Good morning. Thank you for joining Guggenheim's 2025 healthcare innovations conference. I am Debjit, one of the therapeutic analysts here, and my privilege to welcome our next presenting company, Edgewise Therapeutics. Joining from Edgewise are CEO Kevin Koch and Chief Operating Officer Behrad Derakhshan. I'm sorry. I tried. Okay, let's get started with 7500. The Part D data are forthcoming. Obviously, there were some enrollment challenges, especially screening patients in the Part B, Part C. What have you guys done not to sort of have that be an overhang on the data?
Yeah, so I'll give you how we thought about the data. We thought the data was spectacular coming out of the initial April disclosure. What was really exciting about it was, essentially, we saw no relationship between the ejection fraction changes and the concentration of the drug, which is highly differentiated from the CMIs. That is a really important concept. The second thing we observed, and we're really excited to be able to report, is that we saw dramatic changes in feel and function. This includes things in the New York Heart Association changes, as well as KCCQ scores, and then robust changes in key biomarkers like NT-proBNP. That is a very important observation that we had. The compound was generally well tolerated, except that we had observed a couple of cases of atrial fibrillation. As you all know, in HCM, atrial fibrillation is highly pervasive.
It depends on the nature of those individual patients, on their risk factors. What we noticed as we were moving through the study is really two things. One is that the core lab that reads all of the echoes was different than the PI echoes that were observed. In this study, we were using the PI echoes and their interpretation of the data to enroll the patients. What we found is that two of the patients were not HCM patients because of the wall thickness of the patients. The other two, and in fact, this occurred in all those patients, had very high risk factors for atrial fibrillation, with the two main ones being the left atrial reservoir strain and the size of the atria. These folks had very high levels of pathology associated for AF.
What do we do to address that particular issue? Because other than that, the data was spectacular. We, instead of having simply the PI read the echo and make the call of who was enrolled in the patient, incorporated our own internal look at the patients from an expert in echo to take an initial look of it to confirm what the PIs said. We have a core lab, which can look at all of the echoes and ask the question, is this person high or low risk for AF? Of course, we want to enroll a representative group. What we did was match what was observed in Sequoia and Explorer. We integrated that, and we started integrating that particular part of the program about April of this year. We have been recruiting patients since April.
Now, what we first did, and this was a very fast and easy protocol amendment, was we went back and dosed 25 mg patients in BNC. You remember, when we dosed at 50 mg, that was an active dose. Good clinical practice would mean that we would go and look at what is the minimally effective dose for these patients. We have enrolled multiple patients in both non-obstructive and obstructive at 25 mg in the BNC portion, which is the 28-day portion of the study. We also, in parallel, did a protocol amendment, which was more substantial because you have to build a new database to actually do a three-month study, which is the Part D. That Part D took about a month and a half to get that database built, get it through the FDA, get it through the IRBs.
We started screening patients in June. That screening has been robust. When you get these clinical sites up and running, they do not all happen at once. They happen in kind of a rolling fashion. We now have about 20 sites open. We have been screening actively since June. We have actually treated quite a few patients at this point. We will provide an update later in the year. That process now allows us to remove patients that are very high risk, and we should be able to now look at the efficacy of the drug in isolation from any AE profile.
That was very helpful. Take a step back. The data that are expected end of the year may not necessarily be the Part D data, but we'll get the Part B and Part C follow-up at the 25 mg?
Yeah, so we're still discussing what's the best forum to talk about the 25 mg efficacy data. What I think we want to provide is an update of where we are in the Part D studies and where we are in the 25 mg studies. Some level of granularity of how many patients, when's the anticipated readout for efficacy, and what is the AE profile of the drug over this timeframe when we've been treating patients.
Got it. For the Part D segment, given that in the earlier part, the AFibs happened relatively quickly.
Yeah, they're all within the 28-day period.
Exactly. When you get a given update on the Part D, at least there would be enough in there to take the safety overhang off the program.
Yeah, I think that's been the only overhang people have said. The efficacy is spectacular. The limited or no changes in EF has been differentiating. Now let's focus everybody's attention on the efficacy of the drug and the potential of how we'd run the phase threes.
Got it. That's helpful. If you think about the broader landscape in obstructive HCM, where do you think 7500 fits in?
I think it fits in across the board in HCM. Right now, what is holding back the market opportunity within HCM for the CMIs are the REMs. The need to titrate the drug with three or four echoes to get to the right dose, to get to the right therapeutic index, and to trying to avoid the ejection fraction changes. Most of the cardiology groups that are in the community do not have that echo capability or want to be involved in so many echoes and the risk associated with titrating by echoes to be dosing the drug. Invariably, where the drug, where mavacamten is being dosed, in their center of excellence. Even though they have great capacity, that capacity is limited.
What you need to do is have a drug that has no safety issue associated with ejection fraction, does not need to have serial echoes, and shows as good or better efficacy in the HCM population.
When we were talking to docs initially, most of the physicians were sort of aligned with, say, up to 30% of the patients come in with borderline EF, and 7500 becomes their go-to drug in that setting. You're basically saying you can just completely get rid of the titration phase of current.
Right, that would be the goal. Yeah, it would be the goal. How we're running the Part D study is we're looking at 25 mg, 50 mg, 100 mg, and even up to 150 mg or even 200 mg. We're getting every piece of data along the way and trying to say, where can I replace an echo? Can I replace it with zero echoes or no more than one echo just to give some people some guidance of where they are? With obstructive HCM patients, the docs like to see what's happened to the gradient. In non-obstructive, there is no guidepost with an echo. There's no reason to do an echo in non-obstructive unless you have a safety issue.
We want to eliminate the safety issue, and then that should allow us to eliminate as many as all the echoes or maybe just have one as a guidepost.
I think the LVEF piece is a really important element to this because even at the centers of excellence, I mean, Kevin's talking about the opportunity beyond the centers of excellence, but even at the centers of excellence, there are kind of capacity in terms of how many patients they can initiate, take through the titration, and then monitor long term. I think there's one example, KOL out in San Francisco. He doubled his echo staff and still only has less than 10% of his patients on MABA. That tells you something. He's at capacity. He's got everyone that he can on the drug, but he has other patients that could benefit. I think the onerous titration schedule is limiting that. I think one of the concerns they always have is once they put the patient, is that patient's EF going to drop?
Because they can't predict that. I think that becomes a bigger problem in non-obstructive. We saw that with Odyssey, right? More than a quarter of the patients had their ejection fractions drop and had a dose interruption or had to stop dosing. That is a big problem, both in obstructive from getting everyone who's eligible on the drug, but in non-obstructive in particular, where overall your baseline EFs are much lower.
Do you think there is room on the efficacy side, or this is more?
Yeah, you know, it's interesting, right? If you read the AHA guidelines on how to treat patients with obstructive HCM in particular, but also non-obstructive, it's not to gradient. It's not to peak VO2. It's to feel and function. And feel and function, you can do it two ways. It's the physician-administered, which is typically the NYHA, and then it's the patient perception, which is the KCCQ. Both in the obstructive population, both on NYHA and KCCQ, we saw really robust responses. In fact, the KCCQ observations were better than anything we've seen therapeutically out there. Almost as good as having a septal reduction therapy, which is when the surgeon goes in and removes the obstruction physically. That tells you that we're having a very dramatic effect in terms of efficacy. To your question, I think, yeah, absolutely, there's room to improve on efficacy there too.
Let's talk about the feel and function, especially in the non-obstructive patients.
Sure.
What, mechanistically, how does a modulator outdo a CMI?
I think a key aspect of it, and we've described this biochemically, is that we're a partial modulator of the ATPase, so the myosin. We only inhibit that part of the mechanism down to about 60%. Now, if you look at the preclinical data for the CMIs, you can completely eliminate the contraction. That buffer we have allows us to get to a higher level of target engagement. The other aspect of the mechanism is we're changing the rate of the contraction. At different portions of the beat, you're changing the rate in which you're changing the contraction, and that allows you to modulate the contraction in the early portion of where the obstruction is actually occurring when the mitral valve is coming over and engaging the ventricle. That mechanistically is different.
Those two points allow us to get to a higher level of engagement without seeing EF changes, which would drive greater efficacy. I think both of those aspects of what the drug does is highly differentiated and leads to potential for greater efficacy because of the safety margin.
Got it. Let's fast forward to 2026. When you have the full Part D data available, how are you thinking about your next steps from a registration perspective?
Yeah, so it's an interesting time. It would be quite straightforward to run trials that are identical to what have been done in Sequoia and Explorer and Odyssey. I think we've taken an approach where there's multiple opportunities here. One might be going head-to-head against beta blocker, like in a MAPL-type structure. Might be an interesting way. We've seen that beta blocker actually can be detrimental for some patients. To randomize against beta blocker might be an exciting way of running the studies. Alternatively, we think that we could run a head-up, a very similar study to, say, Sequoia and Explorer, and probably be highly successful. What we noticed in the KCCQ, we were seeing 25-point changes in the KCCQ. placebo-adjusted, you're only seeing 7-point changes from the CMI.
Within four weeks.
Within a fixed dose.
That's a key thing. With a fixed dose, without any titration to optimize for efficacy, we were seeing 25-point changes. That's dramatically different than what the CMI is. I think one thing that people fail to really recognize is lowering ejection fraction cannot be good for your feel and function. Literally, if you go from 65 to 45, how would you expect to feel better? You're changing the amount of oxygenated blood that's being pumped through the body. The act of lowering ejection fraction in and of itself causes a decrease and a blunting of the efficacy you can see with the CMIs.
We saw that with Odyssey, right, in non-obstructive patients. I think they just had an analysis, excuse me, at AHA where they looked at people within certain thresholds at baseline on LVEF. I think it was 60, people above 60 and people below 60. The conclusion was the lower your starting ejection fraction, the worse you respond on some outcomes. I think it tells you that low ejection or dropping ejection fraction makes it very difficult then to really fulfill full efficacy in these individuals. You're starting from a deficit.
Given the big delta you're seeing, 20-plus points, your studies don't necessarily have to be as big as being currently done.
Yeah, so the primary endpoints will be probably KCCQ and peak VO2. The peak VO2, we're not measuring peak VO2. We felt that the time it would take to get that up and that the ends would be too small. What has been observed, at least in the obstructive HCM patient population, is the tightest correlation with peak VO2, which would be one of our primary endpoints, is the depth of the NT-proBNP response. We noticed this as well in our studies, that if you got the patients below a threshold with NT-proBNP, those patients were the ones that responded both on KCCQ and responded on New York Heart Association changes. We think we can use NT-proBNP as a guidepost of how the depth of the response needs to be. I think I have been saying this now for almost a year.
Getting people to normal, normal is normal. Forget about the numbers. Normal is normal. You get people below a certain threshold in NT-proBNP, you get them in the normal range, it's normal. You get them to New York Heart Association one, they're normal. You get them to a KCCQ above 80, they're relatively normal.
Got it. Let's move on to HFpEF because we are being mindful of the clock because we can spend the whole day talking about this. You touched upon NT-proBNP. Maybe that's a good segue to HFpEF. Thoughts on the next steps for the program?
Yeah, all the KOLs we've spoke to have something you can measure early. As we grind through the toxicology studies and everything else we need to do with a new drug, the thing you can measure in three months, and we would anticipate running a three-month phase 2A study in HFpEF, the thing you can measure in three months is NT-proBNP. You won't have enough ends to measure rigorously KCCQ and other measures. You can get a feel for that, but you're not going to be able to run it rigorously. The plan would be to look at NT-proBNP. The bogey given to us by most of the KOLs is that if you see a 20% decrease in NT-proBNP, you ultimately will see clinical benefit in much larger studies. That's been seen across multiple heart failure studies.
I think what we can say is if we can generate that kind of response in a three-month study, that would be very exciting. The first aspect of a, say, a 2A in HFpEF would be, what do I see on NT-proBNP? Is the drug tolerated? Can I get to, what is the dose range in which I'd like to dose these patients? I think we can do that effectively. We will start that study in 2026, and that'll be an important really milestone for the company in that patient population.
Given how large that segment is, how quickly can you enroll and can you get to that data point by the end of the year?
Yeah, we're inclined to run a small placebo-controlled study, but we haven't with the exact design. We may have a much more adaptive design where we can expand the enrollment, but we're still talking about the optimal strategy for HFpEF. We are in the middle of the healthy volunteer studies now. We should have data in the first half of 2026 describing that particular molecule and the advantages of what we might be able to observe in HFpEF with that particular molecule.
Got it. Let's switch to sevasemten . Maybe thoughts on the strategy there. Are you committed to go doing a DMD registration study, or do you need more feedback from the FDA? Because that's sort of in the air right now, at least from our understanding.
It's really a combination of two things. One is, what's the background of the patients that you would like to treat in DMD? There are patients who are on just steroids. There are patients who are on steroids plus exon skippers, and there are patients who are on steroid and gene therapy. In both FOX and LYNX, and FOX, we've dosed patients who are on a background of gene therapy. We know that those patients, unfortunately, are progressing. The rate of that progression is difficult to ascertain because that data has not been published. From what we have seen, though, and we talked about in June, is that we seem to have a competitive advantage and seem to have benefit of stabilizing the disease in patients on prior gene therapy.
From that standpoint, we need to see both the competitive landscape of which gene therapy is providing what kind of benefit, and then also getting more data over the next six months to nine months to see, do we stabilize disease? As these patients get older and older, the anticipation is that they would start to decline if we're stabilizing the disease. That would be a very strong support of adding the gene therapy patients into a phase three study. A more straightforward design would be to have patients on a background of steroid, which would be the classical kind of structure. I would anticipate that we would treat older patients in the decline phase because the decline phase is where we can see the greatest delta between sevasemten and the actual placebo group.
We would want to see additional data coming out of the open label and see disease stabilization. We also want to discuss with the agency, and we'll be doing that end of this year, beginning of next year, of what would be the endpoints. We're speaking to both the EMA and we're speaking to the agency, the FDA, in regards to what would be the primary endpoint. We favor the SV95 because the SV95 can be very highly quantitative. It's a wearable device. It's approved in Europe as a primary endpoint for Duchenne. We saw some of our best data and most differentiating data using SV95 as the primary endpoint.
We also can use North Star, but as you might guess, North Star in patients who are 4-6 years of age, trying to get a four-year-old to do something repetitively of 17 different actions is difficult and highly subjective. We would favor using the SV95 as a primary endpoint, but we have to talk to the agency about that. There are multiple companies speaking with the agency right now about the SV95. We do not have any information of how they are going to react to utilizing that as a primary endpoint. Ultimately, we will speak with the FDA on that as well as looking at the data.
Yeah, I mean, the Duchenne data we put out in July of this year has actually matured really well in the sense that we've watched what's happened in the background for whether it's a gene therapy or now we've seen with one of the HDACs out there, we've seen three deaths, right? The side effect profile is really problematic for kids who are going on this. I think we almost generated some of the best North Star data and some of the best SV95 data in Duchenne, and that's matured really well. I think it's to Kevin's point, it's really trying to get alignment with the FDA. What's the duration of the study? What's the primary endpoint? Obviously waiting to see what happens with Grand Canyon before we make that investment in order to move forward with that phase III.
Since you brought up Grand Canyon, that's going to read out sometime in the end of the year.
In the fourth quarter of 2026, yeah.
Depending upon how that plays out, do you want to commercialize our DMD medicine, or do you want to become a primary cardiovascular company?
You know, it's interesting. We just yesterday walked through our commercial plan. We have $520 million in the bank. We've got.
60.
560, sorry. 560 in the bank.
I could take the 40.
You could take the 40. You can have extra 40. We have $560 million in the bank. That gets us through 2028, and that includes a long-term backer. We are set up. It is a lean commercial lift for us. We know where the patients are. We know how to drive them to care centers. There is just, there is no other therapy available. I think it is a multi-billion dollar opportunity globally. I think we are the only company that is active in Becker right now with a pivotal study. I think, yes, the plan is work towards commercialization, and whatever happens on our way there happens, right?
Awesome. Thank you for your time. Unfortunately, 25 minutes is never enough. I appreciate the time. Thank you for making the trip over.
Of course.
Looking forward to a really exciting 2026.
Yeah, great. Thank you.
Thank you. Thank you so much, Kevin.
Thank you all for coming.