Welcome, everybody, to the last day of the 45th Annual TD Cowen Healthcare Conference here in Boston. I'm covering analyst Ritu Baral, and thanks for joining us for the fireside chat with Praxis Precision Medicines. Here with me on the stage is CEO Marcio Souza. We have other members of management in the audience, but I guess let's start with your some epilepsy programs. Relutrigine, can you review how your sodium channel modulators are differentiated from those used in the SOC, I'm sorry, in standard care, and the competitors in development? Yeah, absolutely. I appreciate it. I appreciate I appreciate starting there as well. Coming with the—and thanks again, Ritu, to have started like that. I appreciate being here once again at Cowen. When you think about the way patients with GE are treating right now, it's really blunt force, right? We're trying to give these patients, for the most part, there are some approved treatments for a limited number of them. It's really like old, nonspecific, unfortunately fairly toxic treatments that are often given to these kids. I say kids because the majority of them are kids. There are, of course, a number of adults and adolescents and so on with the condition as well. One would like to, as we did, and others in this space, to be as precise as possible while tackling the issue here, independently of the mechanism, right? I think sometimes we get a little fixated on the mechanism, but at the end of the day, we should be asking what's happening when there's a lot of seizure activity on these kids, and how can we stop it? And how can we stop in a way that the kids are still there, right? I mean still there because you can completely sedate anyone and stop all seizure activities. That is actually not too complex to be done from that perspective. That's what they do with status, right? They give you propofol and put you into a coma and seizures, right? Point of view, exactly. When you look into relutrigine's profile, what's very unique and unexpected, and to some extent was a combination of very good screening and design, but a lot of luck as well, it's incredibly potent and selective to the ultimate place by which epileptogenic activity goes through, right? You cannot have seizures without sodium channels. Now, people talk a little bit like a lot about what happens and how can we stop and different mechanisms that fundamentally it has to go through the sodium channel activity to have a seizure, right? Modulating that and blocking that in some cases is fundamental. Like if you can do that safely, you're going to reduce or eliminate, as you've seen in many cases. When we were testing the drug initially, we were focusing on one of the most important of those isoforms. If we can completely block or partially block that, we should reduce across the boards in GEs, and that's NAV1.6, that when mutated, called for SCN8A. What's curious and, of course, quite interesting and exciting for us is the more we tested that across different models of disease, excuse me, the more we've seen the activity was maintained. Asking back the question, why? Why is that so different than other mechanisms? Two of the answers I gave already, it's incredibly potent, it's very selective, but avoids almost altogether a toxic current, right? Those channels are very dynamic in the brain, but that is an identified current that when you block one of the three major currents on this channel, you have toxicity. I will call that the tonic block. You avoid that pretty much altogether. Toxicity means sedation. Sedation. Disease. Dizziness, lack of life if you block completely, because at the end of the day, if you completely stop action potential, there would be no high-level activities to be done. No seizures either, but at that point in time, you're not having any other activities from these kids. We did a very robust preclinical package, early clinical, safety pharmacology. Every step of the way, and you're going to hear this again when we discuss vormatrigine, we're seeing very, very good activity. The highest bar, the worst case you can test, if that's the hypothesis, is by getting patients who seize the most. That is SCN2A and SCN8A kids, all right? They have uncontrollable seizures. They use sodium channel blockers as standard of care. The only way to actually show efficacy is if you are significantly better than anything that is out there. That is incredibly important because there's always a lot of questions about, well, you're stacking the mechanism, right? What is happening when you're stacking the mechanism? We did that experiment. When you look into the results of the board, that was our phase II study on SCN2A and SCN8A. We actually did the riskiest experiments. We got very fragile kids seizing significantly more than other GEs that were studied before. The bar is incredibly high using a mechanism that is already partially blocking one of the targets. We had not only similar to no safety events other than what was happening on the standard of care already, but phenomenal efficacy, like 46% seizure reduction on the four months that these patients were treated. Quite surprisingly, even to us, to be perfectly honest, 30% of these patients, 33% to be precise, became seizure-free. That was phenomenal, and I think we should all be proud and humble. This is something that actually a couple of the experts brought up in response to my question yesterday at the epilepsy panel. One of them said that it's important there's a dose-response curve, and you need to get up that curve, but you can't get up that curve with standard of care because dose-limiting side effects. The question is, is there another option to help you get further up that curve? The other KOL actually weighed in and said different drugs bind to the sodium channel different ways, and therefore can affect different types of inhibition with different characteristics. Can you talk about that aspect of it, the binding? Oh, yeah, absolutely. Yeah, so this is a different pocket, right? We didn't talk much about exactly if you read our paper on the mechanism of 5, 6, 2, now relutrigine, you're going to be able to guess where this pocket is, is in a different location than the others that were explored before. Therefore, the difference on the mechanism here. We are way beyond the limits where the key opinion leaders were telling you earlier this week that you cannot dose. That is quite important, right? Because that's where efficacy is. That's where additional efficacy is. That's where resolution of seizure is. You cannot go from 100, 150 seizures in a week to zero by chance. I want to stress that, right? That is what you've seen in both. I want to finish this part with another data point, right? We kept treating these kids. If you look into our latest AES poster, which is on our website, in the long run, it's 77%. The longer they stay on this drug, it's not that they are losing effect, which is very common on standard of care. You can have these honeymoon periods, like the first week, the second week, and then all of a sudden, hell breaks loose, and they start seizing again, and it's not what's happening here. It's not only a very fast, very robust effect, but that deepens over time. Now, can we go any better than 77%-80%? We really can't, right? Like on data sets like this. That should put the nail in the coffin there in terms of how important it is to this different mechanism, despite its modulating targets that is known. I would argue modulating something we know is a positive, right? We know how the channel works, and we know how to modulate it, the ability to control the seizures, but also the potential liabilities that would come on the other side. Great. Your ongoing phase II study in SCN2A and 8A, how is enrollment going in the registration cohort two of that study, and when should we expect top-line data? Yeah, so that study is 80 patients, right? It's intended to be a registrational, of course, expecting that to be positive. Our current guidance is for next year, first half of next year for readouts. Since you asked about enrollments and how it's going, always a little bit gun shy to talk about enrollments once we are like about a year from a potential readout, right? It's actually quite nice, and I would say that choice of words is not random to see where we are right now. One of the ways I've been framing this in some of the conversations this week, maybe one way for all of you to interpret the first half is like a worst case scenario, and we can put some flags towards being potentially faster than that. Things like site activation, IRBs, all that's going well. Everything is open. Patients are enrolled on both sides of the Atlantic. This is a European-American study, so we only have sites here in the U.S. and in Europe. We do not believe we are going to need sites anywhere else because at current pace, we are probably going to be having a nice surprise in terms of the enrollments finalizing. How should we be thinking of what potential top-line data will look like? I mean, you saw a placebo-adjusted 46% in motor seizures in cohort one. Is that setting the bogey for cohort two? I think that is when you look into like 46 for this patient population, we could never have expected. I believe last time we talked about expectations for this study, we're like, that is absolutely nothing that works. If you can give that additional 10%, 20%, 30%, that would be phenomenal. We are very happy with over 40%. Of course, if we had power for over 40%, this study would be smaller. We are giving ourselves some room for the effects to fluctuate, as oftentimes does when we increase the size of the trial. An effect that is in the neighborhood of what we've seen before would be incredibly meaningful. Do you believe that that minimum is still like the 30%? Absolutely. Thirty percent ranging up to sort of that 45%? I'd say 30-50. I think that's where we should be thinking. You're also planning to start a registrational all DEE study by mid-2025. How are those FDA discussions going, and what are the key considerations as you talk to your advisors for the agency? We had completed the interactions with the FDA in relation to Emeralds, that is our GE study, broader GE study. We're refining the protocol based on that right now. I think some of the key areas that we are discussing is now we're talking about adults as well, right? Not only children for some of those GEs. Should we put like a maximum on adults versus kids or not? Should we stratify it? Things like that. Fine details as we're looking to this and refining a little bit the inclusion criteria. I think one of the biggest questions on the table has been the inclusion of SCN1A patients on this cohort. The majority, not all, about 85%-90% of the patients with SCN1A mutations have mutations on the intraneuron or that affect the channel there as a loss of function, right? Historically, uninformedly, may I say, most people associated that with contraindication to sodium channel modulators. Because if it was a non-functional channel, you couldn't modulate it. Right. It's misguided in a sense, and we probably don't have enough time to say why. We generated very compelling data, and we had that discussion with the FDA. We're going to be publishing that data very soon, showing that when you actually properly modulate on the excitatory neurons, that's what we care here, right? That's what seizures go through. We actually have really incredible results preclinically at this point, and we expect to translate clinically on that population. The question should still be out on whether or not that should be a subgroup or not on the overall trial. Can you review the market opportunity for all DEEs versus just the 2A8A? Yeah, yeah. 2A and 8A, about 5,000 patients in the U.S., I would say a classic, nicely placed opportunity for a rare disease, right? I think we all in this room can name 10, 15 rare diseases with fewer patients than that that are making significant amounts of returns right now. That was the maximum. Standalone, very important, very clear opportunity would justify multiples of the current market cap with that alone. When you look into the GEs that would be treatable, there are certain anatomical defects and things like that that I think we should all be cautious on thinking that anything pharmacologically could resolve it. When you look into the treatable one, we're talking about around 200,000 patients in the U.S. All DEEs together. Yeah. Some estimates are a little bit higher than that that are out there. We did a very comprehensive look. We believe the number is around 200,000. Within that, the biggest chunks are Dravet, LGS? The biggest individual chunks are LGS and SCN1A, loss of function. Oh, 1A, okay. Dravet. Those are 10,000 on that. They are bigger individually, but they are very small collectively on this. The opportunity here is multi-billion dollars, if you think. When will we have the final indication on that strategy in that trial? Months from now, we should start the study. We should expect that no later than. No later than mid-year. Yeah. You'll let us know upon study start or near study start? Upon study start, yeah. Okay. Okay. Great. I want to move on to Vormatrigine 628. This is your sodium channel blocker, not modulator. This is in the phase II-III Energy study for focal onset seizures. You are also looking at generalized epilepsies. Can you review for us the three different studies and what they are designed to show? You have Radiant, you have Power One, and you have Power Two. Right. I'll put the logic there. I'm going to add a fourth study. There's no pharmacological, but it's quite important for the overall strategy here. I'll do that at the end. Radiant has about two to three key objectives. It's an open label, eight weeks, 50 patients, which we're going to be reporting data mid-year this year, in a few months, I would say. What we're expecting to see from that study is obviously very clear seizure reduction activity on the majority focal, because just by the amount of patients with uncontrollable focal onset seizures, that's going to be the majority of the study. A glimpse of what would happen with primary generalized as well on these patients. I would say very clear guiding on focal indicative or we might have seen in generalized there. At the 30 milligrams, that's where we start here. We are way over an expectation of benefit for drugs in general in this class. You should expect what I'm going to call very clear results from this from a seizure reduction perspective. The 30 milligrams per day, which comes obviously with a lot of the benefits, and I'm sure some of the QLs said that, is we think about seizure control. People talk about seizure reduction. Yeah, but if they have to take three times a day, you can cut that opportunity by five. If they need to take that with foods, you can take that opportunity by two. Those are people like us. Maybe someone in this room has focal onset seizures. It is incredibly difficult. The more complicated you make those drugs. Vormatrigine has none of those complications, right? Once a day, no food effect, very simple. Therapeutic levels achieve that the first hour to the second hour. First day, you are already at therapeutic levels. That is mid-year. Power One is recruiting parallel. We're going to have the readouts towards the end of the year. That is your classical, I would call, focal onset seizures study. One-to-one randomized. Patients are going to stay for 12 weeks. We're seeing for the rate of change during the periods. We're expecting, of course, very clear results there as well. That would be serving as potentially one of the pivotal studies. You mentioned the bar in focal, but in Radiant, you're also doing generalized. How do you think of the efficacy bar in those two populations? Is it a different bar? Yeah, I think it is. In focal, it's pretty clear. There's a set of competitors out there, right? While you could have a very significant opportunity if you get lower than competitors, I think no one here, including ourselves, would be too happy if that's what happens. I think we can benchmark and say at or better than what we think is the better possible opportunity out there is what we should be looking at for focal. Generalized is almost the opposite. What does work for refractory generalized patients? Nothing, right? The generalized is a little bit more binary as well. You're probably going to tend to see larger numbers because they have lower seizures. They can't really reduce by a few, right, on that. I think we should expect similar numerical numbers, but the impact of reducing a generalized seizure is massive, right? Those are full-on disabling for this patient. You could have a lower percentage reduction with a bigger clinical impact. Yeah. Maybe what I'm trying to say is we might end up having very similar percentage reduction, but the clinical impact of 40% or 50% reduction in generalized is massive. Got it. I don't believe I covered Power Two. No, you didn't. Yeah. Power Two, we are starting in the next quarter or so. That study is ready. Could start today. That's the one with the 10 milligram arm, right? No longer. Maybe that's the reason you gave already, the reason why we are not starting. I think our go-to hypothesis with the agency was they would like us to explore a lower dose. It does not seem to be the case. I think they are at this point in time comfortable with us actually exploring only higher therapeutic dose, immune to my ear, because that's where we know we're going to see more results and more positive results with this drug. The question here, we are guiding on two things. We're gating a little bit in two elements. One, as Radiant mature, there are scenarios in Radiant where the efficacy is what we expect and would continue Power Two as we design. There are scenarios on Radiant that the efficacy is so much better than we expected, then we need to rethink Power Two as well, right? Get that faster to registration, for example. Based on the conversations we had so far with the agents, I think there's a lot of openness for that. Because everything is converging in the next few months, there's no truly delay on actually getting through this extra exercise, number one. Two, we're not in the business of competing with ourselves. It's going really, really well for Power One right now. Once we finish this, it's going to be, obviously, we do very high-quality studies in general, but it's going to be very high-quality, but very fast enrollments as well. We don't want to break that by being too greedy and enrolling two competitive studies at the same time. There's only so many sites. There's only so many pools of patients. We're staggering a little bit. Go slow to go fast with Power Two. That should be still a readout early next year as well if we start this study in the next few months. Got it. I do want to round back to your essential tremor program. Last week, you disclosed the preplanned interim hit futility on study one. You noted some underlying assumptions on the statistical, what they noted, underlying assumptions of the statistical model may have influenced the outcome and suggested exploration of other statistical models. Can you walk us through a little bit of what model was utilized and what other options that the DSMB may have been suggesting? I believe you mentioned they did not suggest anything in particular, but what are the options you are considering? Yeah. Two points there on the five minutes we have just so. One, I'm very happy that we discussed the epilepsy first year, right? The epilepsy portfolio has more than clear potential to be multiples of what we have right now from a market cap perspective and should be the main thesis on the company moving forward. When you start with, I'll get it out there, on everyone's models should be zero, ET. I think we should take at face value that there is no way to figure this out. Having said that, we can't simply ignore what we were told and what we are seeing. When you look into both of those things and the IDMC, and I applaud them for being proper in terms of the communication with us, because that could have created more problems than odds. They gave us the formality, as you mentioned. They also said, and I believe it's in the best interest of patients, that you should continue exploring this. If you believe that that's the right thing to do, you should continue exploring that. They were with the knowledge that you all have as well, that this study is basically done, right? We have patients being enrolling throughout. We talked about the pace of enrollment before. We are really at the mark that we expected to be right now. You're not really seeing any idiosyncratic or otherwise safety events that would preclude us from continuing. The model that was used is what I'm going to call your classical mixed model repeated measures with regular assumptions. What are the things now? Those are the facts. If I go two seconds for speculation lens here, which you should take with an incredible grain of salt, all right? We know what was happening at week eight. The reason why we know what was happening at week eight is because study two runs in parallel to study one, but it is blinded to everyone but a small group at Praxis. That study never needed to be blinded. We decided to blind so we could monitor them more perfectly. That gives us a window on what effect would be. When you compare the effect with the blinded and then truly blinded effect on study one, we can understand what the committee was saying. Like there is something going on here, and this something is likely favoring drug. We have got to try to figure it out. Our going hypothesis right now, it's something happened within the time series, all right? Within all the time points, not necessarily with the raw effects of the drug. Therefore, that could be we're not certain that is. We should assume that isn't. There could be other ways mathematically, therefore models to explore this data that would be more appropriate for this disease and this trial. That's what we are doing right now internally and externally. Fast forward to Q3, study two reads out. I think we all discussed this. Ad nauseam before, study two always had a higher probability of success than study one. Study two reads out positive. Are you going to change the SAP of the withdrawal study two? Not in study two. No. In study one, very, very likely. Yeah. We can combine studies at day 56, which was always an analysis we're planning to do. If we find a better way to analyze it, we might be in a situation that is an upside to all of us. That is the situation we're in. I think that, of course, we're all going to be incredibly happy, but we shouldn't assume that that's where we're going to be. Great. We still have one epilepsy program and a minute left. This is 222. This is 222. Your ASO. And this is for 2A SCN2A. You recently disclosed you've got global regulatory alignment. Cohort one is going to look at 1 milligram against sham. What is the sham procedure that's going to be conducted? And how is the, I guess, why was this selected? Yeah. We had a very robust conversation with the FDA. I was very happy I was there. It was a water situation. It was not a lack of water situation. Internal jokes. You can ask the current team about it. Very, very positive in terms of I encourage all of you to look into the design of this study and the design of other ASOs being developed for these studies. There is a very clear difference here, right? A priori, you would expect those studies are disease modifying. They should be measuring similar things. The understanding of the effects would dictate things like duration of the study, choice of endpoints, safety measures. What we got from this study is a six-month study, which means there is an understanding. There's a very robust effect that can be derived in as short as six months, which is a huge win. The second is the sham here, and I will not going to describe because I don't want any potential patients to want to hear about this and like, oh, so I was on sham. It's very simple. It's done behind doors in the OR. Patients or even the blinded team at the site wouldn't know what it is. It does not increase the risk of patients. Why is that important? Because other sham procedures being used and currently used and designed do increase the risk for patients. Injecting a needle in the spine is common. It's infection risk, I mean. There is infection risk. There are general risks of post-lumbar syndrome. CSF pressure. There are a few things. It should give you a lot of confidence as well about what we haven't seen. What we haven't seen is CSF protein elevation and other markers that would necessitate doing things on that control population to try to see if it's disease or not. We are happy that this study is meant to be registrational. That's how it was discussed. We are going to be starting enrolling pretty soon. We intend to have actually quite fast enrollments. When are data catalysts from that? Next year is when we expect to have that all set and done and ready to submit. Great. We are at time. Marcio, thank you so much. This was very helpful. Thank you, everyone, for joining us. Thanks for having us.