Good morning, everyone. Welcome to the Jefferies Global Healthcare Conference 2025. With me today, I have the great pleasure of introducing Praxis Precision Medicines CEO and dear friend, Marcio Souza. Marcio, how are you doing this morning?
Doing great. Living the dream.
Amazing. We'll just jump right into it because there's so much exciting work you guys are doing at Praxis. Top of mind, you've developed several really promising sodium channel modulators through your platform. At a high level, can you give us what stage we're at in clinical development of vormatrigine and relutrigine?
Absolutely. As you mentioned, I appreciate being here as well. It's always fun to be in this conference and meet so many people. As mentioned, right, we have a platform of really trying to, and I think at this point we can see it's succeeding at transforming or translating some quite interesting ideas from inhibition, excitation, balance in the brain, neuronal networks into molecules. It's no secret to anyone that sodium channels are incredibly important on that. Some of them are actually fundamental for that here. Coming from that idea, we're like, okay, how can we do this better? Because we know they work really well. Certain ways to, on that case, completely block the currents are very effective, but they come with a lot of other potential issues.
That was the original idea for the series by which relutrigine, our candidate for developmental and epileptic encephalopathies, was found, and then vormatrigine. One could argue there are many other potential indications and adjacencies for this kind of molecule, and some of them you might be hearing a little bit in the future from us about. Vormatrigine, particularly, we're developing for what we call adults, probably a little bit more restrictive than what it really is, right? It's more for common, I would say, epilepsies, for example, focal onset seizures and generalized epilepsies as well. We are running two studies right now. They're going really well in all aspects. I would say the first aspect, they want to make sure we get the best possible patients on those studies, and we're doing a really good job on doing that together with other partners that work.
That very first one to read out's coming up mid-year, which kind of upon us is Radiant. What we expect to see there on Radiant is, I would say it's a large enough cohort of patients that it should be clear what you're seeing in general from a pharmacology perspective since we are collecting a fair bit of PK on that study to further understand that in the patient population. Of course, safety, but quite importantly as well, what happened to seizures, right? The process we took for Radiant, I think it's quite important based on some of the questions we got from investors recently, from you, Emma, recently.
It's very rigorous, and I think what we end up on the other side of that now is with one fairly, what I'm going to call, high screen failure rates coming into the study, which is, I see as a positive here, right? Really being quite strict in terms of what you're trying to get. A population that by and large is very similar to the most recent focal onset seizure studies conducted. We are pretty happy with that. We set the bar from a seizure reduction before 40-60%. We know there's a lot can happen in that range or beyond the range, and we're looking forward to reporting that pretty soon. By the end of the year, we expect to report the larger study for focal onset seizures as well. It's called POWER1, and by next year, POWER2.
If you fast forward, I know it sometimes lives month by month in this business, but when you fast forward the next 12 to 18 months, we're talking about having a registration package completed for vormatrigine, which is incredibly exciting.
Yeah, absolutely. Maybe you took us a little bit through the, taking a step back, the history of development of relutrigine and then vormatrigine. High level, what are the differences between the two molecules? How does it impact the pharmacology and ultimately what indications they serve best?
Yeah. Intentionally, we're looking into the patient first, right, when you're developing these drugs. And when you look into an adult patient with adults, I'll go and say epilepsy in general, but normally focal or generalized here, those are people who are attempting or have a life like us, like a relatively, I'm going to call normal life, if normal even exists. They're like work or they're attempting to work or they have like going to school. It was important to have the pharmacological properties or the drug-like properties that match that, right? Like a drug that can give peace of mind to the prescriber, first and foremost, for the patient reducing their seizures, being very convenient when you are delivering. Like once a day, no needs to be with or without foods. We know life is complicated in general, act very fast.
Like within the few hours to the few days here, it had obviously a profound impact. Maybe what we underestimated beforehand was the impact of the drug-drug interactions. When you look into that before, that was not a major factor. I think right now, the more and more we hear about this community, both prescribers and patients, we see how important that is. We're actually in a super lucky position here to have a very clean DDI profile with vormatrigine. That is a little bit what I would characterize as an upside down view of this market, right? There are about 3 million patients in the U.S. with adults on co-epilepsy, and the vast majority of them are undertreated or really poorly controlled. That's what these drugs, these next generation drugs as we call, would be used.
When you go back to relutrigine, the story was very different, right? We're talking about kids that are seizing tens to hundreds of times per day. They have G-tubes, J-tubes. They have all sorts of other medications. They're very, very fragile. It was important to consider all of that when you're developing. While the molecules, both vormatrigine and relutrigine, are incredibly potent, they're very similar in terms of how they interact with the channel. They produce quite remarkable preclinical and early clinical work. Relutrigine needed to be very easy to be given in a different way. It's like a suspension. It can be given a different dose for different body size and weight. We adapted on that regard. It's also developed towards the orphan drug markets. They're, of course, advantaged as well from a regulatory perspective, from a speed to markets and so on.
Independent of how you look, each one of them has very, very significant market opportunities in front of them and potential for really disrupting this market. You might have seen, I'm sure you've seen, you wrote about this, the EMBOLD results from last year, which in our last updates a few weeks back, patients reach about 90% seizure reduction after 11 weeks.
Remarkable.
It has never been seen before. I would actually argue beyond remarkable there.
Before we get to those beyond remarkable results, I know most top of mind is going to be RADIANT readout soon by Medea, right, for vormatrigine. What was the rationale behind entering an open label study? Because that's fairly unique in both focal and generalized epilepsy in phase for the first phase two.
A number of things, right? We had to make choices as we're doing, as we're going along in drug development. One of that is the kinds of burden that we're going to put patients through for the kind of data that we require. What do we need to learn? There are two major things, and there may be many more at this point in time, that we wanted to learn from Radiants. The first one is it's a very straightforward pharmacology relationship when you can just collect in an open label, you know what they were given, you know their compliance, and it can inform the POPPK model and other models in a very straightforward manner. That was one of them. Doing open label made that very simple.
The second was magnitudes of like, well, how much can we actually get with 30 milligrams in terms of seizure reduction in a broad number of patients? We could translate that into this dose for POWER2, which had not started yet. Very confident with the POWER1 dose selection that we did based on all the work that was done before. One must ask, right? When we're talking about a market with millions of patients, to expect that it's just like a relatively simple experiment is going to get to all the angles there probably is very arrogant. If you can understand patients' response, can we get everyone to a close or complete response, or maybe some patients need more, or no, we're good at 30 milligrams, and maybe that's the top dose ever for this drug. That's what we wanted to sort out.
We believe Radiant's going to be quite informative.
In terms of the, discussed the patient baselines a little bit, or at least alluded to high screen failure rate being a good indicator for the trial. You have allowed inclusion of patients on background sodium channel blockers to enter the study. Talk me through a little bit of the rationale behind that and kind of why it's smart to include those patients.
Yeah. If you go back to the work that was done with anti-seizure medications and like the impact on patients with focal or generalized, some restriction on generalized, but let's focus on focal onset seizures for a second. The Human Epilepsy Project did like this beautiful publication a few years back that shows that if you could give nothing else to these patients, the bottom line is you should give a sodium channel blocker, right? I know there's a lot of stuff like hand waving on different mechanisms, on different things, but that is what happens, right? It's like a lot of those patients like take it because you cannot have a seizure, period, without those channels being like active, hyperactive in this case. So we knew like the importance there. Of course, it's obvious that they would be on certain backgrounds.
The question that we had was twofold. One is, does it matter from an efficacy perspective? I think all the preclinical work we did, all the structural biology work we did, understanding binding sites, dynamics, currents, everything that you can imagine says that no, it does not. We can absolutely drive all the efficacy we expect for vormatrigine independently of the background. It was theoretical until a few months back, right? Once we actually showed the results from involves. I know I am borrowing from relutrigine here, but they are basically the same problem statement where we are seeing patients at supratherapeutic dose of incredibly strong sodium channel blockers where we would expect only toxicity for an extra blockage, no extra efficacy. What we are seeing is seizure freedom, a lot of seizure reduction, and no safety liability.
I think I would call that case closed for now. Are there going to be cases that a combination of a drug is not going to be tolerated? Yes. It is absolutely not because of mechanistically.
Great. In terms of the split of patients entering the study between focal and generalized, around what are you expecting? Beyond maybe talking about maybe benchmarks in both indications, this is just an eight-week study. Power One is going to be longer. How should we think about efficacy and safety in Radiant at eight weeks versus Power One, which will be at 12 weeks?
Yeah. The baselines, and of course, Power One is not done yet. I am going to say the baselines right now for Radiant, which we know really well, and the projected Power One baseline so far, are incredibly similar, which is a very good thing in my view. It is also very similar to other baselines out there. I think we have like a good basis to begin with. I think what we know from studies being conducted recently is time actually is a good thing, meaning like if you put a little bit more weeks, we are going to get a little bit more efficacy out of that. If you are very efficacious towards the ends and that is more of a trend downwards, you are going to have a lot more efficacy.
One way to say, and this is by no means a hedging or a caveat, is probably RADIANT is the worst we can see here. Meaning, if it is very good as we expect it to be, I think we should be incredibly confident that the four extra weeks on POWER1 are only going to make things better, not worse for that.
In terms of safety, similar expectation that safety rates could be higher at 12 weeks or no, you're going to see plateauing of safety on these.
Most of the events, right? Most publications, like if you look into the recent published work, and we tend to agree what's happening there as well, things happen in the first week or two. The majority of the things, like there's idiosyncratic events for most of the drugs, we don't expect that with vormatrigine, but you do have like significant like urinary retention with potassium openers. We do have like dry eyes with some others or Stevens-Johnson syndrome or you name it, which none of those are expected with vormatrigine, but those tend to be like idiosyncratic and by definition, you cannot predict. The vast majority, like a recent publication, I'm sure you guys can figure out which one, there's like a Kaplan-Meier like curve I'm going to call.
You can see all the events for discontinuation happen in the first week and a half due to adverse events. I'm not even sure that is the drug itself as much as it's the patients feeling this additional uneasiness, and that's what happens. It is basically an indirect evidence that more weeks should not make the safety any different, but certainly has a potential for the efficacy to be even better.
Great. How is enrollment proceeding for POWER1 and for POWER2? Maybe talk a little bit more about how RADIANT is going to be informative of dose selection there.
Yeah, absolutely. Power One, maybe I'll take a step back here as well, right? We do get a lot of questions in terms of the enrollment pace and already touched on some points like screen failure rates is very high. The process that is used with external reviewers of classification of seizures is very robust. It's still like this is significantly higher than some of the other enrollment like trials at the same time. Kind of what are we guys doing? There are many things and some of them I'm just not going to be giving like tips for others. One that is quite interesting, we talked about Empower before, right? It's our program to understand epilepsy in the United States and in a sense, invite patients that would pre-qualify for these studies for the sites where the sites are running.
We put that under the umbrella of the Energy program right now. That's basically the way it looks. If you look into that umbrella, there are over 5,000 patients that have been referred, consented, confirmation of ICD codes, and the vast majority of them released their medical records. We're able to verify there's a neurologist that goes through each one of those records and so on. It is a disproportional amount of people that actually were able, people living with epilepsy, they're able to look into how they are manifesting. All of that would be an interesting academic work. What is actually super interesting from a practical perspective is that about 30% of the patients getting into Radiants or Power One are coming from this initiative.
No one else dares to put the money and the resource, human resource, technical resource systems to invest in something like this. At the day we did it, it was a gamble. We just didn't know if this would work out. Here we are with like a very significant database and growing, where we believe by the end of the enrollment, Power One and Power Two for that matter, about 30% of the patients in the U.S. would have come from that initiative. We have 70% direct on the sites, either for referrals from their networks or the networks that we facilitate, and then 30% from here. In the grand scheme of things, it's about a year of recruitment for a regular epilepsy study that you shave off by having that. I know your question is just how well it's going.
I think I kind of answered in a little bit more words than how well it's going. It's going really well. We're excited.
Amazing. If there's any other points over matching, we'd be happy to, of course, listen to them. If not, we could switch over to rilutrigine because the results from the Bold cohort were beyond remarkable. Maybe we could just segue into that.
Yeah. So relutrigine, the way we decided to run the developments, the first phase of developments was first to try this drug on SCN2A and SCN8A mutations that leads to the epilepsy. It's a little bit upside down in a sense that those are incredibly difficult. One of the most, if not the most difficult disease to treat. We wanted to set the bar is like if this can be treated, if we can actually go through this effectively, it opens like a lot for other disease in general, maybe even for other disease in general. We're quite successful, right? It's the 16-week study, over 40% seizure reduction, phenomenal results in terms of the other aspects and the other aspects like communication, alertness, that disruptive behavior, which is super important for this patient. We checked the box big time last year.
There are still a few questions missing. One is what happens when you keep them longer, right? As you've seen about three weeks back, they get better. They get better. Their family situation gets better. Everything gets better there. Another check the box. When we're seeing over 30% seizure freedom in the patients on the first cohorts, what we decided to do is say instead of going through the run of the mill, finish day study, then start a new study and so on, these patients have no time for that, which is very important for us as a company. We simply added cohorts to that as registration of cohorts, right? That is going very well. The enrollments, incredibly well. Screen failure is even higher in these patients for different reasons. For example, they die during screening, right? This is a very, very severe condition.
They have to change multiple medications. They go to the ICU. They have the status. Not exactly the same reasons why in other studies, but it's going really, really well. Very proud of the team at Praxis and our partners that are running that. We expect by no later than early next year to have the results for these cohorts. That would serve as registration or study. In the meantime, we did a lot of work preclinically, mechanistically to understand what else can this go. I think as I stand here, sitting here today, it's very hard to see where this couldn't go when you talk about seizures. We decided to talk to the FDA, talk about a study that would really treat like seizures in these patients with disease. We call that Emeralds. That study started.
We have that up and running right now. We're going to be getting patients in imminently. If the interest is an indicator, this might end up being the fastest enrolling study we ever had because it's quite overwhelming, the interest that we have across the board. It's going to be a global study. It is a global study. Multiple geographies that we're recruiting, we'll be recruiting right now. It's only the U.S. that's going to be open in the next few weeks. That should be done relatively quick. That comes right after, which would be hopefully a first NDA for relutrigine next year as a supplemental NDA that brings the opportunity for like several hundred million to maybe $1 billion with the first indication to multiple billion dollars in potential revenues here. Again, incredibly exciting to have yet another drug with such large potential.
Yeah, absolutely. Maybe touching a little bit of base on EMBOLD cohort two and then the EMERALD study. In EMBOLD cohort one, placebo was quite low. And logically, I could kind of understand why it would be. But what would you say your relative expectations are for placebo response in cohort two?
Yeah. In well-run trials and pediatric like in disease in general, despite a little bit of a rhetoric out there, if you actually go and look into the numbers, which I always prefer, it's incredibly low in general. There were two, unfortunately, not successful studies in LGS done recently in Ravina that the placebo rate was 6%. I actually would argue this is no difference than 1 or 2% and 6%, but very, very small. Some others like a little bit higher, but never really canonical out of control as some people want us to believe. It's a blinded study. Of course, we monitor the blinded study right now. I don't think there's any reason to believe that we're just going to be a completely different placebo response rate.
I think it's very disrespectful to actually call this a placebo rate when the kids having the seizure have no idea if they're taking an investigational medicine. There's an observation error that we all have, observation error in everything in life. And that error is probably around 2-5%.
Yeah. And then in terms of the Emerald study, is it registrational? And can you compare and contrast how severe these patients are compared to the ultra rare SCN2A and 8A patients who are extremely severe?
Yeah, more diverse. And I mean diverse etiologically, group of patients. I think it's fair to assume they're going to be less severe from a seizure count perspective. I think all indications we have right now is that they're going to be significantly less severe. What we're talking is we had over 50 countable motor seizures per month on the previous study for these patients. Meaning if there's even small type of seizures, that wouldn't be countable. They're canonical less complex ones for those patients. Just imagine one seizure. I hope no one in this room or in the webcast ever had one, which is unlikely, but maybe one seizure can change your life completely. Imagine having 50 in four weeks. There has not never been a study that severe disease.
We will expect the next ones to be a little bit less severe than that.
Yeah. Maybe talking a little bit about, Praxis is much more than just sodium channel modulators. You also have an ASO platform, Elsunersen being top of mind. How may that therapy be complementary with relutrigine in SCN2A patients?
Yeah, no, absolutely. I think the example that is the easiest to borrow for everyone is what happened with SMA, right? If you go back, how one therapy that's going to be, it's very effective. Like now, right now we have multimodal therapies, all mechanistically doing very similar things or basically the same thing, right? With gene therapy or AAV-based gene therapy and our splicing modulator and an ASO, for example. We do not see that very differently here. Those diseases are very complex. They are incredibly heterogeneous. Every indicator we have is that the combination of those drugs might actually bring extra benefits for these patients, for eight parts of those patients. Not everyone is going to be able to go into an intrathecal delivery like once a month, of course.
The markets in general is broader for a neuro drug, but that's going to be a lot there. We do have, I would say, and I believe that's the very first time I'm going to be talking about this. We do have patients under emergency access combining those drugs right now. In due time, we're going to have certain preliminary results for those cases to talk about how it's going. You can imagine that we wouldn't be excited. I wouldn't be talking to you about this as I am if I thought the results were not exactly in the direction we expect them to be.
No, absolutely. It's amazing what you're providing to those patients. Maybe just high level, can you walk us again through the just numerous catalysts you have from now until early 2026?
Sounds good. We only have 50 seconds. I'm not sure if it's going to be possible. Just kidding. Radiant, it's coming up, of course, a lot of the focus. In general, I'm going to pair that with, it's mid-year. I'll pair that with POWER1 later in the year. That kind of constitutes a lot of the evidence for the vormatrigine program. I don't want to skip over the fact that we are pretty close to nominating the ASO PRAX-100 for loss-of-function in the 2A gene that leads to the autism spectrum disorder phenotype, incredibly large and might need incredibly large. Like overall, like scientific interesting, we showed in our R&D day a couple of weeks back, over about to over 200% increase in protein expression. That is the benchmark for Praxis.
If you're going to get to a heparin sufficiency mark, you're going to get to 200. We think that anything below that is suboptimal. It's probably not going to be successful in the clinic. We're seeing gap programs coming up next as well with something similar and then our PCDH19. On the new things, those are the ones. I know it's on everyone's model as a zero, but it's then important for us to read out the Essential3 program for essential tremor that's coming up as well soon in the following months. We are going to see what happens there overall with the program, either wrap it up or move full-blown to a potential NDA if it's positive. Quite a binary event as well that is not in any one model.
Many, many things for next year as well that I think we just run out of time. I'm going to leave for our next conversation.
Looking forward to it. Thank you so much.