Welcome everyone to our chat with Edgewise Therapeutics.
I'm Joseph Schwartz from the BioPharma Equity Research team at Leerink Partners, and it's my pleasure to be joined by Kevin Koch, CEO, and Behrad Derakhshan, Chief Operating Officer.
Thanks so much for joining us for the update today.
Thank you, Joe.
Maybe, Kevin, if you wouldn't mind just starting us off with a quick level set on Edgewise's recent accomplishments and goals for the next 6-12 months.
Yeah. No, it's exciting year for the company, really. In fact, we just came back from the Muscular Dystrophy Association meeting. Had two really important posters, one where we showed a benefit on some of the cardiovascular endpoints, early days from the CANYON study, which I think is really important from the standpoint of utilization of drug, commercialization of the drug, the unmet medical need.
That data showed that in patients who had a low ejection fraction, we actually saw increases in ejection fraction in those patients. Also, NT-proBNP, which is a biomarker associated with heart failure. In fact, the patients who were normal had no changes in their NT-proBNP, but the patients who were not on sevasemten actually had increases, so a negative effect on those biomarkers.
That was really nice data we showed. The other really great piece we showed in Becker is that the open label extension, the patients who are on that open label extension all remained progression free. Really stopped the disease in its tracks. Of course, our big readout for Becker will be our Phase III, which is on track for disclosure by the end of the year.
\We believe that is a position to have, be able to file for an application. It's exciting year in the Becker space. Of course, I think the most near term excitement is coming from our cardiovascular program, 7500 for the treatment of hypertrophic cardiomyopathy.
We put out a short high level summary of some of the observations in the Part D portion of our study, which is the 12-week study, in both obstructive HCM and non-obstructive HCM. We showed which was the key differentiating factor, that we had no changes in ejection fraction with increasing dose and concentration of the drug, as we have seen previously, and also that the safety profile was tolerable. That was important.
That led to a lot of renewed and additional interest in the company, and now we've provided some guidance on the data release we'll have in the end of the Q2 this year for 7500. We stopped screening patients in December.
Our last person who initiated dosing was in January.
It takes about four months to get through the dosing, through the core lab evaluation, and get it out the door. We'll be providing about, 50-60 patients' data.
Mm-hmm.
Right now, we haven't provided an exact number. Probably a few extra in non-obstructive. Really productive year, really important year for the company. With 7500, we're planning to move to Phase III, and we are currently in discussions with the FDA. I think a couple of rounds to get them engaged, and then we'll have a final end of Phase II meeting to lock down the Phase III protocols.
Fantastic. Thanks for that great introduction. Let's stick with 7500 to start. Can you tell us what constitutes a clear Phase II win for 7500 in your mind that could justify rapid Phase III initiation now in the next-
I think it's a key, it's a novel mechanism. As the CMIs have this challenging aspect of their mechanism that they need to lower ejection fraction, which means that they have to drive a heart failure phenotype to see efficacy.
That leads to two really important events. One is if you tell a patient, "I have cardiomyopathy and I have heart failure," but I may give them more heart failure, that doesn't make the patient feel better. Second piece is that leads to a black box warning and a REMS, and that REMS limits the uptake of those drugs into centers of excellence where they have the staff, the bureaucracy, the echo capability to get people to that optimal dose.
I think what we show is that essentially our key differentiating factor is not having that ejection fraction change, potentially not having a REMS, and expanding the market of NHCM out to the community cardiologists.
The community cardiologist is where most of the HCM patients reside. I think that's really the value proposition. Now, for a clear win, we need to continue to show what we've shown before, no changes in ejection fraction. We need to show that as we saw in the first in April, that we see the same levels of efficacy that we saw there because they were outstanding, and then we have no safety signals that would keep us from moving the drug forward.
Yep. Makes sense. Is there anything in the upcoming 12-week data that we should be looking for or you might emphasize beyond what we've already seen, or is it just?
Yeah, I'd be very similar, more of the same and a path to Phase III. I think we'll provide our best estimate and guess of how we can run that Phase III, and what, what would be the timing of that. I think we're still committed to initiating first patient in by the end of this year.
Okay.
The other thing, Joe, that is a little bit different than what we previously talked about in Part B and C is remember those were fixed doses, and so you kind of saw efficacy at the fixed dose. The question is, are you able to resolve, metrics, whether it be a biomarker or a feel and function beyond a fixed dose? What I mean is, let's say I'm treating an obstructive patient in Part D. Let's say their gradient response, so their gradient at rest goes below 30, at Valsalva below 50. That's the criteria that defines a responder in the obstructive group.
Physician at 50 could decide, "Okay, I'm done dosing that individual 'cause they met that threshold." But if that individual is an NYHA Class II or their NT-proBNP hasn't quite reached the threshold of normal, now the physician has the discretion to say, "You know what? I'm going to move to the 100 and see if I can get better efficacy out of that patient." You're driving this phenomenon that Kevin has introduced kind of earlier in the year around getting people as close to normal as you can by not hyper-focusing on one metric, but rather looking at a number of metrics together and saying, "How close can I get this individual to being as normal as they can?
Essentially a responder analysis for multiple endpoints of normal, which is below 150 NT-proBNP, New York Heart Association class I, KCCQ north of 90, and of course, the gradients of the obstructive patients at a level below threshold. How many patients can you get to that place? 'Cause they're essentially normal.
Right
At that point. Normal is important.
Yes.
Now, what we're thinking about is, so how would we position this product in the community docs that would allow it to expand the utilization of the drug? We have to make it as easy as possible for the physician to get someone to the optimal dose, but without an echo. Our base case here is to have an echo at baseline, to have an echo at end of study.
With that strategy, I don't need to have an echo to get them to the right dose. I'll use other measures of feel and function, which is actually in the AHA guidelines, you dose patients to how they feel, not dosing patients to gradient, not dosing patients to other measures. It's really how they feel. If you can do that's a huge win.
That shift is huge, right? Because you're essentially going in an environment where with the current standard of molecules used in that space, they're not used to making a decision independent of LVEF. It's always in the back of their mind. In the early days of mavacamten, what we would hear is, "I'm dosing to the minimally efficacious dose because I'm always worried that someone's going to have an LVEF drop, and now I've got a different problem that I have to solve." Like, removing LVEF out of the equation gives the physician a lot of freedom to focus on optimizing responders, patient response.
That's a different shift than what they're currently doing. I think it'll be a big differentiator in non-obstructive in particular, when you don't have gradients to guide your response. You're really determined.
You're really relying on feel and function and biomarkers to drive your response rate.
That's interesting. Could that shift the long-term role of sarcomere modulation in HCM and take it from just a titratable tool to maybe a disease-modifying therapy?
You know, preclinically, these are always things preclinically you have to take them with a grain of salt, but we've shown robust data in remodeling of the heart in the R403Q pig model of non-obstructive.
That was with five months of dosing. There we saw that the left atria, which is typically enlarged in these animals, was decreased, which is one of the key endpoints for that remodeling, and we did this both by MRI and by echo measures. Really important observations in these longer term studies in the pig.
I think ultimately this type of mechanism, because no one's really tested the hypothesis of a drug mechanism that leads with diastolic potential, meaning the primary efficacy of this drug is being driven by relaxation of the ventricle during filling, which allows more oxygenated blood to be pumped. If you don't have an ejection fraction change and you allow for greater filling of the ventricle, the patient should feel better.
If they feel better, that'll reflect itself in the New York Heart Association changes and the KCCQ changes. You could have a dramatic effect on how the patient feels, which is really the goal of every treating physician for their patient.
Yep. That makes sense. Interesting. So, do you think zero incidence of LVEF below 50% is sufficient to credibly argue against a REMS requirement, or do you think an even broader safety margin might be required?
You know, we've to the extent we can do this, we have seen and actually reported a wide ranges in concentration of the drug that did not see ejection fraction changes. Obviously, ejection fraction as a measure in and of itself has 5%-10% variability. We've demonstrated that in the placebo groups in our normal healthy volunteers. What happens if one case goes below 50? It's probably.
You have to ask, is it durable? Is it symptomatic? as we described in December, we've not seen that case yet. The dataset we've produced of, over, well, it's over probably 100 patients now, if you throw in the normal healthy volunteers, that continues to be the trend.
I think as the ends continue to grow, you become more and more comfortable. We have talked to former FDA administrators and advisors, and they have looked at the dataset and have made the comment that this is sufficient data to de-risk the potential for REMS, given that it's a novel mechanism, and you're not seeing the changes in ejection fraction.
Cool. Some of the early cohorts treated with 7500 showed some AFib events, but you've also cited that there was around a 10% baseline AFib during screening, and they were obviously very sick patients and not titrated.
How will the upcoming dataset separate baseline disease pathology from potentially drug-related AFib? How definitive can this dataset be?
Yeah
in your mind versus having to wait.
Yeah, you're trying to prove a negative.
Phase III.
You're trying to prove a negative, but I think our strategy of incorporating the Zio Patch. The Zio Patch monitors very precisely changes in arrhythmias and among many other activities that can be measured with the Zio Patch.
What you're citing is that we saw a significant number of patients who had actually screened into the study. They met all of our other criteria, but they did not have AFib, but yet they had AFib in our screening paradigm where before they had gotten drug.
Huh.
What it tells you is that AFib is pervasive in this population. When you say just HCM as a population, it's well known that the patients have a significant amount of AFib, and there are a number of comorbidities associated with it.
I think you also could argue that, because of the interest in this therapeutic indication, many of the patients that are available in the U.S. have higher risk of AFib. I think this Zio Patch observation tells you that this is probably higher than you think of people who are actually in clinical trials in the U.S. I think what it does is, it puts a level of quantitation of what we're seeing.
I still believe, and we've had this consistent message, that if you look at the placebo groups in Phase III studies, in both obstructive and non-obstructive, you see rates anywhere from 2% up to 8% or 9%. For ODYSSEY-HCM was 8%, EXPLORER-HCM was 8%, SEQUOIA was in the 2+% . You're also seeing in real world with utilization of Mava, you're seeing 15% rates of AFib.
There's tremendous amounts of AFib in this population. Now, how is that different relative to ejection fraction changes? AFib, if you're a community cardiologist, you have to deal with AFib every day. In HFpEF and heart failure, AFib can be up to 40% or 50% of the patient population.
Most treating cardiologists know how to deal with it because it's part of their practice.
They just know it, and it's symptomatic. You know when you have it. Ejection fraction is a silent killer. You don't know when you have it until you have edema and you're in heart failure. That's the difference.
Right.
You don't know it. You're all of a sudden in trouble. That's a big deal. That's, I think, a key differentiating factor, and I think, I expect this to be in the background rates with the Part D data.
Yeah. We've consistently said 2%-8% is in the background, and so that's the same guidance we gave last year, and we're going to be giving the same guidance this year.
Okay
coming out of this.
That's helpful.
Yeah.
Cool. Let's talk about NHCM for a moment. How do you view the potential read-through of the upcoming ACACIA-HCM study?
Yeah
To EDG-7500?
Sure. I think there's a couple of things. We've largely already designed kind of our Phase III. We have a very good idea of what we want to do. We've got a couple of options that we're considering. We've a lot of the learnings that we have in non-obstructive actually came out of what we saw from ODYSSEY, right? There's a couple of interesting observations there.
One is, how broad do you go in terms of number of sites? Because I think the broader you go, what we see, the trend is that the fidelity of the KCCQ seems to get worse. We've seen that in broad HFpEF trials too, right? If you go really broad, you introduce a lot of noise in the system. We're working with sites who have a lot of experience in doing KCCQ.
We did that already in Part D, but we've doubled down on that as we've thought about kind of the Phase III initiation. The other thing is really what happens there is almost independent to us because I think we have a mechanistic rationale that's a little bit different to the CMIs, in that this drug works preferentially during early diastole, so we speed the rate of relaxation.
Some of the best data that we have that kind of talks to that is, as you remember, we studied the 25 milligram dose in a single dose in part B and C. What we learned there is at 25 milligrams, you see NT-proBNP drop 30% plus within a week, but gradient doesn't move in the first week.
It takes 2 weeks for the gradient to respond to the 25 milligrams, whereas you've already hit your peak NT-proBNP at week 1 of dosing with the 25 milligram. That tells you that you're having your diastolic effect preferentially, and then your hemodynamic response trails that. That's kind of an interesting observation.
Remember, we designed the drug because when we developed this particular lead series and this mechanism, we noticed that at low calcium, which is the diastolic portion of the cycle, you saw greater potency.
Yeah.
Now we've demonstrated the exact same result in humans.
Yeah.
I think that's an important aspect of it, and that is the key pathology in non-obstructive.
Yeah.
You don't have the obstruction.
Mm-hmm.
The flow is being gated by physical obstruction.
Right.
You have the wall thickness and how compliant the wall is an important aspect, and if you can relax that ventricle, you'll fill the ventricle with more blood.
Mm-hmm.
Better off for the patient.
Yeah. Kind of how does it impact us? I think all it does it makes us look at is there any operational change that we have to make on how we conduct the study versus anything fundamental around the type of study that we're running and the expectations that we have on efficacy? 'Cause we're really relying on our own data to power the study. We're not relying on whether it's ONSET or ACACIA-HCM to think about how to do that.
Okay. In the scenario where ACACIA-HCM demonstrates strong efficacy with minimal LVEF compromise, what would be EDG-7500's differentiation thesis?
I think you have to think about in that scenario, where does aficamten get used, right? We already know they have a schedule that's very closely tied to when you adjust the dose, you have to perform an echo. I think if you look at the experience with Mava, BMS has done a very nice job at kind of REMS certifying a lot of community cardiologists.
The problem is you don't find a lot of community cardiologists prescribing mavacamten, mostly because they don't want to deal with the echo burden, right? It's an onerous management that they have to do, and they have to hire staff that can do the focal echos. I don't think that problem has been solved with-
I mean, realistically, Mava has robust efficacy in obstructive HCM.
Yeah.
Continues not to be used.
In the broader community.
Broader community.
Right.
You would have to even have better than what you're seeing in obstructive to somehow dislodge current practice and change current practice in the community docs.
The other thing which I think is a little bit more fundamental is in obstructive, you have the benefit of measuring gradients. So you know if your patient is responding because you're seeing their gradients drop below 30 or below 50 rest Valsalva. In non-obstructive, I think you're going to be limited because you're dosing until you see a drop in ejection fraction.
You're dosing on MTD of heart failure.
Yeah. Okay.
We're shifting the philosophy to say, "Hey, what if you could dose to max efficacy in your patient because you don't have to worry about LVEF dropping?
Yeah. Okay. How much of a desire among community cardiologists do you sense that they have to treat HCM?
All the market research we've done suggests that not only do they have the patients that they wish they could do more for them, but right now what your average community guy is doing, they wait for them to become symptomatic and then they refer them to either an HCM center or a tertiary center that can manage them.
The problem is a lot of the times, that's not conducive with how patients want to be managed 'cause maybe they're not close to a center of excellence. It's difficult to kind of wait for the patient to progress, to push them somewhere, and ask them to make a commitment to go somewhere and get monitored and try and get into an echo schedule, get them to an efficacious dose, have the liability that you might drop ejection fraction. It's a tough sell.
I think if you could give them something that is much easier to manage, then I think you fundamentally shift that thinking. And the best way to think about it is your community guys manage a lot of heart failure patients, right? A lot of their decisions on how a patient responds is guided by what happens to their NYHA class? Is this patient NYHA class two, three, one? If you develop a drug which allows them to not worry about LVEF and focus on, can I get them to NYHA class one because it's a tool that they use in clinical practice today, that's not what we're seeing.
That's not what's going to happen with the CMI class of molecules, I don't think, 'cause you're always going to be pushing up against LVEF.
Yep. Thinking about EDG-15400 in HFpEF, you have a Phase I healthy volunteer study with data coming out soon. What PK/PD signal are you looking for?
Well, we have an idea of from the preclinical models, we can, in some respects, track what we've seen with 7500 on the PD side and track that back into 15400. Because from a preclinical standpoint, the molecule is very similar, and we still see the same diastolic benefit that we saw in the preclinical pig model.
The concentration of the drug and how we do this function and look at the free fraction of the drug and look at concentrations, we feel like you'll see a window of where we can work and can we dose the drug.
I think the other aspect of this is as we have a view of the metabolic profile of this drug is more amenable to a older HFpEF population than the HCM population.
That is something that we'd like to confirm, clinically, as well as with additional preclinical data, and that would position that product as a separate entity. I think it's still. It's become less important since the IRA has been modified to allow multiple orphans within one drug class. But I still think from the standpoint of the size of the HFpEF population, and the need and the heterogeneity of that population, you want a drug with a very precise profile, metabolic profile, so you can combine the drug with multiple other entities, that-
There's, of course, the price point difference, right?
Yeah.
Like, you've got an orphan pricing in HCM versus you're kind of in the SGLT2 GLP-1 pricing range when you look at HFpEF, and you're treating 3.5 million people.
I'm comfortable developing, moving both drugs forward. Clearly, 7500 could move into HFpEF as well. I think it still makes sense to have two different drugs.
Okay. Interesting. Well, with a few minutes left, I want to make sure we can talk about sevasemten.
Mm-hmm.
Yeah, you brought up the recent data updates. I guess with GRAND CANYON coming up later this year, I know that's powered for 1.7-point NSAA change. You've indicated the tipping point might be closer to 0.7. Can we talk about clinical meaningfulness of those kinds of magnitudes of changes on NSAA?
Yeah. I mean, we've been consistent, and the physicians have been consistent, that a 1-point change is important. Think about it this way, if you're trying to walk across a street and you want to lift your leg to go on a curb, can't make it. These patients sometimes lose their balance. They fall. It's a point.
Mm-hmm.
If you're, going to the bathroom, need to get up, you need to call somebody to get you up. That's a point. You're on the floor. You can't lift yourself off the floor by yourself. That's a point.
Right.
That's a real thing.
Yeah.
That's why.
For a person that is often typically.
Yeah
has a family, they have a job, but they know they're slowly but surely declining, that puts so much pressure on the family as well as personally for the patient. I think even a point would be valuable. Now, what we saw in the poster that we provided is that over a multiyear period, now we're 3 and 4 years, we see stabilization of disease, no progression of the disease, and that's now five points.
When you start talking about five points on a North Star scale, now you start talking about, "I can walk," or, "I can't walk without some kind of device," or, "I even need a scooter to get around.
Mm.
Not fully non-ambulatory, but 5 points is going from, "I'm walking" or, "I'm not.
Mm-hmm.
That's a pretty huge event in one's life. I think that we can show, and I think even more importantly, as we look at these open label studies, you're seeing this theoretical stabilization, and our natural history is saying that they should be declining by 5 or 6 points. I think that's really important.
I think that weighs in the calculus from the regulatory agencies of, yes, a point in and of itself at a year is perhaps, you know.
A thing.
is something, but it's that it progresses consistently and homogeneously, and that's a bad thing. We're building this narrative.
There's now four or five databases of natural history, and we're working with the agency to develop a way to validate that work using patient-level data to understand that that is the consistent observation, especially when in our target population, where these patients are starting to decline. When they start to decline, they start to decline in a homogeneous way, and they continue to decline. That data.
I mean, you cover a lot of neuromuscular companies. Tell me one in your portfolio that shows you've arrested the disease. I've never seen anything like it.
Most drugs in this space, they slowed the decline.
Yeah.
We just showed that there was no progression.
Well, I look forward to the data later this year.
Well-
It's going to be exciting. Thanks so much for the update, guys.
Thank you.
Thank you.