one of the biotech analysts here at the firm. It is my pleasure to introduce our next presenting company, Stoke Therapeutics. From the company here, we have two executives. We have the Chief Executive Officer, Ian Smith. We also have Chief Patient Officer, Jason Hoitt. Ian, thank you so much. Gentlemen, thank you so much for spending time with us. Perhaps, Ian, you can maybe start us off, make some opening comments, talk about the milestones or the key things that we should be focusing on, and then we'll go into a Q&A.
Yeah, okay. Thank you, Yatin, and thank you for people that are joining us today here in person and those that may be joining us online. Stoke is a company that's focused on genetic medicines for genetic diseases. It has an approach where it upregulates protein from a healthy allele to address the underlying cause of genetic diseases.
We're primarily focused in genetic epilepsies, but we also have medicines in ophthalmic areas as well, and I'm sure we'll talk about those today. The lead medicine is a medicine for Dravet syndrome, which is a disease that, unfortunately, children develop normally through potentially 18 months to two years of age, and then plateau in terms of neurodevelopment for the rest of their lives.
These children suffer from severe motor seizures, as well as learning and neurocognitive disabilities. Our medicine fortunately works to reduce those seizures that they have, but then also potentially show, based on data that we've seen, gives them gain of function to actually lift them up off that plateauing of
potentially a two-year-old. If you're a 10-year-old and you go onto our medicine, we start to add back tasks and function in terms of cognition and behavior to these children. We're in phase III study at this point in time. We'll provide you with an update today, as we did with the investors just a couple of weeks ago in our Q3 call. The study's progressing very nicely .
As of a week or so ago, when we had our Q3 call, we had zero discontinuations. We've enrolled 162 patients, and as you know, Yatin, two-thirds of the patients are already through the primary endpoint of the study, which is at week 28. So, it's progressing very nicely. In terms of milestones in the next year, the next 12-24 months, we have a phase III study readout that should be in Q3 of 2027.
And if that data is supportive, a potential approval in early 2028, then obviously the launch would be within that 2028 period into a patient population that is sizable for a rare disease. And the commercial opportunity for us is significant given that we potentially have a disease-modifying medicine into this disease population. Close behind Dravet, we have a medicine that is in a phase I/II single-dose escalation study in ADOA, which is a genetic progressive loss of eyesight.
And our target there is OPA1, where we look to increase the expression of OPA1 to actually give back eyesight, potentially give back eyesight into these patients that are progressively losing their eyesight and ultimately go legally blind. So company has progressed incredibly well over the last year and continues to do so, and we are looking forward to into 2027 to be a very important year.
Very good, Ian. Thank you for that. Perhaps in the beginning, we will spend some time on the disease itself and the long-term data that you have generated. So Dravet, I think, as a syndrome is beyond, let us say, seizure, right? When we talk to physician, and I think even caregiver, they also talk about the cognitive behavior burden that it possesses. So the question is, you have generated now long-term data, so what has that four-year worth of data from a long-term perspective has sort of taught you about potential to change the course of disease in these patients?
Yeah, thanks for the question, Yatin. Just for everybody's awareness, we ran phase I/II studies in these Dravet children ages two through 18, and we gave them the opportunity to, after the dose escalation studies in the phase I/II, gave them the opportunity to move into an open-label study. Approximately 90% of the patients that were in the phase I/II chose to move into the open-label study, and approximately 70% to 80% of those patients are still in the OLE study four years later.
So the OLE study now has been run for four years. I want to be clear that this is a chronically dosed medicine. It is an ASO, and it is dosed every four months into these children. And this is a disease where these children are suffering seizures, but also do not develop cognitive and neurodevelopment. It is stunted.
What we have seen from that four-year data now of chronic dosing in these patients, and the patient numbers are still around 60 patients, we have seen this durable reduction at doses that are similar to those that we are studying in the phase III of around 80% from their baseline. So seizure reduction around 80%, and that is durable out through four years now. So incredible data in terms of durability of response in seizures.
And then from the cognition and behavioral aspects of these children, we are seeing this gain of task and function as measured by Vineland. Vineland is a measuring tool that allows you to measure receptive communication, expressive communication, interpersonal skills, personal skills. And what we have seen is this accumulation of task and function, gain of function through this four-year period.
And what I mean through the four-year period is that they have more gain of function in year two than they do year one, and more in year three than they had in year two, and more in year four than they had in year three. So we are actually in this rare position where we are running a 52-week registration study, phase III registration study, yet we have data in nearly as many patients as we are studying in the phase III that is supportive of the medicine in terms of seizure reduction and improvement of cognition and behavior. So it is a unique position to have that much data while running the phase III study.
And the other thing that we take confidence from the OLE study is all these patients, well, not all the patients, but most of the patients have been on drug a long time now, where anywhere between three and four years in the four-year OLE study. So we take away confidence on the safety profile of the drug in that we have not seen anything that would create any concern in terms of as we move forward within the phase III.
Got it. Very good. Very helpful, Ian Smith. Now let us just focus on the EMPEROR phase III study. I think you have over-enrolled and have provided some sort of color in terms of how well the translation, how many patients are continuing into the OLE. Can you maybe talk about what is the aim for this study? What would you like to show? How would you characterize a successful outcome, both on the primary endpoint and the Vineland score that you are looking at, Vineland endpoint?
Success is hitting a P-value of less than 0.05, to be honest, and that is reduction in seizures as the primary endpoint at week 28, but also hitting a P-value of less than 0.05 in the cognition and behavioral aspect of measuring via Vineland. That endpoint is one that is measured by taking a combination of different Vineland domains, measuring each domain separately. So there are four domains you measure each four separately. You actually add them together, divide by four, it gives you an average, and that is called an MC4, multicomponent four-
Yeah
score of Vineland. That is the key secondary endpoint that would be important for us to achieve a P-value of less than 0.05. What that would do is it would provide us with a label to allow us to promote the medicine as a seizure reduction and cognition and behavioral benefit drug, i.e., a disease-modifying drug, promote the medicine as a disease-modifying medicine. That is the importance of getting the P-values and having this data on the label.
Got it. So perhaps on this modification in the Vineland, the way it is calculated, I mean, we have gotten some inbound from investors, particularly focus on why this modification is made and why this might be a better way to capture the disease modification in Dravet syndrome, if you can just comment on both of them.
Yeah, sure. It is a good question, frankly, and it is the most frequently asked question that we have had in probably the last three or four weeks since we put out a press release on the outcome with the FDA. So I need to help people understand that where we have come from and where we are today. The protocol in the SAP was initially agreed in terms of a hierarchical assessment of five key Vineland sub-domains. What that means in regulatory FDA speak is that you have to hit the Vineland domains in a priority and in an order.
Yeah.
What that means, if you have got five, if you miss the first Vineland domain in terms of a P-value, even if you hit the next four, they do not count.
Yes.
You would have gotten what's called a nominal P-value, and there would have been some debate about whether it's allowed on the label. Usually nominal P-values aren't, so it may not have made the label. From our perspective, we believe that was the wrong way to measure cognition and behavior with this disease.
We went to the FDA, had a good discussion, and this is following now two years of thorough engagement with the FDA on what Dravet syndrome is, as well as what our medicine is doing for these children with Dravet. The FDA agreed with us in that where we came out was rather than having a hierarchical assessment that could have been a little bit of a guesswork in terms of can you hit them in order, we came out with a basket of four Vineland endpoints. It's a basket.
It means you measure each endpoint individually. You then add them together and you divide by four, and it gives you an MC4 score, a Vineland score, which is based over those four different measures. That, as well as taking away maybe the potential randomness of a hierarchy, it gives you the basket of measures to look at overall.
Another nice byproduct of it is, it enhances the powering of the study. Recently on our Q3 conference call, we provided some data to help people understand how much it enhances the powering of that secondary endpoint. The real win here is that you get to measure cognition and behavior in these children in four different ways rather than in a hierarchical individual methodology. It certainly increases the confidence of hitting that cognition and behavior endpoint to be able to have it defined on the label.
Got it. I think for us and investor, I think understanding a certain percentage reduction in seizure frequency is rather straightforward, right? Here, what is a minimum detectable difference, and how would a caregiver or a clinician consider it? What is clinically meaningful here?
Regarding seizures?
Regarding this Vineland MCP.
Oh, Vineland. Sorry, yeah. It's a good question, but I'd say a score of anywhere between one and two.
To be honest, our study is powered at this point in time, if you take 150 evaluable patients, and remember the study's 162, it has zero discontinuation at this point. But if you take 150 evaluable patients, we need to achieve at least a 1.8 improvement treatment effect in the Vineland score, the MC4 score, and that's for a powering of 90%.
Yeah
If you ask clinically meaningful, one to two. And it's worth just help people understand what Vineland is, and maybe I could just give a quick explanation of what one to two actually means. The way that Vineland is administered, you could take one of the domains, and it's administered via a neuropsychologist asking a caregiver, generally a parent, and consistently the same parent, a list of questions.
But a question could be, "Does your child identify three objects?" And the answer might be never, you would score a zero. It might be sometimes, and you would score a one. And then it could also be usually, and that would score two. And so, if you're able at baseline ask a parent whether their child can identify three objects and they say never, they score a zero. Now roll forward 52 weeks and they are on our medicine, and let's say potentially they can identify three objects and it happens usually, frequently, then that child would now be scored a two.
I see.
That would be known as a treatment effect from baseline, but you compare it to sham in the registration study. But that would be 2 points. But if you can imagine, the neuropsych will also say, "Can you name the three objects?" They do not want to just hear that it is three objects. The parent will say, "It is a cat, it is a dog, it is a table." But moving a child that does not have the ability or the cognition or behavioral abilities to be able to name a table, a dog, or a cat, and moving them to being able to do it usually-
Yeah
i.e., most of the time, is a huge benefit to a parent, a caregiver. I am just giving you one example, but there are lists that go to anywhere to 30 and 50 questions within these domains. They do vary from, can a child walk? Can a child sit up unassisted? Does a child recognize their own name? Does a child use the word mom, dad?
Yeah.
They may sound simple. They are very objective, the questions. They may sound simple, but they are very meaningful. These small gains, if you can imagine having a child that is 10 years old, but having the neurocognition and behavioral age of an 18-month to two-year-old, and moving them up into being something that potentially could be like a five or six-year-old
Yeah
that is a massive gain of function and task and cognition and behavioral aspects for these children and their families. That is why this endpoint is so meaningful. Yes, it leads to penetration in terms of a launch of a disease and will help with pricing of the disease. But this is really meaningful to these patients and their families.
How quickly are you able to detect these differences? Because the other point, or the question that we get is the time, right? You are evaluating Vineland at 52 weeks. I think in the earlier analysis we have seen at least 18 months. So, is 52 weeks enough to capture?
Well, unfortunately, we have not run the same level of dosing we have in a 52-week period in our registration study. We have got data for that same level of dosing in our registration study, but it was measured Vineland at week 68.
Yeah.
We saw cognition and behavioral gains in the various domains of between 8 and 11. That was at week 68.
Yeah
Which is close to week 52. If you then actually say, "Well, what data do you have at week 52?" We do have data at week 52, it is just that it was sub-optimally dosed. It was basically half the level of dose that we are providing a child over this 52-week period of the registration study. When you run the cognition and behavioral MC4,
on that week 52 data, again, it hits a P-value of less than 0.001. So, we have a high confidence of hitting the MC4, the multi-component four. I believe it is 0.01.
Yeah.
Yeah, I just want to make sure I get the numbers right. So, we have data that is supportive as well, but at month 12. I am not sure whether this is a function of dosing or duration.
I kind of look at it as it does appear as though a drug has a dose response. We get deeper seizure reductions with a higher dose. We get greater cognition and behavioral benefits with a higher dose. So, I look at the registration study that is dosing two times 70 milligrams and two times 45 milligrams as being a dose that is going to drive a statistically significant response in the MC4.
Got it. Very helpful. Then with regard to the patient population in the EMPEROR study, could you maybe compare and contrast the severity of these patients relative to the phase I-B, sorry, I/II background therapy? Just trying to understand how.
Yeah, they are actually very similar. As good drug development, as I am sure you are all aware, the phase III is planned off the data we have in the phase I/II and the OLE study. We want to ensure the patients are consistent between the phase I/II and the OLE that go into our phase III study.
Yeah.
Part of the phase III study actually has an eight-week screening period, where we measure baselines of various characteristics of these patients to ensure that consistency between the phase I/II and the OLE. So, for example, obviously the patients need to be between two and 18 years of age.
Yeah.
But some of the key criteria is they need to have a certain level of seizures to be allowed into the phase III study, that were consistent to the phase I/II and the OLEs. They need to be on a stable background of anti-seizure medicine. So, to be clear, our medicine is dosed on top of standard of care anti-seizure medicines.
And so, the patients that went into the phase I/II and the OLE study were on anywhere between three and five standard of care anti-seizure medicines. And again, that's part of the screening criteria going into the phase III. So, there is a very high consistency between the patients that we've planned the study from and that we're performing the study on.
Got it. And then, what is the official timing of the data? I think you're going to be submitting pieces of NDA throughout 2027 in the first half of 2027. What are those module that you will be submitting?
So, for people to build a timeline for data, last patient in to the 52-week study was in June 2026, which means last patient out in a 52-week study is June 2027. Data will obviously have to be QC'd, quality checked, everything else, and that would be when the top-line data comes. So, that would be in Q3 of 2027. We then would anticipate completing our NDA filing shortly thereafter, with a potential approval in early 2028.
Got it. And then-
We have breakthrough designation. It's a pediatric disease, and we would anticipate hopefully an expedited review.
You would be entitled for a PRV as well.
We would be entitled to a PRV.
Got it.
Yeah.
I understand that if you hit the primary endpoint on seizure, you have a drug that will be approved. Just from a labeling and commercial standpoint, what are critical elements that you would like to be included in the label, various endpoints? What about the age limit? Should there be any cutoff?
Yeah, so you are asking specifically about pricing. I will first of all say that the pricing of the medicine is not about the phase III label. It is actually about the totality of the data.
What I mean by totality, it is the phase I/II data, it is the four-year OLE data that will be five-year OLE data by the time that we submit, and it is about obviously getting an approvable drug with hitting a P-value on the primary endpoint in the registration study. Jason Hoitt, who is with me here, is our Chief Patient Officer, but he has done a lot of work, and maybe in his last couple of minutes, he can comment on that work we have done about why it is about the totality of the data and not about a label for pricing.
Yeah, absolutely. I think one thing we have heard from payers loud and clear over the course of this year, and even frankly before, is that they are not constrained by the label, that they are going to look at the totality of the evidence. When we asked them specifically, at the time of a potential approval, we laid out different label scenarios.
We asked, "What would be the most compelling piece of information that you could have in hand at the time of potential approval to help you appropriately reflect coverage for your members?" Almost to a person, we heard the long-term open-label extension data, because as you are writing a medical policy- Knowing and understanding the long-term safety and efficacy is critically important. That is the same as what we heard from doctors.
I think from a label perspective, I think we feel confident in the way that this study was designed in EMPEROR and that we will see statistical significance across the endpoints. That being said, I don't think it's as binary as Vineland hitting or missing to whether or not we're going to see uptake and/or the ability to price at a rare disease-modifying genetically targeted medicine range to the point of the analogs that we've shared publicly.
Got it. What about the age limit? Should it be I mean, should there be an upper end of the age limit?
Given that it's a genetic medicine, we've got breakthrough designation, it should be two years and above.
Two years and above. Can you maybe talk about the addressable patient population?
Yeah. From an epi perspective, you're looking at about 16,000 total patients in the U.S. We've said previously that we see there being about 6,000 immediately addressable at the time of a potential approval. From the epi, that reflects the population that's 25 and younger, and therefore likely in the care of a pediatric epileptologist or a pediatric provider that are most commonly diagnosing and treating the disease, and therefore most on the cutting edge of recent advancements in the disease.
But it also tracks to the total number of patients. If you look at claims data, you see there are about 6,000 total patients that have ever gotten an ICD-10 code for Dravet in the U.S. There are about 6,000 whether you look at it from an epi perspective, or from a claims perspective, claims being across all ages, but epi 25 and younger.
Got it. Very good. Maybe in the last couple minutes, on your second program, the ADOA, I think you've gone through cohort one and two, cohort three and four likely to be therapeutic. The question is, when we see the data, how would you characterize a first readout on visual function and biomarker?
Yeah. ADOA is autosomal dominant optic atrophy. It's a genetic progressive loss of eyesight, all the way through to legally blind. We are in a phase I/II study, single dose escalation. We're through our first two cohorts that were primarily for safety. They were at 0.1 milligrams and 0.3 milligrams. We have just commenced the third cohort, which is at 0.5 milligrams.
If we move through that safely, we will commence a 0.7 milligram cohort four dose, which would be starting in early 2027. Cohort three, you're correct, Yatin. Cohort three and four, certainly cohort three is entering the therapeutic level of dosing that we anticipate based on our PD work pre-clinically. The cohort four is right in what we would term the therapeutic area, dosing level.
Given that cohort three and four, we should have safety efficacy data in the first half of 2027. Safety efficacy data, obviously safety would be the dosing profile and the patient response. Efficacy will be measured via low contrast visual acuity in terms of measuring a number of letters of potentially gaining a number of letters.
As I said, these patients unfortunately lose their eyesight. We're hoping to increase mitochondrial function and therefore increase the energy in the cell of the eye, and therefore increase vision or improve vision or bring it back. We're also using a biomarker, FPF, which measures mitochondrial activity and we see it as a correlation to mitochondrial activity, which gives back the potential to give back eyesight.
Very good. That's all I had for you guys, gentlemen. Thank you so much.
Thank you, Yatin.
Thank you.
Thank you for listening in.