the first drug in the history of Becker for these individuals who have an extremely high unmet need. The closest catalyst that I think we're gonna talk about a little bit today is the part D data and obstructive and non-obstructive data. Very exciting program in hypertrophic cardiomyopathy. Tremendous unmet need still remains in the obstructive and non-obstructive patients, despite recent advances in that space.
And then, as you know, we've got a second cardiac program, EDG-15400, that's slated to go into phase I in the second half of this year. And so a lot of activities going on, a lot of value-creating milestones in the next 12 months for the company. So really exciting, and looking forward to talking more with you about it.
Awesome. That's a great segue. Let's start on the CIRRUS Part D data.
Sure.
As a company, are you convinced the AFib issues that you had, you dealt with last year are now firmly behind you? The screening is working, at least for the,
Sure
... 20 patients who have you presented data in December.
Yeah. Maybe I'll describe some of the changes we made to kind of reduce the observations that we had in some of the early trials. What we really essentially did coming out of Part B and C is take a look at some of the inclusion, exclusion criteria that our peers in the CMI space had done, and what we realized is we hadn't necessarily matched them. So what we did proactively in Part D is we amended our screening criteria to ensure that, first of all, we were including patients who were actually hypertrophic, right? So we had relied primarily on site-read echoes to select patients, and what we noticed is we had a couple of patients who actually didn't meet the wall thickness criteria for HCM. So now we're relying on core lab to make that determination. So that's one change.
The second is, we had a 90-day look-back period on history of AF before recruiting patients. We have now since amended that to have a 180-day look-back period, with which matches like for like for what we've seen with the CMIs. We've also looked at, started to look more closely at the patient history, right? What are the level of comorbidities, in particular, looking at, you know, mitral valve stenosis. If patients have moderate to severe mitral valve stenosis, there's a high... You know, these are high-risk patients, so we're excluding those. I think most importantly is we proactively implemented a cardiac, continuous cardiac monitoring. We basically went and said: Look, we want to dissociate that there's any relationship with the mechanism of 7500 and AFib. So we monitored more than anyone in the CMI space to look for AFib.
We went looking for it, basically. So, what we did is we have a cardiac monitor with a patch that individuals wear through the four weeks of screening, and then for two weeks, every time we dose escalate individuals. And what we reported at the end of last year is, during that period of extensive screening, over 2,000 measures of screening, we saw no AFib burden in these individuals. We saw no relationship between plasma concentration and dose and incidence of AFib. So that's really encouraging. So, in answer to your question, I think we've made tremendous progress in getting away and finally getting beyond the AFib narrative.
Got it. And, in terms of the timing for the CIRRUS D data, it's gonna sort of read out right around the time of the Acacia study.
Yep.
Are you gonna let the Acacia read out first and then put out the data, or how are you thinking from a cadence perspective?
I think from a cadence perspective, we're really... What's core to what we're trying to do at Edgewise is really make sure that the rigor and then the integrity of the data that we put out is true to kind of, the quality of the data we're generating. So at the end of the day, we're not really guided by what our competitors are doing. I think we want to make sure we put out the best data package that we can, coming out of what is a pretty significant data update, right? We're not just gonna report on non-obstructive, we're reporting about obstructive and non-obstructive data. And really, what I think is really unique about what we're trying to do is really dose optimize every individual to their max efficacious dose, because, as you know, we don't have this relationship between plasma concentration and LVEF.
So, we feel pretty confident that both in obstructive and non-obstructive, we have a lot of room to improve what's already out there. And so we're not guided by what our competitors are doing and when they're putting out data, but largely by when we're ready to put out the quality of the data that we think is commensurate with what we're doing.
With respect to potential REMS-
Yeah
... What have you heard from your physicians or KOLs?
Yeah.
How can you best avoid the REMS?
Yeah, obviously, you know, if you think about the wealth of data we've generated to date, healthy volunteers, we did the single dose obstructive, we've done Part B, obstructive, Part C, non-obstructive, we've got a ton of Part D data. You've kind of seen the LVEF graph that we continually update as we get more data points. I think we're starting to get to a place where, you know, physicians are getting comfortable around the fact that we've dissociated LVEF from plasma concentration. Now the question is like, what's the bar with the FDA?
So, the best that we can do right now as we kind of capture data and go for an end of phase II, is really reached out to and had the ex-head of the cardiorenal division at the FDA, and we put kind of a little bit more of the data than you guys have seen publicly in front of him and said, "Look, this is the data package that we have." And his feedback was like, two things: One is you're a different mechanism to the CMI class, so there's no reason for what's good for A is good for B. So that's one.
The second is, you've already got a lot of data, and by the time you go to the end of phase II, you're gonna have even more data that hopefully continues to show that there's no relationship between plasma concentration of 7500 and LVEF. At that point, it's gonna be very difficult for the FDA to say, "Okay, we're gonna make you do intensive echo monitoring." So our ingoing discussion with the FDA is, "Hey, look at other drugs in this area that are not in the CMI class. What if we do an echo at baseline and an echo at end of study, and that's it?
Got it. So assuming the profile holds-
Yeah.
Where do you think 7500 fits in in the obstructive setting?
Yeah.
And then what happens if you don't have any echo or REMS monitoring, where does it fit?
So I think there's still room in the obstructive population to improve. I think we've seen real, you know, really nice data with the CMIs, but I think there's still room to improve, and I think where the opportunity really lies in obstructive, if you think about it, the majority of patients are actually sitting in the community. They're not at the centers of excellence. The center of excellence have largely been quite good at using Camzyos. I think if you think about going outside of the centers of excellence, what matters to a community physician? So think about what a community physician is managing. Community physicians manage patients with heart failure. The way they assess how a drug in heart failure is working is they rely on feel and function. Predominantly, they use NYHA.
So if you have a drug that spares LVEF, that doesn't require rigorous outpatient echo monitoring, that allows a physician to dose a patient to efficacy without concerns on safety, and can use feel and function as a measure to determine that, that's a win. I also think even in the centers of excellence, if you think about the echo burden, right, the CMIs have the black box warning related to heart failure because it drops in ejection fraction.
If you now come with a drug that doesn't have that black box warning, what you can do is essentially say, "Hey, doc, put the patient on the drug and then move them through doses as you see fit, but you never have to do an echo in order to dose optimize." They might then have the flexibility of introducing echoes as part of the norm of their practice in HCM. So if you think about how physicians do echoes today is, if I've got a patient who's an obstructive or a non-obstructive HCM, I'm gonna do echoes every six to 12 months if they're not on a CMI, 'cause that's just how I would, you know, manage their degree of disease progression. That's easy. But if you have to do echoes within a very tight timeframe, that's the limitation.
Having so many echoes in a very tight timeframe limits the ability to expand beyond the center of excellence, but also limits the ability of the center of excellence to put as many patients as they want on the CMIs. And we've seen this with Camzyos, and, you know, we'll see what happens with the next gen.
Got it. Going back to the AF, background rate versus what you're, you know, capturing right now, is there a window that after that, if you pass, let's say, three to four weeks, it's very unlikely you're gonna land up at AF if you do the right screening?
I don't think there's a correlation between, you know, time and AF. I don't think we think that there's necessarily a correlation between... I mean, at least based on the observations that we had previously from the part B and C, where we saw those four cases, one, there was no concentration relationship, and two, there was no temporal relationship. So I think it really comes down to what is the background of that individual? How complex is that individual? What comorbidities do those individuals have, and how does that increase their risk of AFib over time, right? So I think that, I don't think it's necessarily you get past, you know, week four, and then you're free. I think it's a question of like, where was that patient when they came in, and how are they responding to therapy?
Got it. So in the non-obstructive setting-
Yeah.
-the KCCQ numbers are clearly better, at least now-
Sure.
Clearly better than CMIs.
Yeah.
Where do you think that you have to be to become, let's say, the standard of care in-
Yeah
... the non-obstructive setting?
Yeah, it's a really good question, right? I think what's interesting about non-obstructive is a diastolic disease, right? So you don't have a gradient. If you think about a lot of the benefits that you see on diastolic parameters in obstructive patients is a function of the fact that you've moved this physical barrier and relieved the gradient in these obstruction patients. Now, 7500 is a positive lusitrope. What that means is that you have a drug that improves ventricular filling, improves cardiac output, and reduces the potential for having heart failure, right? Essentially, what you're doing is you're increasing the volume of blood that pumps through the heart between each heartbeat. That's a good thing, right?
So, so our mechanism is preferentially working during early relaxation, during diastole, and that is the core central mechanism that's driving the HCM pathophysiology. So I think when you think about that, I mean, you pointed it out, at 50 mg fixed dose over four weeks, we saw 16-point improvements in KCCQ. That's better than some of our peers have shown in similar types of trials. So I think what's really exciting about Part D is now you've got a trial where you're essentially, your goal is to dose optimize every patient for efficacy. So the physician has the discretion, and what we're using for efficacy in non-obstructive and Part D is-...
You either meet the 50%, at least the 50% reduction in NT-proBNP, or you meet 200 picograms per milliliter, which is how we're defining the threshold for normal in these individuals. So now the physician has the ability to dose optimize every patient on NT-proBNP, but at the same time, he's looking at, "Okay, what's their NYHA? What's their KCCQ? Do I need to move them to the next dose to maximize their feel and function?" Because they're never pushing up against LVEF. And this is a fundamental difference to the CMI class of molecules that have a preferential effect during systole. When you dose those molecules in non-obstructive, what you're running up against is you're not dosing to max efficacy because you don't have a gradient to measure using echo. You're dosing to LVEF.
That's your limit. But if you don't have LVEF as a limit, then you can continue to dose and essentially get more patients to their efficacious dose and classify them as a responder. I think that's the key, again, going back to this comment around how do you fit into the community practice, which is where a lot of these patients are sitting, is make it simple. No echoes, feel and function.
So given what you just laid out, assuming Acacia is going to be a positive study, would Edgewise as a company choose to go in a registration study non-obstructive first to be second in market and be clearly differentiated, or we go simultaneously in obstructive as well and be third to market there?
I think you go all in on HCM, right? I think you have to look at the data we've generated in obstructive. I go back to the fixed-dose data we put out last year, right? At the 100 milligram fixed dose in four weeks, 89% of patients achieved gradient responses of below 30 at rest and below 50 at Valsalva. Approximately 60% of patients had NT-proBNP drops in the normal range. We had the most impressive KCCQ improvements of 23 points. That's almost close to what you see with septal reduction therapy surgery. Coupled with that, 78% of patients had at least a one-point improvement in NYHA class. That tells me that you've got a tremendous room to improve, and no LVEF drops, by the way.
So that means that you still have a very competitive profile in obstructive, and I think you're right. I think in non-obstructive, the opportunity to me is, you know, sky's the limit, frankly. If we can continue to decouple plasma concentration of 7500 from LVEF, which seems to be there, we've decoupled those two, I think we can push in non-obstructive and really get out to the community. I think that's where a lot of the opportunity is sitting, and I think that's where we're going to win.
When you think about the registration study, then do you think you'll have the same sort of a titration regimen that you're doing in Part D right now?
Probably not. And I can tell you one fundamental difference and one of the reasons we did extensive kind of monitoring in Part D is essentially where we want to get to in obstructive patients. 'Cause remember, part of the value proposition is you don't want to do a lot of echoes. So what we're trying to do is build a level of evidence in Part D, particularly in the obstructive, to say: Look, I actually don't need to do serial echoes to look at gradients, because I can look at NT-proBNP and feel and function and guide my dosing in obstructive patients based on measures that are not echo-related. Now, if the physician wants to do an echo to look at gradient, they can do that, but that's not a mandated requirement for as part of the study, ultimately.
So I think that's going to be a fundamental difference. And I think, you know, when you think about non-obstructive, I think it's going to be pretty similar, right? You're going to base your decisions based on feel and function and biomarkers. And then, you know, the primary endpoint there is KCCQ/peak VO2. I think we need to have a discussion with the FDA to see, like, is there a preference for one, or do we do kind of the co-primary endpoint step strategy? So right now, I know I can anticipate kind of the next question is like, what are we thinking about from a trial design? I think there's two things to think about. One is, do we run parallel trials in obstructive and non-obstructive, similar to what we've seen for the CMIs?
And then the second is, is there a potential to accelerate by doing a single study? And I think we have to explore both of those situations, right? We've seen Sonata with the SGLT2s from Lexicon and doing a single study with a KCCQ endpoint. So that's kind of an interesting situation. So we have a good idea of how we want to do our phase III trials. We've got multiple different plans, and I think it's a function of when we get the data and we speak to the FDA, how are they going to respond to the data and the schedule that we're going to propose to them in order to, you know, maximize the success.
Got it. So, let's fast-forward to the time you have the public disclosure on the CIRRUS Part D data.
Sure.
Would you already have met with the FDA to get clarity on the registration program, or that would be a-
That would come after.
Come after that.
Yeah, 'cause we've got to get the end of phase II meeting done, so we've got to capture the package, make sure it's cleaned up, and, and so on and so forth. So... But we, we'll provide some—I mean, we're still on track to initiate the phase III in the fourth quarter of this year, and probably in anticipation of that, we'll talk about kind of the phase III design.
Got it. And is there a world in which you could see a synergistic use of 7500-
Mm
... with CMIs in the obstructive setting?
It's kind of interesting, right? We've talked about this a little bit, and if we ever chose to kind of explore that, I think you would—what you would do is you'd probably do it in a phase IV study. But practically speaking, right, if I'm a physician and I'm limited on what I'm able to do in obstructive because I'm pushing up against LVEF, I might just switch that patient to 7500 because you're not losing on efficacy and you're gaining on lack of drops in LVEF. So I think the—what I think might happen is you'd probably get more of a switch dynamic rather than using 7500 on top of a CMI.
... Got it. So let's finish up with the cardiac side, with your 15400, I don't know how you-
15400
EDG-15400. How are you thinking about healthy volunteer data?
Yeah. So, you know, we're pretty far into getting that study completed. It's SAD MAD healthy volunteers. We're gonna put out that data in the second quarter. You know, at that time, we'll also kind of describe and highlight some of the key features around EDG-15400 that are different when you think about EDG-15400 relative to EDG-7500. Same target, but slight differences. And then we'll also disclose kind of our plans for HFpEF. That trial is on schedule to kick off in kind of the third, fourth quarter of this year. We have a sense of what the bogey is. We've spoken to a lot of people who are really, you know, well-versed in the HFpEF space.
We've got an SAB that we meet with regularly, and the general consensus is, if we're able to show a 25%-30% reduction in NT-proBNP in phase II, that there's probably good translatability to phase III feel and function endpoint, like a KCCQ, for example. So that's the bogey. Obviously, you wanna see more, but that's kind of the bogey that's there.
Got it. So, let's move to sevasemten, GRAND CANYON.
Sure.
Have you had any recent interactions with the FDA, given the complete uncertainty that we are dealing with the agency right now?
We haven't. We had pretty clear guidance from the FDA the last time we met with them, kind of, second quarter of last year, where they were like: "Look, you know, pivotal Grand Canyon study is a single registrational study," you know, and they accepted North Star as a primary endpoint. That kind of got lost a little bit because people you know, were really looking forward to an accelerated approval. But actually, just getting buy-in from the FDA, that in neuro division, that your study is a good study and it's powered adequately, and so on and so forth, was a really good outcome. So, you know, we're just kind of waiting to see what happens right now. We're anticipating Grand Canyon fourth quarter on track.
You know, what's been quite impressive about that is, as you, as you know, you always build a buffer on dropout rates in these studies. We've had a very low dropout rate. These patients are very sticky, and it's somewhat reflective of what we observe in, from patients from our ARCH. So 99% of patients who are eligible to enroll in the MESA open label have enrolled and continue to be dosed with sevasemten. That's pretty remarkable. I've been in rare disease for 15 years. I've never seen an open label have that level of stickiness.
That's good to hear. In terms of, you know, how are you thinking about it internally, given all the data that was generated heading into the Grand Canyon study?
Yeah.
Is there any scenario where this disappoints?
Well, yeah. You know, it's kind of interesting, right? 'Cause if you remember... Originally, we targeted 120 patients. Once we kinda told the sites that, "Hey, we were ready to close," because we were getting close to that 120, all of a sudden, you know, there were a lot of patients that kind of flooded in. And we were very careful, we didn't wanna kind of get ahead of ourselves, so we screened them to make sure that they were still the right patients and they met the criteria. And ultimately, we ended up getting 175 patients. So that moved our powering. So we were powered 98% to show a 1.8-point improvement at 18 months on GRAND CANYON. The tipping point for stat sig is 0.8.
So there's a little bit of room there for us, right? It's drug development, you know, functional endpoints in a neuromuscular disease. Anything can happen. But I think I take comfort in looking at a lot of the natural history data, looking at a lot of the placebo data that we have, and building a lot of models. Every scenario that we look, it seems like this drug is doing something that is not in line with natural history. So that's why I'm really kind of positive going into Grand Canyon.
Given the high compliance rate that you're seeing in Grand Canyon and the 99%-
Yeah
...roll over to the open label study, what does that tell you commercially?
That tells me that once patients go on, they wanna stay on because they feel better, right? And this is the value of the MESA data, because it's not only like, you know, convincing a patient to go on, there's also the payer narrative that you have to ultimately build to say, "Hey, once the patient is on drug, they continue to benefit." So there's an argument to kind of maintain that orphan drug pricing because you've got continued benefit beyond just the study period that you looked at. So I think commercially, you know, the idea is to get a broad label. So you wanna get patients at the point where they're diagnosed during adolescence, because the argument is, I don't want them to get to the cliff before I treat them.
I wanna delay or prevent them from ever getting to the cliff. That's a key value driver to this market, and I think this open label data, this stickiness, talks to, one, the unmet need that you see in Becker that's been underappreciated because of the noise around Duchenne, and two, that if you feel better, you're gonna wanna continue to be involved in a trial where the schedule of visits is not immaterial to these people who obviously, clearly have functional deficits. So very encouraging from a commercial standpoint.
Given that the data are sort of, let's call it, imminent in 2026.
Yeah
... have you started building out a commercial organization right now?
Yeah, we've you know, R. Michael Carruthers, who's our CFO, he just joined us from SpringWorks. He obviously just launched a couple of drugs in rare disease. You know, I have one of what I would call top teams in rare disease commercial. They're setting up a really nice infrastructure right now, where we've made some spend at risk, but we've been very careful, understanding that, as you said, the data is imminent. You know, we're ahead of where we need to be, which I'm really glad about, and if the data reads out positive, I think we're gonna be ready to launch this. You know, hopefully, the approval will be sometime in the, you know, end of the fourth quarter of 2027, and our first full year of revenue will be 2028, if everything goes well.
Yeah, very well set up to execute on that.
In the last second or so, any thoughts on DMD?
Yeah.
What's the go, no-go decision there?
So Duchenne is, as you know, very complex. There's been recent setbacks. You know, it's been hard to, to kind of find a signal in these individuals. I go back to the data we presented last summer. I think it's some of the best data that I've seen, where, you know, we saw reductions in the rate of decline in, in, in individuals who obviously have a very progressive disease. So I think for us, it's... We've got a clean cohort of 10 mg patients that we need to look at, both naive or experienced with gene therapy, and then at that point, we have to look at that data, we have to speak to the FDA and agree on what is the primary endpoint? Is it SV95 or North Star?
What's the duration of the study, and what are the patients we wanna look at, right? This is an evolving competitive landscape, and so what we're gonna see is, the patients we're looking at today may not be the commercial patients at that point, and we wanna predict what's gonna happen then. So ultimately, we wanna make sure that we have the right patients in order to maximize the reach of the drug if we pursue that strategy in Duchenne. That decision will probably happen after we turn the card on Grand Canyon. But we're getting sites ready should we need to kick off that phase III.
Awesome. Fortunately, we're on the clock, and I'm really looking forward to the CIRRUS Part D update-