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

Thanks, everyone, for joining us today. For this session, we have Edgewise Therapeutics. And for those that don't know them, on stage today, I have their CEO, Kevin; and their Chief Business Officer, Behrad, with us. Kevin, thanks for being here.

So I wanted to jump right into it. You recently just read out some really intriguing data from the DUNE study. In that study, it had arms focused on McArdle, limb-girdle but also on Becker's. And can you talk about maybe what you learned about sevasemten with respect to Becker's disorder from the DUNE study?

Yes. So it was a great study. I mean it's really -- I would call it a mechanistic study. It was run by a very well-known investigator from Copenhagen, especially in adult neurology. And what he has been known for is trying to understand the relationship between exercise and biomarker data associated with that exercise. And as you probably know, in both Duchenne and Becker but other muscular dystrophies, they have elevated creatine kinase. So they have muscle damage associated with activity, and contraction of the muscles leads to damage. So what he's been doing is trying to quantitate that across multiple different indications. And so what we decided to do is a single-site study and evaluate patient populations where things like creatine kinase were elevated with exercise, and he's published papers on this. So what we did was just to describe the trial, we had -- the patient would come in. They would get a resting biomarker, which would be creatine kinase or troponin I2, which is a biomarker associated specifically with fast skeletal muscle. And at rest, they would get a baseline. Then they would exercise the patient. That exercise challenge was about a 20-minute 80% of maximal bicycle challenge followed by a leg press challenge. So that was a fairly rigorous study and exercise challenge. The expectation is that you would see an increase in the biomarkers because of that exercise, okay? So those are the baselines, a rest and an exercise. We then started dosing sevasemten, 5506, we've just got our name for a drug, for 3 months. And then you repeat that experiment. So before they do the exercise, before they take the drug, they get a baseline or resting. After 3 months of drug, you give them exactly the same exercise, and you look at the effect on the exercise. And what we saw was -- we did this placebo controlled. So there were 6 patients -- 6 treated patients and 3 placebos. And so what we saw was a statistically significant change in biomarkers looking at rest to rest. So basically, that validates the biomarker work that we had seen in the ARCH open-label study, and then we measured exercise to exercise. And in both cases, what we saw was about a 45 or so change in creatine kinase, which is very analogous to what we observed in ARCH, and then an 80% change in TNNI2, which we also saw in ARCH. But -- so we saw that both at rest over a 3-month period but also with exercise. So why is that really important? So first, it's the first time we've studied patients in a placebo-controlled way where we've demonstrated clear differentiation from placebo and benefit on the muscle damage biomarker. I think the other key is that our CANYON study -- now CANYON is a Phase II study in Becker muscular dystrophy with a defined patient population. And the primary endpoint of CANYON is CK, so the biomarker. So what we've done is actually demonstrate that -- and increase the probability of success that we will hit on the primary endpoint of our CANYON study. So that's really important because you'll hit the primary, and that's -- I think that's now a higher probability of success. But I think hitting the primary for CANYON was a requirement. We're going to the agency and showing them the entire data package that this drug has benefit in Becker patients. There is no standard of care in Becker patients. There's no treatment for Becker patients. The drug has been shown to be well tolerated in the Becker population. And so these patients desperately need a drug or treatment. And so -- and as we've said in the past, if we see a trend towards benefit on the North Star endpoint, which is a functional endpoint, and we see a trend in things like MRI and patient-reported outcomes, we would go to the agency and discuss with them the opportunity for accelerated approval for this drug. So now we've hit -- I think we have increased the probability of us going to the agency and having the discussion about how we might actually have this drug approved for an accelerated approval. The other aspect, and we gave an update of our various studies, is that GRAND CANYON is recruiting well. We started enrolling GRAND CANYON, which is our pivotal study last October. And we anticipate sometime fourth quarter, first quarter of '25 to have that fully enrolled. So when we're reading out the CANYON study, we will have essentially a follow-up study, a pivotal study fully enrolled that we can also support the safety of the drug and that we will test the clinical hypothesis.

Got it. A lot there. Maybe want to drill down on CANYON a little bit. So you mentioned, as we understand, the primary point there is CK, but a lot of focus on function. So what do you think is possible to show on function on NSAA when you get to the CANYON readout? Do you think you can hit stat sig? Obviously, you have some powering assumptions from how you've designed GRAND CANYON. I guess how are you thinking about the likelihood of that occurring in CANYON itself?

Well, the ARCH data was informative. Over a 2-year period, we saw an increase in function in Becker patients. And based on their natural history, we would have expected them to decline by as much as 2.4 points. So we had almost a 3-point delta at 2 years. This is a 1-year study. So I think we would anticipate that we would see more than 1 point in change between placebo and treated. This study is 30 treated patients and placebos. So that was the design of the study for 1 year. So I would say that we're not powered for North Star statistical significance. I think the trend would be very powerful. We saw many of the patients improve function. So I think that result, I think, has a reasonable probability of success in at least getting a conversation with the agency. But we are not powered for that. And -- but I think directionally, since there is no treatment, it's certainly something we could discuss with them.

Got it.

What's really nice about kind of the data that's set to support the North Star [ dimension credits ], if you look at the Duchenne space versus the Becker space, natural history in Duchenne is a little bit outdated. But it's also confounding because if you look at that data, it's kids at different ages and different points in their disease trajectory with different steroid regimens. And they can be on their growth curve or they could be on their decline curve, and it's really difficult to power a trial when you've got that much of a heterogeneous natural history. With Becker, there's a couple of things that work in your favor. One is you've got 3 independent studies that have been generated in the last 5 years. So it's not like you're working off of old data. you're working off of current data. You also have patient population that doesn't have a standard of care. So there's no steroid use in these individuals. There's no drugs approved for Becker. So when you look at the natural history and you use the cutoffs that we've used to kind of select patients in Becker, which has been a limitation in the Becker field, we didn't have the data to think about, is there a subpopulation that has a decline, so we can power a trial. The consistency is between 5 and 32 individuals are declining 1.2 points on an annual basis. So that is current data, which helps us feel confident that going into kind of the data readouts, both CANYON and ultimately GRAND CANYON, that you've got a chance to at least see a trend with CANYON. And certainly, we power it for -- overpowering it in GRAND CANYON.

I think the other key aspect of this is that there's been a lot of discussion of, well, is North Star a good measure. I think you need to think about the context of when you use North Star. Now when you think about it, North Star in a 4- to 70-year-old population where you have 17 different activities, just think about convincing a 4-year-old to do 17 consecutive activities and how quantitative that might be with a 4-year-old. Now think about it, and what would I do if I went to an adult. And can an adult actually perform the North Star in a meaningful way? And so I think for a -- and this comes from our data of screening to baseline, that the ability to reproduce the North Star in the adult population is much higher. And in fact, and even in the Duchenne populations, many of you've seen givinostat was approved by the agency. They were treating older kids 6 to 15 years of age. The vast majority were in kind of the 7 to 9 range. Now we have an older kid performing a 17-point scale. It's going to be easier than a 4-year-old. So I don't think it has anything to do with the measure. I think it has to do with the population chosen at that time to actually do the measure. And I think also this idea of taking patients that might be a more homogeneous group and a decline is something where you can blunt that decline and show statistical significance. So I think we've really enhanced our confidence of the potential of seeing something meaningful in both CANYON and especially in GRAND CANYON.

Okay. On the regulatory side with respect to Becker, obviously, you guys are in the CDER division, which has historically been somewhat more conservative. I'm sure you guys read through the givinostat documents as well. There's some pushback there. I guess how are you thinking about how amenable the agency actually is to an approval -- an earlier approval of a small molecule drug like yours? I guess can you comment on is -- the nature of the tone of the regulators you've had so far?

I think we've had both consultants and KOLs in with at the meeting and been surprised at the level of interest about the mechanism and the curiosity. I think it comes back to this -- the first prerequisite is the safety profile. So we've shown now for 2 years and from our other safety database that we have a relatively well-tolerated molecule. So that's the prerequisite for success. I think the second level of these plasma biomarkers and the utilization of this muscle damage effect we see in these diseases, I think the agency would argue that they don't disagree that muscle damage is something you would want to alleviate. They understand that muscle damage precedes loss of muscle, which precedes functional loss. Their main issue with these surrogate markers is that can you measure them quantitatively. So what we've done and shown now that the DUNE data, you can run a placebo-controlled study and demonstrate statistical significance and not see crazy wide variability in at least the CK biomarker as well as the TNNI2 biomarker. So now the other question that they asked us was there's an observation if you take population. So if you take a Duchenne kid, you take 5-, 6-, 7-year-olds, they will have CKs of 20,000. If you look out after the disease has progressed and you look out when they're 9 or 10 or 11 years of age, their CK has dropped. So the disease progression follows decreases in biomarker. Now you think about usually increases of biomarker are what people expect when the disease progresses, but you're decreasing the biomarker because they have less muscle, okay? But if you think about the Becker population over a 1-year period, there should really be no decline. It's a slow progressing disease. So you should have a minimal decline in the placebo within the noise of the assay, yet we're going to have deep declines like we showed in ARCH with our biomarker. So that should be a stark difference between that -- the decrease in biomarkers associated with a drug effect and not just the natural progression of the disease. So I think we can support this notion that the biomarker result and the quantitation of that biomarker result is real. We'll support that with MRI data, and that's going to be differentiated from the natural history of the patients within our study. So all of those things together, I think, make a strong argument for which -- and this is really the argument we want to have and the discussion with the agency we want to have. Is blocking muscle damage good? And is it likely to predict clinical benefit? I want to have that discussion with the agency. I can't understand why decreasing muscle damage wouldn't be a good thing.

Makes sense.

Is there anything [indiscernible] question the validity of the biomarkers? Or is it just sort of not [indiscernible]?

It's the variability. And so the variability is driven by activity. So -- and there are some good papers on this that if you have a Duchenne kid in particular and you have them off running on a jungle gym, you will see -- 2 to 4 hours later, you're going to see an increase in creatine kinase. On the flip side, if you tell the patient to be quiet and don't move around and have them do that for hours, they -- you'll have a lower CK. So the level of the biomarker can be driven by activity. So now -- but what now we've shown is that in this controlled data, you can measure that. And especially in the Duchenne studies and in the GRAND CANYON study, we have an activity monitor called a side device associated with these measures. We don't have this in CANYON, but the side device measures what's called the stride velocity 95th percentile, which is kind of the maximum stride velocity that you can see in a child or in a Becker patient. And that stride velocity 95th percentile is a primary endpoint in the EMA for Duchenne. And there's been a lot of work validating the correlation of stride velocity 95th percentile with North Star time to stand and force their climb. So I think that -- I think there's beginning to come up with an aggregate of data of how you quantitate both function and then tie that back to biomarker simultaneously.