Good morning, everyone. Thanks for joining us today. I'm Ami Fadia, Biotech Analyst at Needham. It's my pleasure to be kicking off the day with Marcio Souza. He's the President and CEO of Praxis Precision Medicines, sorry about that. I'll like to turn it over to Marcio to give us a quick introduction of the company, and then we can jump straight into the fireside chat.
Thanks, Ami. I like the ambition there, like upgrading us a little bits, as a company. It's like, and good morning, everyone, and I hope everyone's having a good beginning of the week. So happy to be here. Praxis, we're a CNS-based company, like really looking for, like, solve some of the bigger issues, in CNS and brain disease in general.
We have currently 4 programs in the clinic, a number of programs, preclinical as well. I'm sure we're gonna be discussing, some of them today, including our phase 3 programs in essential tremor, which are gonna be reading out later this year. Very exciting with our epilepsy portfolio as well, with 628s we're developing for focal onset seizures, and a couple other applications that we see for that program, in the future.
In our rare disease portfolio, specifically with PRAX-562 and PRAX-222, we're going for this very high unmet needs, untreatable, developmental encephalopathies. So I'll hand it back to you, but it's an exciting year for us, and I look forward to have the discussion today.
Okay. Thanks, Marcio. I guess, the big focus, most recently for Praxis is on PRAX-628 for focal epilepsy. You presented some very interesting data on it very recently. Can you maybe just start by talking about, you know, the mechanism of the drug and what makes it so interesting?
Yeah, absolutely. So 628 is quite fascinating, actually, as a molecule itself and as a potential treatment here for seizures in general, because it acts in a different way than pretty much any other drug that's been developed before. While the actual target itself is well known, the ability to modulate sodium channels has been a puzzle, I would say, for many of us doing drug development for decades and decades, right? There are a number of drugs on the class that completely block the channel, which has been very useful, and I think that's what got us to where we are today in terms of treating these patients.
But it always came with a huge drawback that, at the moments like you dump in, hyper-excitability in general and the action potentials in general get reduced. It comes with a lot of tolerabilities and side effects. So the ideal drug would be one that can reduce the overall activity.
So in case of the seizure conditions like epilepsy, you wouldn't have it, or you would reduce to a minimum the number of crises and would allow patients to have higher functions, as we are having right now in this discussion without feeling like either dizzy or somnolent or having psychiatric manifestations or many other side effects that are here existing. When you look into the market itself, the majority of the patients are not achieving the resolution of seizures because of side effects.
While it might not manifest necessarily, directly as a side effect, it, it is by the virtue of not getting to therapeutic levels. So when you look into the, the 628's development so far, we had some exquisite preclinical data, right? It's the most potent, molecule ever developed, like when you look into MES, EC50, for example. And then we moved into a phase one, which basically continued to, to look very good, both our single-ascending and multiple-ascending, data, very high exposure levels as we expected without those side effects that, we just discussed.
We made a decision last year to add on top of the EEG data, the quantitative EEG data that we had already for the phase one, knowing the drug's in the brain, knowing it's engaging, to add an intermediate step, what I'll call a phase II-A , in patients who have a response, patients with epilepsy, with a diagnosis of epilepsy that also manifests a reaction to exposure to lights.
So when you expose them to certain frequencies, they have this generalized reaction on their EEG that you can measure on the EEG that is pre-seizure. And in the past, drugs that have been shown to reduce or eliminate the signal moved on to be quite successful, and while they were successful on managing the seizures, they also had the same issue with in terms of safety discussed before.
So for a drug like 628 where we can be successful in both is incredibly exciting. So we just discussed the results of this study like a little bit over a week ago, I guess, and was exactly what we were expecting to be very effective on eliminating the signal there on the EEG and fairly safe as well for this stage of development. So we're very excited. We're moving the program to late stage development now. We intend to start the first registration study later this year following shortly thereafter by the second study as well. So very exciting times for that and with the real intent here to resolve this long-lasting issue that our patients are still having seizures after so many treatments.
So what you're saying is that this is sort of a very predictable model, that one can rely on and has been used for developing other drugs. And in some ways it's predictable what, you know, type of profile you can see at a later stage in development.
Right.
You know, perhaps can you kind of talk about how we should interpret this very compelling data that we saw from the seven patients and how, perhaps, you know, the FDA may view that data as supportive of future development? Do you think you need to go and discuss that with the FDA at all, or can you move sort of to a larger trial from here?
Yeah, for the most part, this data was really to de-risk the program internally, for us, right? Philosophically, the way they were looking to drug development is you should be adding events that de-risk. So when you are allocating capital, we know that the risk is getting better and better, the risk profile, as it goes along, in development. So that was key for us, right?
Like when we were designing these experiments and we might have discussed that, last year, I mean, it's not, an experiment to some extent that completely predicts the rate of response, in the future. So we're not looking for we had like a 100% response, right, on this. Does that mean you're gonna have a 100% response in focal seizures? Knock on wood that we do, but, that is not the relationship that we're expecting.
It's more whether or not we're gonna see a significant reduction as is being seen before. Equally important for us was the therapeutic index, right? It's whether or not we're gonna get there with the best possible safety the compound get. So that isn't really, I would say, a topic of discussion for the agency right now other than the fact that it looks pretty safe in patients. And I think it's a really big deal before we go to a bigger study that already have that exposure on top of other drugs as well because this study was not only standalone like other drugs. Some of them that were supposed to eliminate the signal and was not, which reinforced the idea that patients are under-treated for many reasons. They are not responding to the therapies right now.
The entire package becomes quite encouraging and quite important for us and for our investors, of course, one more check in the box that the drug's doing what we expect it to do. And I think that's incredibly important right now.
Now, you tested a 15 and 45-mg dose in the study. You know, at a recent conference call, you mentioned that you know, going forward, you may take at least a 20-mg dose forward. Can you talk about kind of the rationale of picking the 15 and 45-mg? Were they kind of bookends of what you wanted to test and then why 20? Why not test higher or actually why not even go forward with 15 because you did see pretty reasonably good responses?
Yeah, no, absolutely. So the biggest difference between what we did with the PPR study, that is the one we've been discussing, and what we are intending to do with a focal onset seizure study is that the on the PPR is a very simple, I would say, once so we just give once the drug, right? And we measure the time course throughout the day using like patients are hooked up to the EEG. So you see the entire course of that. What you don't see there is what actually happens, what I would say in real life as you're taking the drug every day. So when you give 15 and 45, how did you come up with that, right?
We wanted to create a range of exposures, right, that would give us a good idea on where we should be from an exposure standpoint on the following study. Is that perfect? Absolutely not. But when you use the data that we have in animal models, the data from the phase I, and now the data from this PPR study, we know that at certain levels of concentration, we are right on the therapeutic zone.
You look into 628's, the equivalent EC50 from the MES when you extrapolate to humans are on the high 20s nanograms per mL. When you're giving, for example, 20 milligrams, we are 10 to 15-fold higher than that. So while we are and we got this question a fair bit in the last few days, like, why not go higher and so on? We are higher. We're higher than anyone ever being.
Like if you were to benchmark by what other competitors, other drugs in the class could have done, we'd be talking about single digits, milligrams here. We'll be talking about 1, 2, 3 milligrams. So by going with 20, having a drug that is both expected to be very effective and that delivers because we've done the multiple dosing with 20 and 30, we know to the extent that the data is limited by the number of subjects, but can show us it's safe, it is what this patient needs, right? They don't need only a drug that solves the seizure, but one that is fairly safe as well. Now, luckily, we don't have to offset one by the other.
So we can deliver both here, right, with very high exposures, giving 20 and 30 milligrams, and at the same time, those that are expected to be pretty safe.
Yeah.
That's the rationale.
That makes sense. I guess, you know, before we sort of talk about the next steps on 628, I think the one question is, you know, in the - in this sort of patient population, which is sort of a subset of the larger patient population that you're targeting within focal epilepsy, how representative are these patients, of the broader focal epilepsy population, in terms of the type of efficacy that we saw here? I understand that this is a smaller subset and, you know, we may not see 100% responses.
We should not sort of expect that. But again, to some extent, you know, this model is very predictable what we see in larger trials. So how should we interpret what kind of the, you know, the ultimate clinical profile of this drug could look like? What does that mean for where it would fit within the treatment paradigm, for focal epilepsy?
Yeah, so if you look into, I think a very simple way to look into focal epilepsy is that it's a signal that starts somewhere in the brain, right, that creates this, unfortunately, may I say, that seizure condition that can or cannot generalize and create a number of manifestations for these patients. So in that regard, it's actually very similar to what we did, right?
We simulated the visual cortex and we've seen a generalization and we stopped that generalization. So if you were to oversimplify it, it would be like a one-to-one relationship here. Now we know we humans are a lot more complex than that. So it might not be exactly. But we do know that it's expected to be quite efficacious. Now we're looking to the treatment paradigm right now.
In the U.S. and elsewhere, we have a couple classes of drugs that are pretty much dominating the markets, right? Like the SV2As and sodium channel modulation in general is where you get 85%+ of the market using those agents. The majority of the patients are failing for safety firsts. And why we record this, the failure as efficacy was really because they couldn't be exposed to levels that were efficacious of those drugs or because they couldn't tolerate and they stopped before that. So that is really the limitation that we have right now.
When you start a patient who are like being seen by a neurologist for example, versus a level three epilepsy center, as ideally all patients will be seen by experts, but they are not, that there are several limitations in terms of the care. They're gonna start with the easiest drug to give, right? And unfortunately, a lot of those patients fail.
So most of the drugs in development right now are pretty good. There are other companies developing fantastic drugs for epilepsy, and they're needed. Why? Because there are over 2 million patients in the US. Between 300,000 and 500,000 are failing every year. So it's not getting any better, right? So we need to add new drugs that are gonna make these patients' lives better. But everyone is talking about a third line.
So they failed multiple times and now you're adding at the ends. A drug with the profile of 628, of course, is gonna start there, but it's incredibly likely it's gonna move down the adoption quicker than any other drug that I know of in development goods. And that is the, I think that's what is beautiful here, to be honest, with this profile is that would give that peace of mind.
And you might have heard in our webcast last week, one of like the key experts in the country talking about the excitement is actually to move down knowing patients that day one would be covered because of the exposure profile instead of letting them go home and maybe not knowing if they're gonna have another seizure, which is not only traumatic, but can be fatal, as well. The ability to adopt this drug broadly is one of the most exciting parts about 628s.
Okay, so perhaps maybe, you know, if you explain to us sort of what are the next steps in terms of moving this forward from a clinical development standpoint? And then maybe I'll couple it with a second question, which is, you know, this is gonna be in focal epilepsy, but how do you think about tapping into, you know, other types of epilepsies such as generalized epilepsy?
Yeah, it's maybe I'll start with that, right? When you're looking to the utility of a drug like 628 and the models we tested and the response you've seen on the PPR, it fits quite well not only for focal onset seizures, but for generalized as well.
So the entire plan has to include both indications, that something we're actively discussing, the timing and when to run, but it's a when, not an if, going to both indications. Now looking to the next steps here for 628. So the first is the next efficacy trial, which we intend to start second half of the year around mid-year, get patients on the second half, which would be our adult classical study in focal where you see patients who failed previous treatments. So normally two anti-seizure medications that fail.
They're gonna come with an accountable number of seizures, about one per week, or so between one and two. That's normally what we see. And there is a choice in terms of the duration, right? Some people do eight weeks, some people do 12 weeks. We are believers on the 12 weeks paradigm for a study like this. Is more time to look into a number of things, that seems to be what's in favor by the FDA as well, in terms of the duration. So that's where we're going.
One of the things we define as a company that is quite important is to understand where these patients are coming from. And we tend to invest a fair bit on understanding the patients before starting the trial so we can de-risk the patient profile, increase the quality of the patients, and more patients available at day one.
So that's what we are doing right now, with the aim to start the study later this year and to read it out before the end of next year for the first study. In parallel to that, continue to develop the second study and have that started next year. So that would be, arguably the registration package for the U.S. And what we are looking is how can we fit a generalized, study on this mix as well? So from the get-go, we can have the broadest possible, label here.
So, a lot going on there. Obviously we couldn't do that until a few weeks ago. But now with, with the capital, with the ability to actually, invest on these studies and accelerate the development, which is quite key. That's what we are, we're doing.
Okay. That's exciting. Let me just take a quick pause and remind our listeners that if you have any questions that you'd like me to ask Marcio, feel free to send it over through the dashboard and I'll incorporate them in the questions that I'm asking him. So with that, let's switch topics to ulixa in essential tremor. You're gonna be reading out results from that study in the second half of this year. And maybe if you could sort of, you know, in the Essential3 study, you have two sort of simultaneous phase III studies, including a kind of a 12-week parallel design study and a 12-week randomized withdrawal study design. Can you elaborate on the rationale for designing these trials this way?
Yeah, absolutely. Absolutely. So our candidate for essential tremor, right, ulixacaltamide hydrochloride is actually quite fascinating because if you're looking to the T-type calcium channel, it's such a fundamental mechanism for control of tremor. So the ability to properly modulate that with a drug that can be given to these patients can keep the therapeutic levels and is safe, which is key for this population, has been kind of one of the key questions for the field for many years.
And we believe that with ulixa, we can get there. We come from what I would call quite successful phase III study, right, where we're looking into the dose and we could set like a dose for the phase III. We're looking to the proper endpoints. We explored the number of endpoints.
We settled and we got a very good agreement with the agency on using the ADL as a primary endpoint here, which perfectly linked to how patients function and obviously the design of these studies. When you're looking to creating substantial evidence of effectiveness with two studies, the question becomes like, what is the simplest number one, the most de-risked and the one that's gonna inform the most they use?
And when you're looking to all those criteria, and in the discussions we have with the FDA, it became pretty clear that we could do two studies that are identical. And I would argue that's what most people do. Or we could look into a combination of designs, to your question, which would answer more questions, but in general would de-risk the program more as well.
De-risking from a number of patients required, de-risking from level of information from the overall expected response. So we settled with parallel group study. It will call Study 1 in the Essential3 program, which are 200 patients per arm, 1:1 randomized to 12 weeks. We are expecting to see; we're looking for change from baseline there, so how patients progress in the study and the difference seen. What is super interesting about essential tremor: placebo tends to minimize, to be minimized or to be tapped, around 2 weeks or so of treatment.
That's what we've seen before. That's what we're seeing before. So that is a pretty interesting chance here when you're looking to extending as placebo stay stable, and we've seen that before in our open-label extension studies, the more ability for the drug to show the results.
So we're already pretty well powered with the trial, but we're even more excited about those features. Then we had to answer the question to ourselves. It's what is the best second study to show efficacy? And we ran two pilot studies before, showing that when you remove the drug, the loss of effect that was gained with the drug is pretty dramatic and quite quick.
So that lends itself fairly nicely for a randomized withdrawal study. And we discussed that with the agency, and they were in agreement with our package in general. But for us, having a 200-patient randomized withdrawal study, it's quite an upside, right? Since that study is really well powered, it allows us to look into the treatments as patients will be in the real world, meaning they're gonna be only randomized after they reach a level of response.
And then you remove the drug for the responders, half of them, since it's a 1:1 after that point. And you expect the patients that you remove the drug to fail and you measure at the end. So allows for a very comprehensive package, arguably for a very comprehensive label. We are marching through the study right now. We expect to read it out by the end of the year for an NDA submission next year. So accelerating quite nicely in terms of registration for the very first in modern developments essential tremor drug to get to that stage. So incredibly exciting.
That's great. I think sort of the concept and the rationale around the randomized withdrawal study makes complete sense. Can you sort of maybe dig in, can we dig in a little bit more on the parallel design study and try to sort of think, if you could sort of help us understand what's the powering of the study and what can we look at from the prior study, to give us a sense of, you know, how the study is best set up to succeed?
Yeah, absolutely. So if you're looking to Essential1, then when you apply the same metrics, right? So the same general hypothesis here. In Essential1, when you look into the entire population, right? So this is not like subgroup analysis or anything like that. It's really the entire population looking to the same endpoints as it was measured exactly the same way, was calculated exactly the same way.
We reach nominal significance from statistical standpoints with ulixa versus placebo at a trial that was 3.5 times smaller than what Study 1 is. Obviously it's not only size, but size matters a lot when you're talking about power, right? So we, we're looking into the uncertainty of that measure. So yes, we were nominally significant on the endpoint. The endpoint did not change. The population did not change.
But what are the uncertainties as we scale the study, right, that come with it? And we wanted to make sure that this study is really well powered for Essential1. We had no previous study or prior to power Essential1. So we are learning with that study. Now we do. And when you account for all of that, the Essential3 is over 90% powered to detect a difference on these endpoints. That is the ADL, 11 items as requested by the FDA.
When you look into the change from baseline to either week eight or week 12 as the primary endpoint is, we feel really confident because we discounted the things that could be, and I mean discounted like in a positive way, the things that could be variability measures here, like we increased the standard deviation from what you're seeing before, expecting that in a larger study, we're gonna have a little bit more variability. So far that's not what you're seeing. But we're like, certain that that was the proper thing to do. So if anything, would even have more certainty, making sure that the patients are assessed equally throughout the study.
So to reduce the more intrinsic variability for the patients, adding an additional check for stability before randomization, and making sure that the certification factors that we learned in Essential1 were applied in Essential3, meaning propranolol used at baseline, which is about 20% of the patients or so. Family history, which is about half of the patients have family history of essential tremor, and the presence of intention tremor on top of essential tremor, which is about 20%, as well as we expect. So when you control those variables, we tend to have an even better show for the drug, I would say a better result for the drug when compared to placebo. And that's what we're expecting to see later this year.
The powering of the study that you talked about, is that including or excluding the intention tremor patients?
Including patients with intention tremors, on top of that as well. So the key for the previous study where we're seeing a variability there was because patients with intention tremor were not balanced between placebo and drug. So now that we are balancing that prospectively, we believe that it's even less of an impact in general, if at all, for Essential3.
And just, you know, as a reminder for our listeners, obviously, based on Essential1, because of this slight imbalance between the two arms, it wasn't clear whether the intention tremor, what we were seeing, right? We couldn't really tell what we were seeing with the intention tremor subpopulation. And the intent of designing Essential3 by including intention tremor is to, you know, study the widest possible patient population.
But if you find that there's something different about intention tremor patients, right? And for whatever reason they don't respond to this particular mechanism, then how does that impact your abilities to sort of provide that, you know, positive study? How's that built into the trial design and statistics?
So we did see numerical improvements on top of like whatever preexisting condition these patients had, right? Including intention tremor on Essential1. So that gives us more certainty on the mechanism that is not necessarily that we don't have an action, but that action is influenced by an underlying condition for the patients. So balancing became more important. We did have conversations with the FDA about the intention tremor. I think there's a very good appreciation by the agency that it might have some influence on the ability to show results. But because of the calculations and all the observations we had did not show that impact on a trial that is designed as Essential3. Yeah, for a smaller study that was not balanced, it's a different impact.
But even if you look with all those variables and you assess the same endpoints, it's still nominally significant on Essential1. So it did not impact to that level by controlling everything on this study. We should not have any issues with intention tremor now. Fast forward, and there is an issue, right? Which biology always surprises us. I'm certain, I'm very positive that would be a lot of willingness to look into that. And it's one of the ways we're analyzing, it's one of the discussions we had, as well, why we don't believe that was gonna be necessary. We're pretty certain that would be well accepted if that's the case.
Okay. Let's switch gears, and I next want to talk about PRAX-562. Perhaps, maybe you can just sort of summarize for us what is it that you're looking to learn from the EMBOLD study and, maybe remind us when that data is expected and what would be success from that study?
Yeah. So we are expecting the results from EMBOLD by mid-year. We're very excited about that as well. There is a lot of learnings there. It is a proof of concept study, right? So really, looking into learning as much as possible from this patient population. When you go in, your classical DE population is patients that are very young, right? They're by definition, if they have a developmental encephalopathy, they have to be young. And they have intractable seizures. So they're seizing a lot, a lot more than we see on the other conditions that we're seeing. And they have not responded to pretty much any medication. So when you're looking into 562 in those genetic models, right?
Now moving from the more generic models like the MES and the 6-Hz, but going to the genetic models, we're looking to the ability to actually completely eliminate the seizure on those models. Of course, as we talked before about SCN8As, the translation of that is likely not to be complete elimination, but gave us a hint on the power of this mechanism for those patients we are looking for, the use of, like 562 on top of other treatments.
We cannot remove the background therapy for these patients, right? Otherwise they would go into status and really bad things could happen to them. So looking for an additional on top of the maximum they could ever have seizure reduction, other improvements, quality general quality of life for those patients.
So we designed a 16-week study, where patients are both treated with placebo or drug and switch from placebo to drug. So we can have a general idea of what happened. That design, the exact weeks is blinded to the investigator, to the patient, to everyone. So allow us for the best possible comparison without an influence here from will, of those, in general.
And we expect to, to having enough information at the end of the study to power our pivotal study for 2A, SCN2A and 8A , leading to a single registration study and an approval on those and NDA and approval on those indications in short order. One thing that's interesting for this program as well, we have pediatric designation for each one of those indications. And that is only the beginning, right?
The actual mechanism and the preclinical models that we have would allow for an expansion to many other DEs that are quite exclusive from other DE treatments in development as well. So, coming up soon and stay tuned on that, and excited to move that one as well for late stage.
Okay. We have five minutes remaining. This is not gonna do full justice to the Solidus ASO platform. But, if you could maybe, you know, put the platform in context, as you talk about PRAX-222, the EMBRAVE Part 1 data, that was presented last year, showed very nice seizure reduction and then, you know, reduction in seizure-free days. So maybe if you could sort of talk about your discussions, ongoing discussions with the FDA there and what the next steps there could be. So a lot packed into that question.
No, no, that's great. I'll try to do justice for that in the next four minutes. The platform itself where you're looking is for the best possible ASOs for a given disease, what it makes the most sense to treat in a given way. And what I mean by that is sometimes it's a retained intron, sometimes it's a reduction of expression, sometimes it's a general increase on the expression that the biology and like population genetics, genetics in general are leading to.
Looking to the preclinical work for 2A , for SCN2A gain-of-function and PRAX-222, it is by far the best possible outcome you could have on those animals. And then we cautiously move into patients, right? Cautiously because those patients are incredibly severe. And you're not expecting to see much.
Despite all the positivities that were coming from preclinical, we were blown away by the results we're seeing. Not only the ones we actually talked publicly on the reduction in seizures, increase on seizure-free days, which are quite important, but everything else that comes with it. All the quality of life measures, development measures that were reported, we're receiving the reports from the parents, directly from the sites, and being measured and seen.
So globally affecting the disease. That's what we are looking to do for all drugs at Praxis, not only the ASO platform, but for all the other ones to come. We are moving forward later this year. We're gonna have our registration program enrolling patients for 222, which should be a one study for global registration.
We're gaining alignment with all important regulatory bodies in the world, and that should be clear, and then we'll talk about that publicly pretty soon. What is next for the platform is on the next 12 months, and as early as the next quarter or so, we should sequentially declare the next three candidates. The next one is PCDH19. Another one without any treatments, very large population as rare disease goal.
And we have quite exciting, actually, data we're gonna be talking pretty soon as well on what is possible with that, followed by SYNGAP1 and SCN2A loss of function; there's the leading cause of autism from a genetic standpoint. So very exciting on that as well. We're gonna continue to be diligent on how we deploy capital, how we deploy human capital, financial capital for those programs.
But I think as we've shown, on the last several years, we're very good on moving things along and being very serious about the science behind it.
That was actually a great summary of the Solidus platform in five minutes or four minutes. Well, with that, I want to thank Marcio for taking the time to join us today and, for all our listeners for, joining. Thank you so much and everybody have a great rest of the day.