Praxis Precision Medicines, Inc. (PRAX) Earnings Call Transcript & Summary

Good morning, and welcome, everyone. Thanks for being here in New York City with us and in the webcast. It's a really -- real honor and a privilege to have all of you here for our second Analyst Day, our Epilepsy Day today. As you can guess, we're going to be making some forward-looking statements. I'm not going to be reading out the text here, but I do encourage every and each one of you to look into our SEC filings, especially our Form 10-K filed with EDGAR. So why are we here today, right? We're going to be talking mostly about the epilepsy franchise, but it would be disingenuous not to talk about the other 2 franchise, like each one of them have significant value creation for patients, first and foremost, it should starts and ends with who we are serving as a company, as a community, as investors. The second is there's a number of real big events for Praxis coming up. And that was the disingenuous part, I'm just not addressing that upfront. I'm incredibly excited and all the team at Praxis and everyone that are in the community, and we know health care professionals are as well with 944, [ 221 ] Part B, the randomized withdrawal phase for our essential tremor study reading out, I guess, now I can say in a few weeks, since May is right over the weekends, starting at least. We're not going to release data over the weekend, so we're clear. And right after that, our psychiatry franchise, having the -- one of the biggest readouts, I would say, in the year in biotech, in our view at least, there is Aria. We're so incredibly excited there because as I'm sure it touched other life of many of you, it definitely did on my family, the psychiatry in general, those conditions and depression, most specifically, has such a huge unmet need. So we're really looking forward to that. Now if you were to talk about unmet needs, we can go across the board here, know the disease we're going to be discussing today. And there is not even a way to classify, right? A patient, a family, a system of support the physicians and nurses that are around those patients, there's always the unmet need. I think it's, once again, a little bit disingenuous to talk about higher or lower. So we're going to be talking about the needs of individuals here and their support networks and the people that struggle, to be perfectly honest, on helping them every single day. So that's the general direction, what you're going to hear. But I wanted to bring back to, how do you connect all of this, right? And hopefully, today is going to be more clear than it ever been, for Praxis at least. And what we see is this connection between the different nodes, transmitters, networks, [ seller ] process, disease processes on those 3 franchise has been fundamental to how we thought about the company but it's more fundamental about how we develop, how we reduce to practice everything we are doing every single day moving forward. Now the underpinning of -- or very first spots when you're looking for a program, for a drug, for how to address this, what is the most unequivocal item that we can look into and say, that thing, whatever the thing is, does generate a pathological process that I understand with a one-to-one relationship or a one-to-multiple relationship? And in our view, it's anchoring the entire pipeline in genetics. The aha moment here is that when you study genetics of epilepsy -- for many of these studies being done before, Dr. Petrou is going to be going through some of them and the inspirations there -- originally, we're thinking that's going to be it, right? We're going to look into some of these genes and they're going to be fairly exclusive, but they are not as well, no. And that allow us to really ask, how can we modulate the processes? How can you modulate the networks? How can you really help more patients outside of that? We chose very deliberately to go into movement disorders and into psychiatry on top of epilepsy. There could be many other here. And there's a number of prioritization choice that I want to make before we continue to move here. So the first one, I would say we all know biotech is a beautiful market right now, right? Everyone is incredibly happy with the performance of the sector. But what comes with it, the positive that comes with it is discipline and focus are even more important and focus and discipline on the science, as I am absolutely convinced that at the end of the day, you're going to see how much we have; financial discipline in terms of who the kind of individuals that we want in the companies, how we distribute work inside and outside, but really the day-to-day expense management and so on. So with part of that, one of the things we did decide to do is to discontinue our program in [indiscernible] for 562. There was nothing wrong there on the safety. There's nothing wrong there with how the program was going. It did not meet 2 of the criteria for us. And I wanted to say that right before I move to the next concept, that has been true to how we filter these drugs. So first and foremost, as I just said, every single program is anchored in genetics. So it starts and ends right there, right? If you do not understand the genetic process and it's not to say that if others do a different way, it's wrong, there is no right or wrong. That is just -- the right is the drug that gets to the patient. How we get there, there are many different ways to do so, but that is a filter for us. That is the principle, right? But from there, we chose and we are committed to 2 principles. One is translational tools that are real. And what that means is, we expect something to change. We measure that thing, let's say, electrophysiology, say, quantum state of EEG, a fluid biomarker, whatever it is, and it does change and we can replicate that in humans and that the drug is in the brain. I think if you're looking to the most recent issues that happened in CNS, since CNS gets this bad rep for higher risk, they're mostly because that step was not properly fulfilled. If you can jump straight into a later stage without understanding its proper concentrations of the drugs in the brain and if it's doing what is expected, I would say, of course, something is going to happen that is not desirable. We don't ever go over that step. But then the one I was referring to is the one to the other side of the slide here. If we cannot generate proof of concepts quickly and effectively, our filter for progress, let's say, stop on it. And that's what happened with [indiscernible]. So that has nothing to do with anything else. What it gives us, on the other hand, is capital to make sure we continue to run the company, #1 from my perspective, from [indiscernible] perspective, I'm sure you would appreciate, but reinvest as well as needed on the high-potential programs that we have. And to finish with the circle here, what I would say is we talked about patients again, it's an incredible cliche in the industry, right? I don't believe there is any single company out there that say they're not patient-centric. And I'm not going to go into this, things on the wall -- or things on the Wall Street Journal saying how patient-centric we are. We are. And you can see this every single day on the questions we ask, on the unmet needs we met, on the ability to actually directly ask the tough questions, should we develop a drug here? There is really a need. Can we be best or first or first in class and drive the benefits to the patients? And you're going to see throughout the day that that's exactly what we are doing. Now none of those principles matter if you don't get at the end and generate sufficient cash revenue, free cash flow, hopefully 1 day, to reinvest and to continue to help because it's to do this again and again that it says in our mission, right? So if you go back to today and to the markets in epilepsy, there are 2 major areas we're focusing on. The first one is common epilepsies, and you're going to hear this from Bernard, you're going to hear that from our experts who are going to be speaking today and you're going to hear a little bit from Steve as well how to address the significant number of patients in the market right now that are just struggling with residual seizures or, equally worse or equally bad, with many other things that happen in their life because we have the single focus oftentimes on reduction of the seizures or seizure freedom, but that is a surrogate to everything else that happens on their life. But from a purely size of the market perspective, we're talking about a very significant, multibillion-dollar opportunity for us and very concentrated as well. And one of the things that the epilepsy community did really well was, one, form the medical schools, the programs, really good neurologists that are dedicated to these patients, and that helps a lot when we're getting in a given market. But the second is really understanding why, what's happening and the choices they have as we're going to hear. But that's one side of the equation. The other side of the equation is the rare or more defined epilepsies. And while there are many, as Steve is going to talk a little bit about that, we always go through 2 major filters, just to say that again, 1 is business. If there is no way to actually get this drug to the market and to reinvest that in another drug, it's probably a fool's errand to go into that direction. And it's not the markets to be, it's the market today, meaning the number of patients we understand right now are in needs. Every single program we start, we ask, are those patients actually seizing? Having other comorbidities? And you're going to hear about some really cool initiatives we're having to get that upfront. But even if you put all of that aside and you get exclusively the indications we're going to be talking today, we're talking about a much bigger markets from a purely revenue perspective than anyone else has right now in development for epilepsies. Our press release that was issued about 10 minutes ago says in the headline that we have the largest pipeline in epilepsies in the world. And that's very true in terms of not only addressing the patients, on the collection of it, how we go about the science modalities quite importantly as well, in general, from a revenue potential perspective. How does it fit with the overall pipeline, right? Once again, as you're going to hear throughout the day here, in the next few hours, epilepsy serves us 2 purposes: one, stand-alone therapeutic area. And hopefully, I'm not going to have to reconvince anyone that is a very potentially profitable one. But the second is a gateway or a springboard to many other disease in the brain. Some of them, we might develop ourselves, others, there might be better companies to do it. And we are more than open to have conversations, partnerships and so on to continue to advance in creating value to all these patients out there and shareholders alike. Now you're not going to be hearing from me for much longer here, so I just wanted to take a moment to first introduce and thank 2 of the world experts in epilepsy that are joining us here today. Dr. French, she's going to be talking a lot about really the, I would say, conundrum as you're going to see about treating patients. You can read the bio here, but I think it goes without saying [Audio Gap] provoking us, we should do better. That is, the bar is too low right now. And I'm sure Dr. French, that's what a little bit what we're going to hear from you. Thank you so much for being here. And then Dr. Friedman is going to walk us through a little bit what is the life of a physician in a given day treating a patient. And the first time that we talked about it, and I saw his talk was quite moving because we think cohorts in drug development, right? Every time we're looking to this, we're like, okay, how many patients? What is the cohort? What is the average? What is the general distribution of a given issue? But when a physician is treating a patient, they are looking into the patient's eyes, on their parents eyes and saying, that's what I can do for yourself, for your kids, for your parents and so on. And that's quite powerful as well to understand. So in today's agenda, after this boring introduction by myself, when I get to the actual meat of it, right? So Professor French is going to talk a little bit about what we just discussed, like what is the unmet need right now in epilepsy management? And not the unmet needs superficially, but what is really going on, what can we do for these patients? And I'm sure you're going to find it incredibly enlightening. And then Dr. Petrou, our Chief Scientific Officer, is going to talk about what we do in 2 aspects as a company. The very first one is how do you think about the system? So from a system perspective, how do you get these things together? What is the secrets that we have here? And to be honest, some of them are protected by patents and so on. So it's a little bit of a secret as well. And you're going to hear from Steve, someone I absolutely adore working with every day, and of course, and who does not sleep because he joins every single meeting that we have from Australia, which is I very much appreciate, Steve, from you. And once we talk about the tools we use, Steve is going to say, okay, how did we do it? And there's a lot of cool things that we're going to be hearing that we've been doing is being a little bit on the back burner from the public side but not from the internal side. Dr. Friedman is going to come after and go through this case and the idea of why do we need other drugs, why those drugs, and hopefully, we're going to be having hopes in our strategy, how we say that. So a lot of people here are going to be like the Praxis, at least they're going to say, really, Marcio, hope? It's not hope what is the strategy we have to actually drive this. And Bernard is going to come at the end and talk about our 3 epilepsy programs, going to be in the clinic by the end of the year. Now we're not into bragging as a company. We believe delivery is better than bragging. But Nicole has been reminding me and Bernard that we should be talking a little bit more about what we did on the last 12 months. And we did clear 6 INDs with the FDA during the last 12 months. We think that is just what companies should do. And that's why we have not been talking about them every single day. But I'm going to be forever indebted to the team at Praxis for working so hard, so diligently, so close to the agency instead of complaining on the public calls, driving to get those things delivered. So I just want to say a big thanks to the team because obviously, we're here today, but we're supported by 150 other Praxians out there. Much more to come, many more registrations, trials, obviously, incredibly excited to be kind of here-ish virtually again in a couple of weeks to talk about 944; in several weeks, to talk about 114, and then to continue to build the company. But without further ado, Dr. French.

Thank you. It's a pleasure to be here and also to see how many people are interested in epilepsy. It makes my heart feel good. So I'm going to talk a little bit about unmet need in epilepsy management. Here are my disclosures. I work for something called The Epilepsy Study Consortium, which is a nonprofit, and I am happy to work with many companies that are developing drugs in the space, including Praxis. So by a conservative estimate, you saw some of these statistics already, 50 million people worldwide have epilepsy and the annual incidence ranges from 20 to 73 cases per 100,000. 1 in 26 is the number we use at the Epilepsy Foundation, where I'm also the Chief Medical Officer. And 5% of the people in this room have or will have a seizure in their lifetime. And the most common incidents is in childhood and a little bit of a plateau in the middle years and then a sharp increase as we now understand over the age of 60. So you guys, your risk isn't over yet. About 30% of patients have an identifiable neurologic disorder, genetic underlying predisposition or whatever. And the rest, actually, we don't know why they have epilepsy. And we make a diagnosis in a number of different ways based on what type of seizures they have, the clinical context in which they occur, the EEG which they present with. And we have been very fortunate to have many, many new antiseizure medicines in the pipeline. And I have added the most recent one on this slide, very up-to-date, ganaxolone in the last couple of weeks was the first drug to be approved for CDKL5 deficiency disorder. But despite our success in getting many, many drugs out there, some might say we haven't made the progress that we may be expected to make. So probably 30 years ago, this was -- this initiative was started by Professor Brodie, who many of you may have heard of, in Glasgow, who has a first seizure clinic. And he sort of established the numbers that we all use now, which is that if you gave somebody an antiseizure medicine, and it didn't seem to matter really which one it was, about 50% of the people with newly diagnosed epilepsy would have their seizures come under control, and they could expect a very nice life as long as they continue to take their medication. If they failed that first drug, they still had an opportunity to improve with the second drug but that only added about 13% additional seizure-free people. And many of those people were people who had failed drug number one not because they continue to have seizures but because they had a side effect that they couldn't tolerate. But once people had failed 2 antiseizure medicines, the likelihood if you went to a third monotherapy, a fourth monotherapy, polytherapy or whatever you wanted to do, there was very little additional benefit after that. And in fact, I was on a commission many years ago that established the definition of treatment-resistant epilepsy as failure of 2 adequately used antiseizure medicines to control seizures. So studies in the 1980s, as I said, established that critical ratio of treatment of 60% to 70% response to antiseizure medicines and 30% to 40% treatment resistant. And we all expected, with all these wonderful new drugs coming forward, that there would be a change in that ratio. But the data does not support that. So Kwan and Brodie went back to their clinic just not -- it's actually now a decade ago, but in 2012, the majority of the drugs that you saw on my slide were already available to these people by that time. And yet the number that were seizure-free had only risen from 64% to 68%, which is rather disappointing. So that leaves us in the position that we are in now, where we have treatment for 2/3 of patients, and we can pretty much give them a variety of drugs. None of the drugs are perfect, but we can give them a variety of drugs and they will become seizure-free. So what do we need? That's what we're here to talk about today. We need treatment for that 1/3 of people who have not gained control from the other antiseizure medicines that are currently available. We desperately need treatment for difficult pediatric syndromes and the rare epilepsies, those 35 -- with the 35,000 individuals that we were talking about, because even the small gain that we've made in the common epilepsies, we have not made in the rare epilepsies. And essentially almost all of them continue to have seizures. And many of those, as you know, now are identified as monogenic epilepsies and more and more over time, are being identified. We would love to be able to predict, and I've spent a lot of my own research time looking at, can we take somebody who first shows up in the clinic and predict that somebody is going to do well or somebody is not going to do well so that we can move on to more aggressive therapy earlier? We need improved options for newly diagnosed patients because even though we can get them seizure-free 2/3 of the time, they unfortunately have to tolerate a lot of chronic adverse events in order to do that. And I would love to have more medications where people can take them and not have an impact on their quality of life. We need drugs that attend to people's comorbidities in the common epilepsies and in the rare epilepsies. In the common epilepsy, epilepsies, depression, cognitive slowing, memory impairment are often coming along for the ride. And in the rare epilepsies, depending on the monogenic epilepsy, there is a host of comorbidities that we know of, whether they be cognitive or psychiatric or movement as we heard. And of course, finally, we would love to have drugs that don't only work on seizures but work on the underlying disease. So just a moment to say that even though we get 2/3 of people seizure-free, they are also burdened because we are not treating the disease. By having to take that medication on a daily basis, some of them for the rest of their lives. And I and Dr. Friedman can give you many examples from the clinic of people who had perfect adherence for years and years and years and then missed a single pill and had a seizure. And that kind of burden is one that I think is hard for all of us to even fathom, that missing a single pill could have dire consequences on your life. We seem to have no trouble finding drugs with novel mechanisms. And even recently, for example, more potassium channel openers, lots of GABA-PAMs, mGluR2 PAMs and so on and so on. But will those drugs translate into better treatment, better efficacy? And interestingly enough, to date, it almost seems like the opposite, that we understand the drugs that have mechanisms that treat epilepsy well symptomatically. And we find that many of these novel mechanisms have not translated into better efficacy or better tolerability. You have heard me use the term antiseizure medicine many times already this morning rather than what you probably all understand to be the term, which is antiepileptic drug. And the International League Against Epilepsy, which is our naming body, is actually in the process of officially changing the name, so get used to it, to antiseizure medicine. And the reason is because we feel it's extremely important to signal to everyone, to people who are developing new therapies, to patients, to our community that the medicines that we are currently giving are symptomatic treatments. And I use as my example this little guy here. So all along here, we have been treating -- if he had pneumonia, it's like we're treating his cough, and we may make him feel better or we give him an antipyretic for his fever but we're doing nothing for his underlying pneumonia. And obviously, we would like to not only treat the symptom, we would like to treat the disease. We would like to get mechanistic. We would like to get disease-modifying. And we certainly hope that we are on the cusp of having treatments from all the pharmaceutical companies that will provide us this benefit. And we're sitting here at the dawn of some of these new therapies saying, can we predict which drug is going to be better? Which drug is actually going to help our community? Is going to advance us? And everybody who was on that list that I showed to begin with, certainly thought that they had a drug that was going to move the needle. And it turned out not so much. So how could a drug differentiate that's going to help us in the clinic, that's going to help with the patients that Dr. Friedman is going to tell you about? Well, a drug could differentiate in a number of ways that would be important to us. The first one I've already discussed, which is it could be disease-modifying. It could be targeted to a specific population. Now we're sort of shooting darts in the dark and hoping that it hits. But if we knew that a certain drug was better for a certain population, that would really help us because many of these patients, once they are treatment-resistant, have to go through 10 or 15 drugs and eventually, they may indeed find 1 that benefits them, but the odyssey to get there is pretty terrible. We could have a drug that has a clear and indisputable advance in treating resistant epilepsy. And I will give a shout out to cenobamate because of the recent drug that had been approved. That one certainly did get more patients seizure-free than we've seen in -- for past drugs, and we need more of that and better of that. Seizure freedom, that's what we're looking for. That's what our population is looking for. We need drugs that are better tolerated. Again, we don't want to -- I tell companies when I talk to them, don't presume that just because somebody has treatment-resistant epilepsy, that they're willing to tolerate more side effects to get seizure control. They still have to live their lives. If you were them, you wouldn't want to like sleep 15 hours a day in order to have seizure control, and neither do they. Less issues for women of childbearing potential because we have very few drugs that are safe for the developing child as well as safe for the woman who is taking it during pregnancy. Specific efficacy in difficult syndromes such as Dravet and Lennox-Gastaut. And if we had one that was longer-acting, so that adherence would not be as much of an issue, that would be a blessing as well. Seizure freedom is what our patients are striving for. In add-on studies until cenobamate recently, usually less than 5% of subjects were able to obtain seizure freedom even for the 3 months of the treatment period after we randomized them to drug or placebo. So there is great opportunity to develop new therapies that increase the rate of seizure freedom or even get people to 75% to 90% seizure reduction. This number was created a long time ago of what constitutes a benefit. 50% reduction in seizures constitutes a clinical benefit. And maybe that's true, but I can tell you that our patients certainly are not satisfied when they get 50% reduction. If you're having 7 seizures a month, then you go down to 3.5 seizures a month. You still don't know when you're going to have a seizure. You still can't drive. You still potentially can't go out in public. You still might injure yourself, et cetera. You still might die of sudden unexplained death. But as I said, it all comes down to risk versus benefit. Don't forget that the balance of adverse effects and risk of harm to benefit is important even to individuals with treatment-resistant epilepsy. And there is something else about adverse effects that I want to emphasize that's really important. And that is that most of our drugs that have known mechanisms now, such as sodium channel blockade and a GABA increase, et cetera, have a clear dose-response relationship in clinical trials. The more people get, the better off they do. But unfortunately, if you look at this graph, what you're looking at is sort of a theoretical construct of the percent of people who might achieve seizure freedom with a drug and the dose that they can tolerate. And they can only get benefit up until the point where the side effects become intolerable. I remember vividly a guy who is in one of my clinical trials with a drug that's already on the market, and he was having a seizure every single day. And in the randomized controlled trial, he became seizure free. And I was thrilled. As his investigator, I was completely thrilled, until he came to me and said, "I'm dropping out." And I said, "How can you drop out? You went from a seizure a day to seizure-free." And he said, "I can't stay awake at work. I'm going to lose my job. I'd go to my desk and I fall asleep at my desk. I cannot have that, and I can't stay on this drug." So there is a toxicity limit to how much people can take. And imagine if you could take that same mechanism -- that didn't work. Can we go back? Yes. That -- no, it's not -- the animation isn't working, and it's such a good animation. But -- so we'll tell you about the animation. So if you could move that toxicity block over, you could actually uncover more efficacy. You could get more people seizure-free. And the example that I just gave you was the critical example. You could get more people seizure-free. So you don't only get the benefit of improved tolerability if you can reduce the side effect burden of a mechanism, but you also potentially get the advantage of more seizure freedom. And we've seen that, for example, in sustained-release formulation. Same drug, but just you don't have a peak dose toxicity, you can give people higher exposure over the course of the day and more people will be seizure-free and happy. So let me just change gears -- shift gears for a little bit and talk about precision therapy. So recent studies of emerging antiseizure drugs have targeted the rare epilepsies. Dravet syndrome has been targeted by fenfluramine and cannabidiol; Lennox-Gastaut, clobazam, rufinamide, cannabidiol and fenfluramine. And the question is, we're talking about the benefit of precision medicine, is this precision medicine? So up until this time, the studies have determined that if you take this drug and you give it to the specific population, it works better than placebo. However, if you had taken that same drug and given it to a different population, would it have worked better? Would it have worked worse? We have no idea. Or if you took a different drug and you gave it to that population, maybe even 1 that's already on the market, would it have worked equally as well? So our precision therapies to date, I worry about pseudo-specificity, meaning if I gave a drug to blue-eyed people with depression and it worked, I could claim that my drug was beneficial for blue-eyed people with depression. But that doesn't mean it doesn't work on brown-eyed people nor does it mean it's a better drug for blue-eyed people than some other drug. But it sure gives the impression it does because I tested it on blue-eyed people with depression. So these studies have not proven either that the drug is more effective than other potential therapies or that the drug will be more effective for this syndrome than any other syndrome that could be tested. So we need better precision in our precision therapy. And we hope that we're getting there, right? We want precision therapy with disease modification. We want targeted drugs that actually target to the pathway that we already understand is present in those patients. That's the hope for the future, to correct pathology caused by a specific mutation or mutations. And we got a little hint of that maybe with everolimus for tuberous sclerosis because in tuberous sclerosis, we know what the mechanism is, which is an mTORopathy. Everolimus normalizes the mTOR pathway and also is disease-modifying for some of the other problems that happen in mTORopathy. So you're not only targeting the seizures, you're targeting other comorbidities. And that's where we want to get to, right? And targeted gene therapies is also a hope for the future. One that's already in the clinic in trials as antisense oligonucleotides with haploinsufficiency disorders, with 1 bad gene and 1 good gene, where you're stoking up the protein of -- increasing the amount of protein that's missing because of the bad gene that can't produce it. So do epilepsy patients represent a satisfied market? I do not believe they do. There are many issues with existing antiseizure medicines. 1/3 of them continue to have seizures, 1/3 by estimation, have dose-related side effects. There are no disease-modifying therapies yet. And we are hoping that in the future, we can really address many of these problems and hopefully, we'll be hearing something exciting for the rest of the day. So thank you very much for your attention.

I'll serve as MC just quickly. Just wanted to invite Steve to come here. And as a lot of you know, he's going to do a little introduction, but Steve's being a co-founder of the company, a lot of the ideas from the very beginning come from his lifelong dedication to CNS. He's going to go through a little bit on how we do it, why we do it and why we are so excited here at Praxis.

Thanks, Marcio, for that. As Marcio said, I'll just give a -- if you'll indulge me for 30 seconds, a bit of a background on how I ended up here. I was a career academic, being dedicated to really neurobiology of disease for the past 25 years of my academic career. Up until about a week ago, I was the Director of the Florey Institute of Neuroscience. It was Australia's largest neuroscience institute, about 700 staff members, quite a large operation with a broad focus on neurodegeneration, epilepsy and basic biology. I've been involved with Praxis, as Marcio said, since about 2015 when it founded. And it was founded really on the cusp of an amazing advance that we saw in the genetics of rare disease, as Dr. French was mentioning, at that time, really, for the first time, understood the genetics of these sporadic cases. And that presented an amazing opportunity to really do something about it and consume that knowledge. And that's been a really big part of my career is how do we take the learnings from clinicians, from basic citers and actually turn it into impact. And I've been very, very passionate about that. I've worked with parent groups, some amazing parents I've worked with over the years. I've worked with the founder of Praxis. And I found it almost irresistible to not get more and more involved that for the past 2 or 3 years, I've been working on my day job for 10 hours and then putting another 10 hours in the evening, sleeping about 3 or 4 hours a day. And I've made the commitment. And Marcio set up, I think, such an amazing leadership team and an amazing team within Praxis, but it's just too compelling. I'm leaving my job as Director. I'll have a small footprint to keep the lab going in some projects that are collaborating with Praxis, but I will be doing 95% of my time as Chief Scientific Officer. And I'm just so excited about the future of that. So excited where the field of genetics, where the clinicians and the research have taken us to, to this point. So -- and I want to share a little bit of that excitement today. So that's enough about me. Let me talk a little bit about why we're here today. So this is a great question. Why epilepsy? And why now? And I think whenever you answer questions like that, it's always good to know where you come from. And I won't to do a long history lesson because that's not what we want to do today, but this is where it all started. Why we are talking about why Praxis exists? Why we're talking about making medicines? Why we're informed by genetics? It comes from the studies of a noted American epileptologist called William Lennox. He was studying twins. He's asking the question, is there a genetic element? So he studied identical twins and he studied nonidentical twins. And what -- unequivocally, what he found, if there's epilepsy in an identical twin, the chances of epilepsy occurring in the other twin was very, very high. Way higher than in nonidentical twins. And this is just 1 example of many of 2 lovely girls from the '40s, Carolyn and Eleanor, 16, 17 years of age. They both presented with early morning tonic clonic seizures. They had a very similar EEG. In their early 20s, they both developed psychosis. And it says 2 things: epilepsy and comorbidities do exist, and I know Dr. French talked about that, they're very important things to consider. But it really was the first signs, and he studied many, many of these twins. And so there's another twins actually we contacted when they were so young girls who had absence epilepsy. And we managed to track them down with my colleague, Sam Berkovic. And it's just amazing to see that the story they've had over the last -- this is what, 80 years of advances in genetics. So what is it about epilepsy versus all the other disorders, neurological disorders that have clear genetic mechanisms? There's common disorders like migraine and depression, very -- sorry, very common in the epilepsy that fits in that middle category of common, and you saw the large numbers that Dr. French mentioned, and then the not so common. But what is it about epilepsy? If you look at it, the tribe has spoken. Look, this is the mention of the word epilepsy or seizures associated with genetic conditions in what's arguably the most comprehensive database of human genetic disorders, the Online Mendelian Inheritance in Man database. And you simply look at the occurrence of that. And if you compare it amongst all the neurological disorders, and even around all other diseases, the number of mentions and the number of papers and the number of genetic findings is so much higher in epilepsy. If you look -- if you aggregate those 2 boxes, you've got a multifold increase in the amount of activity in that space. So the field has done this for us. And they haven't done it because they want to crack epilepsy. They've done it because epilepsy is yielding its secrets. Of all these disorders, the genetics of epilepsy is coming more readily than it is for these other diseases put together. And that's an amazing -- it's an amazing opportunity there in this knowledge as there's a huge database of knowledge here. And we're starting to learn so much more about the architecture, and I'll touch on that in a second. When we look at the -- when we look at all the discovery, there are hundreds of genes now being implicated in epilepsy. And if you look at those, what's interesting, there's a lot of complexity. My goodness, how do you deal with 100, 200 different things and each of these genes has got multiple mutations, et cetera? But they are starting to fall into certain classes. Some that might be obvious, an ion channel makes sense because the brain -- for epilepsy, the brain is a moment-to-moment machine. And epilepsy emerges in those very short time frames. You change a level of excitability, you can quickly change the brain state from being normal to epileptic. Ion channels and synaptic proteins participate that in the short term. But then there are longer-term things that happen. The mTOR pathway can drive cortical malformations. And a lot of these things, and we'll talk a bit later at Praxis around tuberous sclerosis, which is a change in the structure of the brain and that region and those little tuberous are epileptogenic zones. And then we're thinking about those chromatin remodeling and how do you change DNA in the long term. We're starting to see patents. So it's really important to see patents to help you guide your decision-making for which targets do you engage with, why do you engage with those targets and how do you engage with them, how do you use that knowledge to increase success in delivery of drugs for our patients. And that's something that drives us, something that I live and breathe daily and I've been doing that for as long as I know. And I'm just, as I said, so delighted to be in a team now that wants to work on that vision, share that vision and make it happen. Now when you think about almost any genetic disease, this is sort of a template, if you will. So how does this disease work? What are the genetics? And there are basically 4 boxes. And I guess those who have been in a business school, which probably a lot of you here, a lot of things look like this, which quadrant do we think about? On the bottom axis, you've got, well, is it a common genetic mutation that we all have single nucleotide polymorphism? We all have these things. And some of them are very common. We share them and the collection of those determines how tall you are, how smart you are, how high you can jump, all these things. Then there's how strong is that genetic change? What's the effect size of that? And that's what's on the vertical axis. Now where does epilepsy live in? We've been -- I've been working in this field for 25 years, and when we started, we had -- so there were 4 eras of epilepsy, really. There were the twins, then there were the families, there are now the sporadic cases, and then there was the common population epilepsy. 4 major chunks of discovery. The last 2 are the 2 that have been -- that founded at Praxis, and I think the very last one I'll talk about later is the future of Praxis as well. Where do we sit? So if you look at the top left quadrant, families sit in that quadrant. It's a rare family, a single gene inherited in that family causes epilepsy. If you're lucky enough to get it from mom or dad, you've got epilepsy. It's a rare gene. It's not in a general population. And if you get it, you epilepsy. And we would say that's a rare gene, the large effect size and it captures rare monogenic epilepsy. So that was -- this first era was twins, something genetic. The second era was families. We're finding rare genes, but we can't really -- and I actually started my involvement with biotechnology in around 2000 for a small Australian company Bionomics, where I was the -- took care of their CNS research on a part-time basis to try and take advantage of that. But it was very quickly apparent that there weren't enough. These genes weren't common. These families were rare as hens' teeth that we couldn't -- it wasn't actionable. It was almost too early to use that knowledge to move forward. But what happened in that third era where complex -- where we started to understand the causes of sporadic, rare and very severe epilepsies. So back in quadrant 1 again, that's where those rare diseases that Dr. French was talking about, that all gave birth to Praxis Precision Medicines. We started to realize that there were a whole slew of genes, and there's been around 200 or so now that have been identified, where mutations in those can give rise to de novo epilepsy. So the parents don't have the mutation. It's a random event that occurs and there's a mutation in there, and then this child now unfortunately has a very severe epilepsy syndrome. And we started -- that started to crack because of the sequencing technology got so cheap. We could sequence whole exomes, whole genomes, parents, et cetera, and find these genes. Now but what's happened now, and this is really very recent, is that we're starting to say, well, what gives rise to common epilepsy? In the rest of the 90% of the population of people with epilepsy, what gives -- is that other rare diseases? Do they have single gene that's causing it? What's actually going on? We have no -- it's only now we're beginning to understand that. And what it is, like the same thing that determines height and weight and all the things that are human features, it's the accumulation of common genetic variation that's driving the emergence of epilepsy in the population. So the same thing that could make you tall, the same potluck of genetic combination can give you epilepsy. And that's really important to understand that because once you understand that, you're much better positioned to know how to act to reverse it. And I think it speaks -- and Dr. French so beautifully set that scene for that in saying, how do we -- it's not just random mechanism. And we can all be very clear in the lab and figure out, well, that's an interesting mechanism. How do we take advantage of that? How can the genetics tell us? It tells us how the system breaks to get epilepsy. There's knowledge in that on how to correct it. And that's -- and then the next era, I think, and something Praxis is actively positioning for, doing the thinking, doing the networking, understanding how to act on this new wave. There's more and more data is going to come out in the subsequent months from now, and we're very much primed to be at the forefront of understanding that data and acting on it for the benefit of our patients. So we went from that -- the twins of William Lennox. This is sort of where we are today, what the genetic landscape of epilepsies. And what you can see on the left is the rare monogenic epilepsies. Those single families I mentioned. The sporadic cases in these kids that have a very powerful mutation that gives rise to a very severe and debilitating epilepsy with no recourse. There are no drugs that are useful for these kids at this time. As you move across, you see the common polygenic epilepsies that are due now to the accumulation of various amounts of genetic variation that we all possess. It's just the accumulation of the wrong set of those common genes that give rise to the epilepsy. To the very right-hand part of that, which is 100% acquired epilepsy, traumatic brain injury or head wound, a tumor, cardiovascular event, some environmental toxin then gives rise to 100% acquired form of a epilepsy. I think this captures -- and then the prevalence is a guide, it's not exactly to scale, but it's just a guide to say where the sort of relative numbers of patients are in this spectrum. This is a very powerful area, and we've been using this at Praxis since day 1. And engaging with -- if you look at our pipeline, how are you going to do depression? How do you get into essential tremor? It's always been informed by our understanding of genetics of epilepsy because genetics and epilepsy shine light on genes that are critical for brain function. Whether that's disease or not, the genetics is incredibly informative of that. Just take for an example, our PRAX-114, which is a GABA-PAM. I mean at the time, we were one of the longest-serving Praxian at this point in time. That's a devious honor. This means I'm old. I think I'm the oldest person at Praxis, we are both. What do you think, Marcio? And he never let me forget that. So the PRAX-114 is a neurosteroid, a GABA-PAM. We know and we knew some 10, 15 years ago that this gene, the target for PRAX-114, is mutated in epilepsy and gives rise to a very common form of epilepsy. It wasn't a common mutation, a very common form of epilepsy. We know that, that same gene and that same system is involved in some mood disorders. Catamenial epilepsy is a form of epilepsy that females get during different times of their menstrual cycle. And that's due to fluctuations in the levels of neurosteroids that act on this particular gene, and it can give rise to mood disorders and can give rise to epilepsy as well. So we knew a lot about the genetics of this, a lot about the neurobiology of this. So the decision to take it into that direction was very much informed by that knowledge. We knew also, and I'll show that later, that it's a good antiepileptic drug. It actually stops seizures. That gives us a lot of validity from a biological perspective that it's got the actions that are consistent with the mechanism that would be helpful in depression. Similarly, for 944, I won't go through all the gory details of that. Very similar for that, a genetic story, a neurobiological story consistent with -- this is a gene that's important for sleep because of a certain rhythm that's generated in the brain that makes you sleep. That same rhythm can go awry and cause the type of epilepsy called absence. That same rhythm occurs in other parts of the brain responsible for movement. So the neurons, the networks that are formed by and by that are very sensitive to the function of these types of these calcium, T-type calcium channels. That was part of that. But then in a more obvious space, when we look at our rare epilepsy programs, they sit firmly in the left-hand part of the spectrum. They are deliberately and expressly designed to reverse the effects of a genetic lesion that causes the disease. And that's almost the simplest case. You have a gain-of-function mutation, can you make a drug that reduces that? You have a loss-of-function mutation, can you make a drug that increases the activity of that gene to directly counter the causal effects of your seizures and your epilepsy? And that's so important because -- and I think, again, Dr. French said it so beautifully that we are good at symptomatic, developing drugs that give symptomatic relief. We do not achieve the reduction of the attendant comorbidities when we're so focused on symptomatic relief. When I did my training, the way to develop an antiepileptic drug was to get a cat, was to put -- buzz it with electricity and give it the drug and see how much you have to turn up the knob on the electricity for the seizures to reemerge. And that gave wonderful drugs that stopped seizures, but they had nothing to do with the root cause. Each of these agents, each of these, we know precisely the genetics that we're targeting, we know exactly where we're going and we're trying to reverse it. That was the mantra, that was the founding of Praxis Precision Medicines, why we wanted to do it. Now there's a sort of a third bucket of that. Because we start the genetics, it has shine light on our understanding of neurobiology. We understand, in this case, we're looking at focal epilepsy with molecules like PRAX-628. We know -- and I'll talk a little bit about that, about what we designed into that molecule in order to make it specific for that particular type of epilepsy. But we're using the knowledge and we have specific ideas that emerge because we know why seizures emerge, what causes excitability in a neuron and therefore, what is the point of sensitivity. What's the Achilles heel of that process? Can we make a drug that targets that weakness? Now this is the future I mentioned for Praxis. This big area in the middle of common epilepsy caused by aggregation of common genetic variation in the population. And we're thinking very strategically now about how do we use this to develop the next generation of drugs that will do both things, stop the seizures cold, but also because they're tackling root cause mechanisms, have a much stronger probability of tackling all the comorbidities. We're not going to put that fellow to sleep now in order to stop his seizures. That's something that's very important to us at Praxis. And you'll see all the different ways, as I speak later in the session, about how we're thinking about achieving that balance between proper efficacy, trying to do as much for the comorbidities whilst not entering into that black hole of toxicity. And this sort of overview -- and I think we're going to pause here for a quick break and a coffee break. This sort of overview, the 3 broad strategies at play at Praxis for how we think about developing drugs. All of these informed fundamentally via genetic mechanism. Each of them can also inform around an indication expansion when we're ready, when the time is right, when the opportunity is there, how do we take this knowledge and move into a different space. And as I said, epilepsy is the sand pit on which whole slew of brain functions can be revealed. And we want to really want to play in that sand pit and we want to have a broader impact on human brain health as we possibly can. So I think I'll leave it there for a short break. We'll have a bio break and a coffee. So thanks, everyone.

Thanks. I think he means short. So maybe you're going to do like a 5-minute break, just stretch, get some coffee, come back. And then we're going to talk a lot more about how the action from this conceptual framework gets drugs into the clinic and helping patients. So back in 5 or so. [Break]

Then hope you got a coffee and a bit of a break. So as I said, I talked a little bit about our broad thinking and our strategy in the previous session. What I want to do now, briefly touch on how -- our model of how we get things done and then just go a little bit deeper into several programs to give you a flavor for how Praxis actually thinks and conducts and executes on activities. This slide -- there's 2 aspects to this slide really. The top part is one of our 4 pillars that Marcio showed us, and it really speaks to our thinking in the translation space. We think translation very, very early. When we're thinking about a program, we think about the patients, who are those patients going to be, how do we identify them, how do we align the genetics, the therapy and the patient group at the same time because that speaks to plausibility, feasibility of moving forward in those areas. We have a whole slew of models that we did. Epilepsy emerges at a very high level. There's a brain -- high brain level behavioral seizures, high-level electrographic seizures, but we start from single genes. We need models that enable us to connect the dots between those different scales, so from the molecular to the neuronal to the network and to the whole brain and behavior. We do that. We're interested -- very much interested in biomarkers, whether they're electrical, whether they're biofluid, even this concept of endophenotypes, changes in behavior or in performance of something that's not part of the core phenotype, it gives you a sign of that disorder and a sign of efficacy, then patient stratification, genomics, informatics, functional genomics. It's a lot. It's a lot for a small company. We're very lean. We're very efficient as an organization. And we think collaborations are the key to us expanding our footprint. And as you can see below, just a sample of the organizations that we've got some deep and -- collaborations with that help us with our mission. And we think this is really important for Praxis to move forward rapidly, to move forward efficiently and to not to have so much inertia that we get caught in an area, we're caught in a box because we've developed, we put too much internal resources, we've built the laboratories and everything, and then that becomes a boat anchor. So we need to remain nimble so that we can move. And we've solved this, we think, with this approach. It's not slowing us down. We all work hard, all the teams work hard with an amazing team of people who reach out and collaborate with these people. So we're very blessed in having highly functioning science team within Praxis and the clinical team that can pull this off. So I'm going to touch on the 3 strategies that I left you with in the first part of the talk. I'm going to give some examples programs that emerged from each of these 3 strategies, always informed by genetics, but have moved forward into programs. So this one, I'm going to focus on nodes of pathological convergence informed by genetics. It's always been the mission, how do you reduce the complexity of -- we've got 20,000 genes in the human body. Each of them can break. They can break in different ways, give rise to different disorders. Sometimes it's good to -- rare diseases, we can tackle them one at a time. But in many cases, we want to understand whether or not from the many come the few. Does it converge to a single point? Is it network? Is it a neural network? Is it intracellular signaling network? But we can do 1 drug, 1 process can subserve a lot of different therapies. A really important area is voltage-gated sodium channels. Historically, epilepsy has many of these. These drugs emerge whether you're screening phenotypically or mechanism-based. Why? Well, the genetics has told us that mutations in these 3 sodium channels, Nav 1.1, 1.2 and 1.6, is the cause of many, many rare diseases in epilepsy, unequivocal mutations cause disease. We also know that these genes are starting to emerge in that common gene issue that's in the common epilepsy patient. We know that Nav 1.1 and 1.2 are powerful signals for developing epilepsy in that other domain, really important genes. Why are they important? Well, this is a little cartoon of a nerve. There's a particular part of the nerve where -- the brain is an electrical organ, all the signaling, the way it speaks to itself in the short term, whilst I'm doing, talking, moving, everything's happening electrically. The slow stuff, this chemical signaling, the structural change, that's all slow. That's when you learn how to play a piano, you have plasticity, that takes a long time, it's an impossible for me, but it takes a long time to do. That's a different process. But moment to moment and the way that our behavior, our mood, depression, all of these things that can be governed by the moment-to-moment function of these channels are the source of that signaling. Essentially, they are like a spring-loaded faucet. You whack them with a hammer, they give a spurt of water and then they rapidly shut. That's the -- and the brain has to talk -- has a -- we know about computers having clocks, and they're measured in billions of cycles per second, gigahertz. The brain has got its own clock, and it's determined by the function, what's the fastest biological process in the brain is an action potential. That's that little squiggle, a little brief squiggle of electricity that occurs when these channels open. That's about a millisecond. So the fastest clock we've got in a single neuron is about 1,000 times per second, and that -- very rarely is that reached. So that's a basic clock. That determines the basis of everything. So you can understand, in an epilepsy, those neurons are going haywire. They're firing at speeds they shouldn't be operating in. So they're in new operational envelopes that they shouldn't be in, driven -- and these channels, either they're directly causing that or their bystanders participating in that process. So very, very important to our signaling. They also can break through genetics. They can break through being in the wrong place at the wrong time and being activated in a certain way. And a really, really tiny thing. This is like the old story about a butterfly flaps its wings and it causes the tornado in China. It's the same sort of thing. This is a tiny, tiny change in these channels. I mentioned that faucet, that you whack it with a hammer, it springs open and springs shut. But if that faucet doesn't close and it drips the tiniest amount of water, that causes a problem in the brain. And you can see here that little black line is what a normal channel does. You can see the blip is the blip of water when the thing opens and rapidly closes. The red is the dripping of the faucet. It's called persistent current in a sodium channel, and it's a tiny, tiny amount of current. We see it in genetic mutations. We see it in channels that have been subject to a lot of activity such as in epilepsy and it causes a major, major impact on excitability. So we think it's a really important note. So in thinking about how do you deal with a pathology caused by sodium channels, you have to deal with this. You can't ignore it. Happens in patients' 8A gain of function, epileptic encephalopathy case. You can see the baseline on the left, you can see that persistent current on the patients on the right. It happens in 2A patients, the black at the top, the faucet shuts off on the red, there's a constant on persistent current on the right is the way that the biophysicists display it, compare the purple to the black. So what does this mean? What effect does it have on the way Praxis thinks in engaging to make a drug? So you can think about sodium channel drugs and a lot of companies and others think about subunit selectivity. We're going to make a 1.6-specific agent because we know the 1.6 channel is important. Nonetheless, even if you achieve subunit specificity, you still have the problem of over-engaging that channel. You can't turn the volume completely too down on the television and still listen. You still need some volume, but you don't want it to blare. And that's the problem, the same problem with a nerve in the disease state, it's blaring. It's going too hard. You want to do something that turns it down but doesn't stop it. And that's -- so even if you've got subunits of activity and you got -- this drug engages with all the channels, it's still going to turn the volume off for that particular thing. So we think in functional terms, how do you stop those channels that are participating in the nerves that are overly activated so the channels they contribute to that over activation. But we can -- we know from an electrical perspective, channels change their function, the persistent current, all of those things contribute to overexcitability of neurons. Can you make a drug that's tailored to the high activity and allows the normal activity to go through? That's what that graph is showing. We think of nerves have this thing called an input-output curve or a transfer function, you stimulate them, they do something. Now generally, in the disease state, a little bit of stimulation does too much. So you've broken that cycle, you can -- and that's what that black represents. You've got too much activity for the input that you've got. But we do want the stuff on the left to happen as normal so that our cognition, our movement, we're not sedated. All those other things, all those things that Jackie said, that gray box is her black hole of toxicity that she couldn't remove from the presentation. How do we avoid that? Well, if we can get rid of the persistent current, that's one way of doing it. If we can have a drug that binds rapidly and unbinds rapidly from the channel, that's another way of doing it. Because your drug needs to interact with the channel in real time. Some drugs bind and latch on and take a minute, half a minute to drop off again, they don't have the right cycle time to be tracking the activity of a neuron. So these are 2 very important criteria that Praxis considered to produce functional cell activity. We think that with a focus on functional selectivity means you can make a drug that dampens down that hyperexcitable state, allows normal -- and should -- we should enjoy more tolerability, and Dr. French mentioned how important it was to have tolerability. And this is one of the mechanisms at Praxis. So this is what we wanted to do, and I'm pleased to say we've done it twice. 562 and 628 are 2 examples of how we started with an idea, and we've delivered 2 molecules that do this. And let's have a look at some of their properties relative to other molecules. This is some of the areas that we're going to put in for 562. SCN8A gain of function, I mentioned. I showed you that. Those patients have persistent current, have too much neural activity. SCN2A, those patients have persistent current, have too much neural activity. Tuberous sclerosis, which is an interesting one. It's a cortical malformation and it gives rise to a focal seizure, but it's a very interesting malformation. We've done a little bit of work on the so-called -- these tubers that are produced. They've got various -- they've got a core in them of really, really weird neurons that have mega-sized and have an odd behavior electrically. Now that core, we know surgically, in some cases, you remove that core, you can actually have a major effect on the amount of seizures. So we know that's the epileptogenic zone. So we want to try and again, stop the activity of those neurons from entraining the whole brain, going from a partial epilepsy to a generalized epilepsy. Now molecules like 562 are some of the best anti-seizure medications I've ever seen. And I've played with a lot, put them in a lot of different models. These things stop seizures cold because they're so focused on this initiating event. So we think they're very well positioned to impact these tuberous sclerosis. Overall, we're looking at about 10,000 patients here in this collection for targeting with 562, still within the rare envelope, but we think we're going to have a huge benefit to lead the board. Now these are the old mainstays, drugs like lamotrigine and carbamazepine that have been around for many, many, many years. Two important figures here to look at, the red curve and the black curve. The black curve is how well do these drugs block that persistent current. And the red curve is how well do they block the other current that tends to lead to toxicity. When you start to get too much red, you get into the toxicity region. The other thing to look at is where they are on the dose-response curve, the potency. Look at PRAX-562, huge left shift. Its sensitivity to persistent current is twofold better than the nearest competitive molecule and manyfold better than many other molecules that are currently on the market. Its potency is many, manyfold greater than anything else on the market. That difference between the red and the blue curve is really important. That potency of the molecule, where it lies, is really important because as we all know, less molecule is always going to be less toxicity for all the other reasons. So we think it's got an incredible properties. It's binding. It's the fastest-binding sodium channel drug I've ever looked at. We're lucky to have Kris Kahlig as part of the team. He's worked on these sorts of molecules in the past, and he can attest to that incredible binding kinetics of this molecule. So it really differentiates at the biophysical level. It differentiates how it affects neurons. And we believe it's going to differentiate on what it does for patients who are taking this molecule. And Bernard will talk more about that. This just talks about the potency difference. Look at that. All the other drugs on the right, PRAX-562 on the left, hugely more potent, big change. The ratio of that -- essentially, that red curve to the black curve, the persistent current versus the other current, sixtyfold. We think that gives an enormous therapeutic window when you have that sort of difference. You can do what Dr. French said you need to do. You hit the wall, you want to go a little bit further with your drug, but you can't because you hit toxicity. These are the sorts of properties of the drugs that we want to develop and deliver that allow you to break through that barrier between efficacy and toxicity. And how well does it work when you look at it? On the left is a seizure model, maximum electric shock. We won't go into the details. It's a mainstay in the industry. It gives us good benchmarking against other molecules. You give this drug to that model, 2 milligrams per kilo produce a 50% reduction in this assay. On the other hand, you use -- how do you demonstrate neurotoxicity? One of the very common and simple ways is to see whether the animal's movement and behavior is impaired by dosing it up. And this is a spontaneous locomotor assay that does that. We dosed up the animal and we say, does it get drowsy, does it move less and you have a look at the dose response to that. The 50% point of that gives you this idea of efficacy. We've got about a seventeen-fold therapeutic index, which is massive. That's a massive therapeutic window there. We also know that it works in genetic models, the 2 patient populations that we want to give the drug to, when we give this drug, we can completely abolish seizures with these patients. With this molecule in the 2A and 8A models, similar to the patients, there are rodents, but they've got genetic changes that cause their channels to play up in the same way they do in humans, hugely efficacious. Also keeps these animals alive longer. So not only stop the seizures, but when it's fed, a lot of these animal models, over time, just have a terrible seizure and die. This drug keeps them alive. And you can see on the right, you withdraw the drug, then they start dying again, excellent evidence that it's having the so-called disease modifying. It's going beyond seizure control. Interestingly, you can -- because this is -- I mean it's such a good anti-seizure medication because of its profile, its envelope of behavior, we put it into some potassium channel models of rare epileptic encephalopathies and it was very effective in those. KCNQ2 on the left, KCNC1 rodent model on the right. This molecule is effective because of how it acts. So it doesn't have to be a mutated channel to work. But it's acting in a way and at a part of the neuron that gives us precision, not that 1:1 correspondence of a gene and we're reversing the effect of that gene. It gives us precision because we have anatomical and we have biophysical and we have neurological evidence for why this is a point of sensitivity, and we're tackling that point of sensitivity. PRAX-628. This is the associated molecule. We've developed 2 for 2 different areas that we're targeting. Adult focal epilepsy with PRAX-628. About 2 million patients in adult focal epilepsy, and these are the patients that are very -- you'll hear more from Dr. Friedman about or the tough time these patients have in getting medications to change their lives. We think PRAX-628, and Bernard will talk more about the properties of the molecule that make it better for patients, but we think it's got the right properties, the biophysical and the neurophysiological properties that are going to be of high benefit to people with focal epilepsy. When we compare it to 628, we see some differences in its biophysical properties but it's still very potent. It still has an excellent ratio of tonic to persistent current block. It's also got PK differences that are more suited to a broader population for daily dosing. 562 has got a longer half-life. As we know, Dr. French mentioned, for a lot of kids and people who have been taking this in the encephalopathy groups, compliance is really important. And here, the effects of missing a single drug will be less because the drug -- the peak to trough ratio is much less for a molecule like this, has a very good properties for having level exposure levels that can be very beneficial for patients. PRAX-628 also isn't metabolized on a 6-hour scale, which is great. Again, slightly more rapid benefit, and then Bernard will talk in more details about this in his session. Again, like the way we did the assay for PRAX-562, with 628, about a seventeen-foldtherapeutic index again. Efficacy on the left, tolerability on the right. Good separation of those 2 things. Just think tenfold difference is great for a drug. We've got seventeen-fold. So I'll leave that section there. I'll move to the -- what I think is maybe the most obvious approach, which is focusing directly on underlying genetic defects. This was the founding principle of Praxis around precision medicine in rare disorders. It's -- these are just some of the programs that we're working on. In each of these cases, the disorder is caused by a mutation in a single gene. There's lots of different mutations because they're -- these are private mutations that these patients have. There's some recurrence and that's just random. You're not -- these kids are so sick, they don't grow up to have families. They can't pass this gene on. But they have these random mutations that either turn the gene on more or turn the gene down. Fundamentally, it's too high or it's too low. It's got to be in that Goldilocks zone in order to work properly. Our approaches, whether they're ASO or whether they're small molecule, are specifically designed to push it back into the Goldilocks zone. And how do we do that? I'll talk a little bit about those. Praxis 222 is an ASO that we've developed collaboration with Ionis for SCN2A gain-of-function epilepsy. This is SCN2A, I mentioned before, is that sodium channel that's in the axon initial segment, and it's got massive gain of function. In these kids, works too hard. They have seizures, they have other mood and behavioral disorders, and they have a huge impact on carers and families alike. And we're very, very committed to helping this patient group. So what do you do? Well, you can use an ASO to reduce the amount of the channel. Antisense oligos are quite flexible in that they have lots of different modes. They can turn a gene down, they can increase the gene, they can increase the expression. They can change the binding of various DNA binding proteins, et cetera. So they're very flexible in the types of different modes that we can use them in. In this mode, simply the ASO binds the so-called gap mark, and that targets that RNA for that SCN2A channel for destruction. So we can reduce the amount of RNA, and we can do it with quite a lot of precision. This is a preclinical study that we undertook. You can see on the left, different levels of ASO cause a different amount of reduction of the mRNA. You can see that very clearly on the left. So we can titrate it. We want to knock down 20%, 30%, 80%, we can titrate that. Doesn't change the protein, doesn't change the structure. On the right, you can see the 2 gray ones, which are our controls, the orange box, given the ASO, we reduce the SCN2A protein. On the far right there, that's actually an image of that axon initial segment, the little -- the white squiggle in the ASO control, you can see that the white's gone because we've reduced the SCN2A, and we've taken it out of the equation. It's not as much in the initial segments as it was. So we've got RNA, we've got protein, and we've got immunohistochemical evidence that we are affecting the thing we want to affect. What happens when you give it to a mouse? So these mice have been genetically engineered to contain a human mutation. This mutation is gain of function. These mice do not live much beyond 20 days of age. These mice, you put them on your hand at 2 days of age, and they're having seizures. One of those seizures will kill these mice. We've never seen 1 go more than 25 days. And that's what that gray bar is that you see. So there's a survival curve over time, all the gray ones that have got the control ASO are dead. As we -- at day 1, we inject ASO into the brain. And we inject enough ASO to knock down the level by 50% or by 80%, and that's the blue and the orange. And in most cases, you can see the rescue of the survival, those animals are living longer. It's a tough time to be working with these animals because the brains are growing so much. So we did a second series of experiments where we gave a booster shot to clamp the reduction at the levels. And you can see what happens when you do that. And that's telling us that a 50% reduction in that middle figure, the animals live now. One injection -- or a second injection at day about 35 or so after -- following their initial injection, those animals now live for 120 days. And only as the ASO gets metabolized, do you see the disease reemerging and the animals start to die again. Similarly, for the higher dose, as you would expect, more ASO on board. It's going to take more half-lives of down -- of metabolism before the knockdown level goes to the point where we're seeing seizures again. They're hugely effective in these models, and we can even dose the animals later in age, which is also a question you have for people. What happens if you dose and the kid's already -- is a year old or 2 years old? You actually still see it, as long as the animals are alive and they're right at that cusp now 4 or 5 days before they started to die, and as long as we can get over the cusp, these animals survive. So really lots of really healthy signs there that we can have a huge disease-modifying effect. When we started these animals, it wasn't just the seizures. All other aspects of this animal look like a wild-type animal. So the movement and other behaviors, all these things were very, very similar to wild type, suggesting that we've not just stopped the seizures, but these animals are indistinguishable from a counterpart, from a wild-type counterpart. Similarly for KCNT1, another devastating epilepsy, this is a very interesting channel. It's actually a -- it lives in that same initial segment region. It actually detects the sodium during the action potential and participates in the formation of the action potential. It's clear now that every single patient that's got this mutation has what we call gain of function. This channel works too hard. These patients have terrible intractable seizures, they're aphasic, they're immobile, they often sit in a cot, have hundreds of seizures and die very young. And there's nothing we can do for them. What they need is an agent that blocks that channel. And that's what Praxis 020 is all about. Our program here, we developed a molecule, a small molecule that can inhibit that channel. Here, we're just testing that it can equally inhibit a wild type and mutant channel because we know these kids have got wild type. It's obviously an autosomal dominant disorder. They've got a copy of each, can we inhibit both? Yes, we can inhibit the mutant. So we don't have to worry about some complicated pharmacogenetic issue. What to do with the seizures? Well, we've developed a model in collaboration with one of our partners, and again, Praxis didn't -- they had the model. We worked with them. We dipped in, we got it done, and we moved out, went on to the next project. Here, we can show that as we give the PRAX-020 molecule, we can essentially abolish the seizures. So when you engage this channel, with the molecule, seizures go away. Other things on the -- we do EEG analysis. A big part of us is -- preclinical and clinical EEG is a big part of Praxis' thinking about translation. Here, we look at the spikes on the EEG. All -- probably -- I've looked at probably 20 or so genetic models of epilepsy in mice. Pretty much every single one of them has EEG abnormalities in between the periods where the seizures are occurring. So we call that interictally and there's lots and lots of spikes. These animals are having 1,000 or 2,000 spikes a day. They're really good markets because you can measure this over a day, you're measuring 1,000 or 2,000 events. You give a drug, you see the reduction, huge reductions in the level of spiking here. The gray is the control and the blue and the orange are dose-dependent reductions in the spiking. So we're seeing changes in seizures, and we're also seeing changes in spike frequency. Some of the spike frequency changes, there are specific -- the pattern of EEG in these patients is very unique, migrating focal seizures with a frontal dominance. You might be able to use EEG measures in our target engagement and efficacy, and these are things that we're thinking about and people -- we work with people like Dan and Jackie, and Bernard's team is thinking deeply around how we would measure efficacy, I mean, in these patients. And really, and this is probably for me, was an amazing moment in my scientific career. I'm thinking, we always worry, are we just helping the seizures. And that's a big -- it's important. And Dan will tell us and Jackie French also will tell us, you have to stop the seizures. Yes. But there are other things that are going on in these patients' lives. These mice, untreated, can't build a nest. Now it seems like a minor thing. You give them -- basically, you give them a tissue into their home cage and they build a little bird's nest. And you can see, on the right, that's what they build, typical mouse, and they sit in that for their comfort. And that's a natural -- and that's a big complex behavior. It requires cognitive and motor skills, it requires the ability to draw on their instincts because this is an instinctual behavior, the mice aren't taught to do this. The animals with the mutation on the left can't do it. They just can't build. And you know what's remarkable, we give the drug and they acquire this ability within 2 days. They're now building nests, and that's just remarkable. And you take the drug off and they can't build a nest. That ability to change its complex behavior that control -- that incorporates cognition, in these time frames, I've never seen that before, where you can turn off and on such a massive cognitive effect so quickly with a single drug that interacts with a single gene. So we think -- what does this mean collectively? That with -- in these patients, because we're going for root cause, not symptomatic relief but doing root cause, we think the patients are going to benefit across a number of different domains. Praxis 080, it's a the molecule, again, gives rise to this disorder called girls clustering epilepsy. It was a disorder seen in females. And when you see that, you often think X-linked. Females have got 2 X chromosomes, males are XY, X-linked disorder. Quite a complicated and interesting mechanism of the disease that Praxis has utilized its detailed knowledge of genetics, of cell biology and of clinical observation in order to come up with what we would think is not an obvious strategy for dealing with these patients. And the disorder is caused by the fact that when you got a mutant in girls in any given cell, you've got 2 X chromosomes. One of them is randomly turned off. It's called X inactivation. You don't have -- you never express both. In males, we've only got 1 X, it's always on. And what happens, of course, if you've got a mutation on one of those X chromosomes, this cell will have the mutant, this will have the wild type. The cell will be wild type, be mutant, wild type, mutant or mixed. And that mixture of cells that are expressing different forms of PCDH19 is the reason behind the pathology. We know males that have got the mutation, because they're carriers, don't ever phenotype. All their PCDH19 is mutant, they don't have a problem. It's only when you got mixtures of wild type and mutant PCDH19 do you get a disorder. But what's really important and what Praxis observed and acted on is that when you look clinically, there are some males that have no PCDH19. It's completely gone. They're carriers because they've got a null gene. They can pass that null onto the female and the female that's even got 1 copy without a PCDH and one with still has the disorder. So any mixture is bad and pathological. If you've got none, you're better off. So Praxis' sort of proprietary approach here is to reduce all the PCDH19 in the females, take it down to 0 because that's far better than having a mixture. We know that people survive well when you remove that. You can do that in a rodent. The males have it naturally, no problems. We think it's a really important, and it's just -- I think it's a great example of how Praxis, informed by genetics, it's not just something we say, it's something we live and breathe on a daily basis. It's something that's ingrained into our DNA, no pun intended. Well, we've made progress on this with an ASO that's going to remove this PCDH19. And this is just on the left, it's the amount of knockdown of the ASOs that we've been screening so far, greater than 90% reduction already. This program began about a year ago. And we're working hard now to turn this into clinical candidates, and we hope to have something for 2023. The final project will be SYNGAP1 loss of function. SYNGAP1 is another -- of the rare and of the ones we've picked, these are the more common amongst the rare, if you want. And then there's about 1,600 of these patients diagnosed in the U.S. and these kids have intellectual problems, very severe behavioral problems and drop seizures. And we're working with a few families in Australia of this. Again, like all of these conditions, huge impact on the -- completely dominates the families while looking after their 2 children with this disorder. It's a really important protein in the synapse. I've talked before about ion channels. Protocadherin is a channel that -- is a protein that's important for how neurons connect to one another. This is an important protein for how neurons talk to one another. Now by removing that blue blob of SYNGAP1, by reducing that to 50%, that whole structure doesn't work as well. That's the receiving side of the synapse. The chemical transmitter is released from the neuron. This is the side that receives it. The lack of PCDH19 reduces a whole bunch of signaling with the synapse. You can't talk -- you're not talking in the same language anymore from one neuron to the other. Again, the simple idea, can we increase the amount of SYNGAP1, and when do you increase it? Luckily, a close colleague of mine, Gavin Rumbaugh has been studying this. He's at the Scripps in Florida. He asked the question, well, if we turn the gene on later in life in a mouse, so the mouse has had the gene off, it's sick, it's got a phenotype, can you turn it on later in life and restore its behavior? Yes, you can. SYNGAP goes 1, the EEG goes, issues are reduced, the seizures are reduced, performance of memory is improved. That tells -- and that's a -- it seems like a small thing, but it's a really important thing because we don't know, the clinicians, the basic regimen, don't really know what's reversible and what's not reversible. And this is telling. And what we're seeing, we saw that with the KCNT1. We saw that the acute impact is what's causing that. It's not some chronic long-term change in the brain as the function of that gene now, here and now that's causing the deficit. So many of the comorbidities could be reversible even later in life. And that's a really important thing, important information to have. We'll test this and understand this more clinically but it's a really important thing to know. So we're not up against this. It's not as if, well, there's nothing you can do. Kid's had his seizures for 6 months, that's the end of the story. And that's been the mantra in the field because we're conservative. We don't know all we could, but I think the data's starting to come out in a very positive way. Again, we've used an ASO again in this program that increases the amount of SynGAP expression. Different mode, I won't go into the details of how that's done today, but we can get almost a threefold increase with our best ASOs, more than enough. We only want a doubling. We don't need a threefold increase. So more than enough to be able to treat. So this is the sort of ASO, again, where we're slated for delivery in 2023 for -- to move into clinical development, the sort of program that could just transform the lives of these patients. Now -- and I'll give one last section of our thinking, this is that final area of focus on implicated genes in common disorders. Now these common disorders can be epilepsy. These common disorders can be major depressive disorders. These common disorders could be movement disorders. How do we use our knowledge of the genetics to make the best decisions to enhance our probability of success, to get the best molecules that treat these patients across many domains of their disease as we can and do it effectively and as Marcio keeps telling me, do it time and time again? So these are the 2 programs I mentioned before, 114 and 944. We call those imputed. The genetic evidence in common diseases wasn't available at the time we were making these decisions. We imputed it from rare cases, we imputed it from our deep knowledge of cell biology. But now what's happening is that we are starting to see these -- they're called GWASs, the genome-wide association studies. And these are things where you take a lot of patients. They used to take 1,000 patients and you say, well, for a patient with epilepsy, what have they got that's different from a patient that doesn't have epilepsy in their genome? And it was quickly apparent, and I was involved in these 20 years ago, you can't do it with 1,000. And what's happening now though is they're doing it with 15,000 and 20,000 and 50,000 patients. Global consortia are being built. And every few months, there's another massive study that comes out, some of the genes -- and what's happening, of course, as the studies get bigger and more powerful, more and more of the truth emerges. Initially, there's false positives, false negatives because they're not powered properly. But things are starting to be powered properly now. And that's what we're reacting to. We're in the midst of those, and what I will do now, I will talk about how 114, what some of the evidence we used through that imputed mechanism. As I told you, it's in development for MDD, PTSD, I'm sure many of you have heard about Praxis' efforts in those areas, but totally inspired by a role is a study that I was part of in 2004, 18 years ago, showing for the first time that a mutation in the target for 114, which is a GABA delta receptor, when it's got loss of function mutations, causes generalized epilepsy, direct link. And as I mentioned before, there were a whole bunch of other studies done around catamenial epilepsy, premenstrual dysphoric disorder, all of it tied into the functions. So we've got direct links to behavior and epilepsy and neural excitability all driven by this receptor, a convergence, a perfect storm of information on which to make decisions that we think are reduced risk. What does it do? Synapses have got inhibitors -- there's 2 major classes of function of neurons in the brain. There's the things that excite the brains and the things that inhibit them, the accelerator and the brake, if you want. That's how you achieve control in pretty much any system, the balance between those 2 things. Now the inhibitory system has got 2 broad modes. It either pulses the breaks off and on, which is called phasic inhibition, or it puts the handbrake on and leaves it, that's tonic inhibition. And even though the handbrake, as you know, isn't as strong as a normal break, if you keep it on a little bit all the time, it's got way more effect on your car than an occasional pulsing of the normal brake. That's tonic inhibition. Massive effect on the global function of the brain. And that's what's interesting about tonic inhibition. Our brains are bathed -- GABA is the inhibitory molecule that comes out of the inhibitory neurons. It spills over and it bathes the entire brain in a low level of GABA. It's around a micromole or so. That low level of GABA interacts with these delta receptors and sets a baseline level of inhibition across the whole brain. And that's really important for setting behavior and depression, for setting sensitivity to seizures and all sorts of things. That's what we're targeting with 114. We have predominantly -- you can see here, we have a tenfold preference for that receptor versus the brake, the phasic receptor. That split is really important to these sorts of disorders. It's a -- we talk about depression, these being global imbalances in neurochemistry, et cetera. And it's the global nature of this target. And our ability -- as we know, biology is local, but -- most biology, but pharmacology is global. You get to the whole brain. This is a really good target because those 2 things come together. Well, we think about that, can we use functional selectivity, can we use gene selectivity to do precisely what we need to do to affect the disorder, limit the side effects and get the efficacy we need out of the system. And sort of unsurprising when you think about the history, one of the things we did was just asked the question, is 114 a good seizure medicine? Yes, it is. This is showing, in a chemically induced seizure, that 114 stops the seizures cold. So we know that it's -- neurobiologically, it's -- obviously, we're getting excellent target engagement. It's doing the sorts of things we'd expect it to do. Positioning from a genetics perspective, the links to behavior, excitability and epilepsy, all of these things compel us to build a case for depression. And the final area I'll touch on, which is the future, and I won't go into this too much because we don't have anything to deliver today on this other than our acknowledgment, I think this is sort of Phase II for Praxis. Phase I was the rare diseases, the genetics and the understanding of rare disorders through monogenic epilepsies. Now -- and as we'll hear from our esteemed colleagues, we heard from Dr. French and we'll hear from Dan Friedman, there's also a lot of people walking out there, a huge percentage of the patients with epilepsy that need better medicines. But we don't really know what's causative, and that's why we've been making symptomatic drugs. The genes are coming out. The GWASs are telling us, look, these 30 genes, I mean, everyone who's got epilepsy and our idea, and you can see them, and it's probably at the place to be doing it, these are called Manhattan plots because of the little skyscraper look. Those vertical excursions are the sort of statistical signal associated whether a gene really is meeting the criteria to be causative in common epilepsy. And you can see, the more things that come across the top, and you can see a sampling of some of those genes on the right. And it's the accumulation of these genes and these natural variations -- all of our genes are different. We've all got natural variations in our genes, but it's this particular set of accumulation of variance that's giving you the risk for epilepsy. So our philosophy and our thinking is, well, what do we do about it? We know that now. We now have the ability to think about root cause and is root cause dealing with 1, 2, 3 of these risk factors. A genetic mutation can knock a gene down or up by 5% to be risk. A drug can do 80%, 100%. So can you overcompensate a single gene and have a massive effect on the seizure propensity in those patients? These are the questions we're asking. These are the framework we're laying down for future decision-making to develop better drugs with higher probability of success. It's been done in cancer. Oncology has been -- has known -- oncology is or was 20 years ago where epilepsy genetics is today, deep knowledge of genotype and phenotype in tumor cells, deep knowledge and what -- and this is just mapping the activity, the commercial activity in programs that are looking at comparing what happens when you have an antibody, a bispecific versus a monospecific therapy that attacks 1 of the genetic markers versus 2 of the genetic markets. You see accumulated benefit as you tackle more and more of these things. Very similar concept and driver that we want to do in epilepsy, how do we tackle that? These are the things that Praxis is asking. Some of the things are proprietary, our thinking, but we'd like to share more in the future around this concept. I'll leave you here with this and I'll pass the podium to Dr. Friedman. Thank you.

Thank you for having me speak today. And I just want to sort of talk here about my perspective as both somebody who's a clinical researcher, studying epilepsy therapeutics and epilepsy outcomes as well as a practicing epileptologist. I treat teenagers and adults in my clinic. And I sort of face day-to-day decisions that run into the limitations of our current therapeutics in my practice. And I just want to go through some of those, echoing a lot of the things that Dr. French said in her talk, but really talking about how it may impact individual patients and go through that. So these are my disclosures. I also work for The Epilepsy Study Consortium and I share the passion of Dr. French for helping new therapies come to market. So I'll start out with a case of a patient I might see in my office. And this is a 28-year-old woman, has a history of depression and presents to my practice after an ER visit for a witnessed convulsive seizure. The evaluation of the ER was pretty unremarkable. She doesn't have a tumor or hemorrhage or anything that's emergent. And upon careful history taking, she's had occasional episodes of a hearing of a buzzing in her ear, then feeling confused that she just kind of blew off as panic attacks. We order an MRI and an EEG as sort of standard diagnostic workup for a new-onset epilepsy. And the MRI shows a lesion in the left temporal lobe and Heschl's gyrus that looks like focal cortical dysplasia, a developmental abnormality that -- in neuronal migration and organization. And an EEG that shows epileptiform discharges, spikes in that same region in the left temporal lobe. And given this information, she falls into one of those 30% of patients that Dr. French mentioned that has an identifiable neurologic or structural cause for her epilepsy. The epilepsy is focal, the seizures start in one place. And she is diagnosed with having epilepsy, which is just an underlying predisposition to have recurrent, unprovoked seizures. And the next step is symptomatic treatment. That's what we have in our armamentarium right now, and we want her -- to prevent her for having more seizures. And why do we want to do this? We want to reduce her risk of mortality from seizures. People with epilepsy have a significantly increased mortality rate over the general population. Much of that mortality is specifically related to the consequences of seizures. People can die from accidents, drownings and sudden unexpected death in epilepsy. We also want to reduce the sort of day-to-day morbidity from seizures. People, even if the seizures are not fatal, people can have fractures, shoulder dislocation, burns from seizures. But ongoing seizures can also lead to long-term cognitive changes and neuropsychiatric changes over time. And we want to improve overall quality of life. We want her to be able to work, to be able to safely go about in the world, including driving and other activities that could be risky if you were to suddenly lose awareness. So how do we pick from our symptomatic treatments? As Dr. French already mentioned, we have a lot of options to choose from, and this is sort of a time line of the introduction, the marketing of the various anti-seizure medications. You see a big inflection in the 1940s, where Merritt and Putnam introduced, probably which -- I think is probably the most successful predictive animal model in CNS disorders, which is the electrocuted cat that Steve mentioned. But that allowed for the identification of many, many compounds, many of which came to market in that time period, and then more so as our -- as we got a little more sophisticated about selecting compounds. But as Dr. French mentioned, despite the fact that we have many of these options to choose from over time, the overall efficacy on the group level hasn't really changed that much with the -- increasing the number of options. We have a lot of options, none of them are 100% perfect, but they do differ from each other in a lot of different ways and mostly in ways of tolerability, off-target side effects, idiosyncratic side effects and effects on comorbid conditions. So I'm going to go through sort of a decision-making process that I've faced day-to-day with patients with newly diagnosed epilepsy on how do I select which is the right symptomatic medication for them. So we'll go back to our 28-year-old woman. She has depression and focal epilepsy. We know that some of our medications actually can exacerbate mood disorders. Probably the one that is most common to exacerbate mood in general practice is levetiracetam, but there are many others that can make mood symptoms worse. So we want to cross those off our list in initial choice of therapy. Some of our medications don't work as well for focal onset epilepsy as they do for generalized epilepsies. Those cross off our list and we're left, in this patient, with these 8 choices. But she's also an oral contraceptive, and she is concerned about weight gain. And so we eliminate the medications that have negative interactions with oral contraceptives, reducing their efficacy. And now we stick to medications that are also weight-neutral. And she also expresses desire to have children in the near future. And now we eliminate lacosamide, which doesn't have a great amount of data for its safety in pregnancy. And we're stuck with really 1 drug to choose from for this woman. So even on an individual patient, there's a lot of room for improvement, giving her more choices. So I want to show this somewhat complicated flow diagram that shows all the ways that epilepsy and the epilepsies could affect what we really care about when we're treating patients, which is quality of life. And our target of therapy right now is seizures, but you could see that there are many ways around the seizure box to get to a negative or positive impacts on quality of life as well. But let's start with seizures. So as Dr. French mentioned in her talk, seizure freedom is what patients care about. They want to be on a medication that will make them seizure-free, and that is the biggest driver of quality of life. This is an example from the pooled analysis of a randomized controlled trial of drugs for focal epilepsy, where the patients that showed any significant -- clinically significant improvement in quality of life measures from baseline were only the ones that became seizure-free in the study. And there was really no difference if you had no effect of the medication or it made you 75% seizure-free. We want seizure freedom. And this is important for drug-resistant epilepsy, but it's also important for initial therapy. When I prescribe -- when I choose a medication for a patient, I want to give them the greatest guarantee upfront that is most likely to work long term and not tell them, it's about a 50% chance that you won't have another seizure. And in terms of where we are in our landscape, most -- in the drug-resistant focal epilepsy population, most drugs, if you look at their clinical trials, look at the placebo-controlled phase, maybe about anywhere from 1% to 5% of patients became seizure-free in that blinded phase. And the only real big difference that Dr. French mentioned was the recent example of cenobamate, where at the highest dose, about 20% of patients became seizure-free in that blind phase. And that has really driven a lot of enthusiasm for cenobamate in the epilepsy community. So where are there additional -- where is there additional room for improvement? The treatments themselves as, again, Dr. French talked about. Tolerability is a huge problem, a significant burden for patients with epilepsy. They are a big driver of negative quality of life, a big driver of -- second only to -- memory issue is a big driver of medication nonadherence. And some side effects we -- are unpredictable, that we can't predict by mechanism of the drug, but some are. And some are related to their mechanism of action, as Dr. French talked about and Steve talked about. We know that based on their targets in the CNS, what kind of side effects -- the negative side effects we might experience, especially in a dose-dependent manner. Fatigue, vertigo, ataxia, sedation, cognitive changes, tremor, mood changes are common side effects of many of our anti-seizure medications because they engage with targets that -- in circuit is not involved in the epilepsy are involved in these sorts of processes. Another potential room for improvement in this example in this patient is teratogenicity. So a lot of the medications that we prescribe, we don't know their safety in the developing brain and their safety in children born to women taking them during their pregnancy. We have some data for older medications. And if you look on the side of the chart, levetiracetam, lamotrigine and oxcarbazepine are really the ones where we have the best data for safety in women of child -- in children born to women of childbearing age. And older medicines are less safe and then there's a whole slew of medications in the middle where we don't know about. And it's important also when we consider the treatment of women who may become pregnant, but it's also important because we give a lot of these medications to infants and children whose brains are still developing. So neurodevelopmental outcomes are very much understudied in our treatments. Another area where there's room for improvement in our therapies is the comorbidities associated with epilepsy. And as Dr. French mentioned, the most common comorbidities that patients with epilepsy experience are mood-related, depression, anxiety and memory and cognitive disturbances. And as patients are -- as you move into more drug-resistant populations, the prevalence of those comorbidities are -- become more common. And the old thinking was, well, this is just the effect of ongoing seizures on the brain or the effects of our drugs that we already said have a lot of off-target CNS side effects. But we've come to understand that, that's not the whole picture. And there are plenty of examples of a bidirectional predisposition, let's say, between depression and epilepsy, memory issues in older age and epilepsy. And there may be shared risk factors because the same neuronal circuits that are not working and giving rise to the seizures may also be involved in regulating mood or memory. And so the underlying disorders that give rise to epilepsy may also give rise to the comorbidities. And that's another opportunity to target because you could see on that comorbidity box, there are a lot of ways to get to a negative quality of life even if seizures are well-controlled. And then finally, as Dr. French alluded -- mentioned in her talk, there's a lot of opportunity, theoretical at this point, to treat the epilepsy, the underlying condition that gives rise to both seizures and the comorbidity. And all of our therapies are symptomatic, they don't address the underlying mechanisms and they need to be taken chronically. As she mentioned, we don't have the eraser that will remove that epileptogenic lesion, which by the way, probably is genetic in etiology, not germline genetics, but a somatic mutation that gives rise to abnormally develop neurons in just one brain area. So no treatments alter the underlying mechanism that leads to the increased seizure susceptibility. We don't have treatments yet that can prevent the development of epilepsy, the acquired epilepsies that happen after a high-risk injury like a stroke or a traumatic brain injury. And we don't have therapies that will make pharmaco-resistant epilepsy now pharmaco-sensitive, maybe take the patient who has to take 3 drugs to control their seizures only need to take one drug at a well-tolerated dose. So there's a lot of promise in identifying new targets, better targets for therapy, including improved efficacy, including disease modification, including modification of comorbidities; at the same time, improving tolerability, limiting off-target effects and having predictable neurodevelopmental outcomes. And so despite the fact that we have 18 marketed drugs for the common epilepsies, as I sort of tried to show you, options fall short when we're faced with an individual patient that brings their own issues, their own comorbidities, their own unique life circumstances to the picture. And these shortcomings present a lot of opportunities for differentiation of new therapies. So thank you for your attention, and I will hand it off to Bernard.

Thanks a lot, Dan. So we had great talks from Steve about how we leverage genetics and use translational tools, and from Jackie and Dan about the clinical need at kind of the patient landscape level, the cohort level and the overall ecosystem. But at a -- from Dan, at a very specific patient level, how you think about that and how that drives selection of a particular drug. What I want to do now is bring all that information together and show you how we integrate that into efficient drug development programs, which we expect to have 3 for epilepsy in clinic this year. So it's a really great time to be talking about it. And of course, that is why we're talking about it now. So our PRAX-562 program is currently in clinic in healthy volunteer studies, and is progressing really nicely along the lines of that very wide therapeutic window that we've described and Steve talked about mechanistically. The kind of cousin molecule, PRAX-628, we expect to be in clinic by the end of the year in healthy volunteer studies. We'll be developing that for focal epilepsies. And then PRAX-222, which is one I'll start off with, this is -- Marcio alluded to it at the outset, this is our seventh IND. I'll start off talking about this one because it really, in an incredible way, exemplifies a lot of the principles that we've been talking about. So we're very excited about this program. Let's talk a little bit more about 222. Steve gave you the background, but as a reminder, it's an antisense oligonucleotide. It's designed to knock down SCN2A transcript and therefore, lead to reduced protein and reduce excitability, reduce seizures in children who have gain of function mutations in that gene. So it will be an intrathecally-administrated ASO. And this is a true, just like Dan was talking about, this is the need: true causal therapies directed at the underlying mechanism. Now I want to next turn to one of our key pillars, understanding the patient population. And we went about this in a different way. So I want to talk about this in some depth. We debated quite a bit, should we do like a big kind of old-fashioned natural history study that would take us years to do, and we decided not to do that. We went instead with this kind of real-world evidence project, which I think has turned to be very successful, done it in 2A, I'll show you data in 8A, and this is reproducible in fast. And I think it'll be a key part of how we approach these DEE programs going forward. But what we do is incorporate real-world data, medical record data, not just text information, that we're able to bring in EEG data and other kinds of complex, kind of very data-heavy measures that we're interested in. And of course, I do want to thank the SCN2A patient community, the 8A community. We couldn't do this without their participation, without them signing up so that we could have access to this incredible wealth of data. So in this particular study, 2A involved 45 patients, we had 45 families. We're showing you here 15 patients, average age of about 7.5. I'm showing the seizure history. It's not just -- see, in gray is the seizures, and these are just the reported ones. In green are episodes of status. Just incredible. What's remarkable about this presentation is how dramatic it is. On average, these children presented with seizures, I think, on day 5 of life. At -- about 80% of them at some point were having more than 10 seizures a day. So it's an incredibly dramatic clinical presentation, difficult to manage from the outset. And what you can see, take home from this slide is that these seizures don't go away. So these are real world data. I mean these children were optimally treated according to the best clinical drugs we have available. Those seizures and episodes of status don't go away. Of course, seizures, as Steve described, are only one part of the picture here, a very important part, but there's profound developmental delay. And that involves different domains, straight up motor delay, so these children can't hold their heads up, they can't sit up, they can't feed; cognitive delay, which manifests in delayed, really absent language in social delayed, the ability to interact in meaningful ways with family or caregivers. So it's a very broad syndrome. It's hard to encapsulate in a couple of slides, but this one gives you a sense of the overall kind of clinical and medical burden for the patients and for the entire family. So in the first year of life, and you can imagine if seizures onset within the first few days, in the first year of life of these children were on almost 15 different medications. Most of those were seizure medications, about 9 of them. They spend nearly 70 days out of their first year of life in the hospital and then 100 days in total over those, on average, 7 to 8 years, lots of procedures, recurrent hospitalizations. So you get the picture. This is a child and a family is in and out of the hospital due to refractory seizures, taking lots of medications and having lots of side effects. So it's somewhat paradoxical. You talk to neurologists and some will say they're satisfied with the standard of care, that they're working well and there are sodium channel blocking drugs out there. Clearly, they're not doing the job. There's no question about that there's a need for mechanism-directed targeted therapeutics. With that, I've got to transition to biomarker translational tool. So Steve went over this really nicely. Between seizures, even in the animal models, there is a lot of residual hyperexcitability, and that shows up as interictal, between seizures, epileptiform discharges, so kind of these epileptiform activity, even though you're not having clinical seizures. So our team really nicely brought in these EEG data from that real-world study and looked at this -- this is over years, what we were able to show was calculate what's called a, like a epileptiform discharge burden scale, so how many of these discharges they are having and how long they last for. What you can see in that histogram there is normal children, age matched, don't have these. There's some very low underlying rate success in that some kids may have a predisposition for epilepsy there, but children with 2A mutations continue to have these at a very high level throughout their lives. So that 2 key implications, 2 key take-homes here. One is even between seizures, the brain is hyperexcitable. And so it comes back to -- I think both Jackie and Dan mentioned it, having drugs with long half-lives so you don't breakthrough seizures -- have breakthrough seizures is real important. And two, since this is persistent, You can use this as a pharmacodynamic biomarker to understand if your drug's having an effect without having to wait to count seizures. There's plenty of seizures to count, but this may be a -- really a more sensitive continuous measure. So we're actually doing an observational study to expand upon these data and use it as a lead into our clinical trial. With that, I'll turn to the clinical trials. So Marcio mentioned at the outset, and it was in our press release this morning, that we've submitted this IND. It was really, really interesting. This has never happened to me and I have been doing this for a while. At our pre-IND meeting, the FDA said, be more innovative, go faster and do a seamless trial such that every patient has the opportunity to contribute efficacy data and has the opportunity to benefit. So that's what we did. We submitted it is a complex study, and I'll complement our team because it integrates all phases of development, 1, 2, 3, into a single trial. So this is a single trial path to registration. We're sure the agency will have comments, but we're waiting to hear what they are, and they've been very engaged, of course. So what does it look like? So the children will be children with, again, function mutations up to the age of 18, having continued seizures, which as you saw, that's most of these children. We'll look at endpoints that cover all those different domains: seizures, we'll look at EEG parameters, we'll look at developmental outcomes and social outcomes. And the trial, as I alluded to, is designed in 3 parts where there is a dose escalation, and we get an initial understanding of safety tolerability and the impact on seizures. That's followed by a confirmatory phase where we increase the number of patients and get more precise understanding of safety and nail down the efficacy. And then there's a long-term extension where we'll see about developmental outcomes, and we'll also see if we can expand the interval between doses to make it as convenient as possible once we're in the commercial and clinical setting. So with that, I'll turn to talking about dose in a little bit more detail. And this is something, Steve mentioned this a few times, really focused. This is how you get it right. So this is not a neurodegenerative disease, right? This is a -- fundamentally a channelopathy. The neuroanatomy is intact broadly and at a cellular level. So we're not worried about like some of the other problems that like recent ASO programs have run into, a neurodegenerative cascade and you're intervening too late. If we get the drug to the right place, at the right exposures, there's a very high probability that they should work. And that's true for most of the programs, all the programs that we talked about. The other key thing here is that we know we can get very, very high levels of knockdown, like 90% plus in the cortex and other key areas in the brain, but we don't need to get that much knockdown. So Steve showed you work showing 80%, 50% knockdown were efficacious, extended survival. But we've extended that, and we've shown a 35% knockdown will continue to get efficacy. So that's a good place for us to start. We know we can achieve that. We've been very rigorous in how we've modeled this, so not just relying on like plasma concentration. Our preclinical work and our tox studies have involved monkeys with intrathecal dosing.. We use tissue levels of the ASOs as well as, of course, message for knockdown to generate this modeling. And this just gives you an example of we can dose escalate. You see over the first 6 months or so, get to a target of 35%, it's actually pretty tight in terms of the variability for an ASO, and then maintain children there with less frequent dosing on about quarterly intervals. So very, very for us again, we think this is the key and really different than, say, a Huntington's disease program where you're dealing with degenerating brain and you need to get the ASO, particularly into deep targets versus cortex. So where are we with this program overall? We're enrolling that observational study focused on EEG, using it as a run-in to understand those interictal discharges, and as soon as we hear back from the agency, we'll be initiating the seamless clinical trial. I'll turn now to the PRAX-562 program. We talked about this one before. Steve grounded us in the mechanism here, the relatively high potency for persistent current relative to peak current as well as the use-dependent features such that really only blocking sodium channels in hyperexcitable neurons. That's what drives what we believe will be a really remarkable therapeutic window. And we're already seeing that. I'll substantiate that in the next couple of slides. The other nice thing about 562 is it does happen to have a very long half-life. And so we don't need -- all we need to do to give a single dose, say 90 milligrams, it accumulates and you end up with very low peak to trough. So for children who are having a lot of recurrent seizures and underlying hyperexcitability, this is actually a really nice feature to prevent those breakthroughs if they miss a dose. So where are we with this program? So we previously did a Phase I study SAD, a 2-week MAD and food effect. And we've reported those. We had very nice tolerability. We well exceeded our target exposures over 2 weeks. We showed on a biomarker of excitability, auditory steady state that we had nice dose-related improvements there, reductions. But the half-life, as we just discussed, turned out to be 5 days. So we decided to do a second study that was longer to get us to steady state, a 28-day study, and to also expand upon the ASSR work and understand it over time. So we'll be wrapping that up over the next couple of months. I'd say it's going very well, and I'll show you some really encouraging data about the therapeutic window. But we anticipate in the second half of the year that we'll take 562 into proof of concept studies in children with 2A and 8A gain of function mutation. That's the targeted mechanistic genetic approach. But also, as Steve mentioned, there are nodes of convergence in shared pathophysiology with focal epilepsies. And so the other cohort in that study will be children with refractory focal seizures due to tuberous sclerosis. So with 2 ways of going, and I'll show both the targeted kind of genetic medicine approach, but also our approach translating that into broader focal seizures and other kinds of epilepsy. So coming back to this TI issue. On the left here, there's a slide showing data from just day 1 of dosing. This is from our first healthy volunteer study. And day 1, at 90 and 120 milligrams, we exceeded the EC50 in the MES model, which is a commonly-used epilepsy model. Now on the right, we have preliminary data, only a dozen subjects, that's quite a few, but preliminary data from 28 days of dosing. And I've never seen anything like this. As Marcio knows, I'm a little obsessed with hitting the maximum tolerated dose. It just gives me comfort that there's an end out there somewhere. We haven't hit that here. So you see the EC50, we're reaching concentrations that are twentyfold above that without dose-limiting toxicity. I don't know if there's any other epilepsy drugs that can do that. And I'll say, too, we're not seeing, in healthy volunteers, kind of on-target sodium channel blocking [ i.e. ] as you expect to see vertigo, nystagmus, ataxia. So this is really consistent. Because we have the ASSR signal, we know we're getting drug in the brain. So this just fits this -- a wide therapeutic window. I'll turn to the other patient population, children with SCA8 mutations, same real-world data, and the visual, I think, says everything. These children have the same picture as with 2A mutations, incredible recurrent seizures, high rates of recurrent status. It's kind of mind-blowing, speaks for itself, and then the same picture about health care utilization, burden on the family, a child in the health care system. So this is the second genetic population that we'll be studying with 562. So again, this study is expected to start in the second half of the year. And I'll turn to our third program that we'll have in clinic for the epilepsies, 628. So again, this is related molecule to 562, very similar mechanistically, preclinically, a very similar therapeutic window due to that potency of persisting current and use-dependent features here. But we designed this to have a different metabolic path and a shorter half-life, making it more suitable for a broadly-used drug in focal epilepsy. It is predicted to have a 36-hour half-life, which is still long. We think it will be helpful if people do skip a dose, but it's not a 5-day half-life. And so this drug is currently in IND-enabling talks. We expect to be in healthy volunteer by the end of the year and then move into focal epilepsy from there. I'm going to come back to this slide and just -- Marcio said we definitely don't like to brag or congratulate ourselves, but with our seventh IND in the works, 3 epilepsy programs slated to be in clinic, the hypotheses here are panning out, which is really encouraging for us. The biology is panning out because it's true and it's real. And so -- and we're excited to have the next 2 candidate nominations. With that, I'll turn it back over to Marcio to close.

I guess there's not a lot more to say. After all this presentation, I just want you -- to thank everyone before the Q&A. And just remind, maybe to wrap, right, this is all very deliberate from a science perspective, from a business perspective, from a regulatory perspective. I know we mentioned a number of INDs with interactions with the agencies, how deliberate we bringing [ ovative ] patient data. We know there's a directive for the FDA to listen to patients to understand the impact in their lives, to totally use the statutes on how they feel, how they survive, how they function. They are like the [indiscernible] system that was shown today, 10 years, all medical records. That's what Bernard showed today. It was not like, oh, give me a couple of records here and there. We collected every single medical records in a 10-year retrospective manner, used machine learning to extract the data and put it back to them. By the way, we're not selecting when we report this. We're just showing when we report this. It's devastating, but it breaks this idea that, oh, if I haven't seen because I'm the [indiscernible] at NYU or Columbia, it didn't happen, nothing against us. It's just because they had to go to another hospital and we were able to get the data, we're able to report this. The path for epilepsies is way more straightforward than for any other disease from a regulatory perspective. That is the added bonus here to some extent because we'll be able to, for example, work with the FDA Green using single trials for its direction. That's the aim, is to get there, not compromise. We'll not compromise safety, not compromising efficacy, but really developing the best possible drugs for its patients. I think they made this absolutely clear. Now with all of that comes financial discipline. I'm not going to repeat it again. I know it's top of mind for everyone. It's something I wanted to be very clear. We care deeply every single day. At Praxis, we're going to continue to only advance the best drugs so we can generate the best return for our patients, for everyone else. So just want to thank all of you. Most importantly, there are hundreds of patients that gave us the data that we're seeing today because they gave theirselves for a seed to grow or their mutations for like samples to be cloned or data for the natural history observational or even hundreds of patients in our clinical trials so far. They are the actual heroes. We are just a little bit of puppets here trying to make this all work, and we will because that's really matters at the end. So we're going to break for a few minutes, 5 or so, kind of rearrange here, get the Q&A going. Thank you so much for everything so far, and see you in a couple of minutes. [Break]

All right. I think we are ready to roll the Q&A. So a lot already.

Laura Chico, Wedbush. On 222 and I guess the epilepsy programs in general, you have a mix of modalities, small molecules, ASOs. Could you talk a little bit about where -- I guess the rationale for what modality makes sense in which indication? And then CNS delivery of ASO specifically has been kind of a tough bar to achieve. So could you talk about what features of ASOs you're targeting, not only to enable delivery but also distribution?

Yes. It's incredibly important, Laura. And I'll ask Steve and Bernard to touch on that. I think the way we look into, is it there? Is the target amenable? And what do we expect that to do? And Bernard can talk a little bit about the distribution, we will look a lot into that, but please, why don't you start, Steve?

Yes. So I think that's a great question. It something we think a lot about at Praxis. Where is really important for a lot of these epilepsy targets, cortical exposure is really important. And we know that's one of the areas where ASOs naturally can achieve exposure levels from intrathecal administration, and particularly in some of the decisions we've made are looking at targets where cortical engagement is going to be high probability of efficacy. We think biologically, these are more tractable. As Bernard said, this isn't a genetic defect that sets off an entire complex cascade that has questionable reversibility. I think we've look at that, so we think biologically we're going to have efficacy. And we're also looking at key collaborations. [ Cerevel Therapeutics ] is one of them that has a catheter and a port system to enable human intracerebral ventricular administration. So we are exploring that as well with them for future progress so we're on top. We don't think every ASO program will require that, but where it is needed, we will have that option to tackle that head on.

Yes. I'll just add, Laura. One of the reasons I joined Praxis, because I like the idea that it wasn't a therapeutic modality platform company. It was a biology and secret sauce is what Steve was talking about. And so we use the modality that makes sense for the target to get to do whatever needs to be done at that target and to get drug there. And so I think our modalities and delivery approaches will advance along those lines. And for example, we talked a lot about we're doing genetic therapies with ASOs. We're not doing AAV gene therapy and for these ion channel targets. We don't want to. We don't think you get the right distribution. You can't -- they're not titratable. And so I actually think it could be kind of dangerous to drug ion channels with AAV gene therapy at this point, until you can regulate them better with cell-specific or regulatable promoters. So I think we're choosing our modalities and our delivery approaches very carefully to match the program.

Okay. So Tazeen Ahmad from Bank of America. To follow on the question of ASOs, so Marcio, out your last job, you had some experience looking at ASOs over various indications. So beyond which today's topic is, which is epilepsy, how do you kind of view future potential opportunities? Would you look at Huntington's, for example, or related indications? And then also to follow up on gene therapy, taking into account what you just said, Bernard, about disadvantages, there are still companies that want to pursue that route, who are pursuing that route in different types of epilepsies. Maybe just get a view from the panel on is there a type of epilepsy that you all think might be more amenable to gene therapy? And would that be an area that you would consider going into over time?

Tazeen, the answer kind of Bernard gave us is what I would start here for a previous question, right, like looking to the targets, the biology, the distribution of the brand and then ask how. And sometimes, it is small molecule and for some of them we showed today it's a small molecule. Other times, it's going to be an ASO. There might be a day that is an AV or any other type of deliverable like gene therapy, not right now, and we can explore a little bit on to why. But the hope's not a strategy has to stay true. And I know sometimes I say that jokingly, but in this case, it's not, meaning do we have to really understand what we are targeting? And as we saw by [ IC's ] presentation today, we first asked the question, does this gene matter? And some of the failures, like recently, I think the question was now resolved, right? Now for others, like you mentioned, like Huntington's then, there's repeats, length and things like that, and [ mine's ] not fully resolved. But that answer is checks, that question is checks, the structure of the brain, the kind of knockdown that you might be necessary on that one, it is still not resolved. When you go to the toxicity, I'm going to go back to epilepsy. We know what happens if you knock down a lot of SCN2A and it's not pretty. So that is a place that you have to stop. So if I come here today and say, "Oh, we can just give this in an indiscriminate way as 222, as an example, we'll be responsible. So the pharmacology, the biology, the basic like pharmacogenetics have to play a role. Is that going to be only epilepsy? Unlikely. It's very likely going to be looking to other areas in the future. But right now, epilepsy just makes the most sense. Now to go back to gene therapy on the strict sense of gene therapy or like AV driver regulation, right now it's just not possible in our view. It's a fine-tuning versus a hammer, and we need better fine tuning, but maybe that's the point.

With high channels in many of these key proteins, precisely where you express them precisely how much you get is critical for the function. And I think AAV, for many of these disorders, is too much of a hammer. And when the technology improvements that Bernard and Marcio alluded to are with us, I think Praxis will be ready to look at that seriously and move forward.

Yes. It does -- it all comes back to the points that Dan and Jackie raised as like, it's about therapeutic window, right? And we get that with small molecules. You can't get that with AAV gene therapy right now.

I'll also add that there may be types of epilepsies that are amenable to that kind of therapy. Probably the lowest hanging fruit are focal epilepsies. So as an alternative to disruptive surgery, you can use a hammer in a very surgical manner. But whether or not this is preferable to other therapies like neurostimulation that we have on the market or for stem cell therapy that may be emerging, we don't know yet.

And [indiscernible] do you think there's a therapy available [indiscernible]? Or is there room for [indiscernible]?

I mean I think there's definitely -- I think there is definitely room for improvement for both. I think fenfluramine was a big breakthrough for Dravet, but there's still plenty of patients who were on fenfluramine who are still having seizures. There hasn't been an effect on the cognitive comorbidities and the gate issues and some of the other long-term effects of Dravet syndrome in the -- we don't think you have with fenluramine, and it didn't show that great efficacy in LGS, which is a heterogeneous disorder that is often due to a lot of the monogenic disorders that we've been talking about. And lumping it as orphan condition as a single entity probably is one of the reasons we have such sort of unsatisfactory outcomes with the therapies.

Ritu Baral, Cowen. First question is on 222. Bernard, what's left to talk about with FDA on that integrated trial design? Do you know sort of like the general sizes of the different stages? And where in that do you think you could generate that first quantum of seizure reduction data at the target dose? Generally, what year?

So we've had -- we had a pre-IND meeting, and the feedback that I talked about in terms of the seamless study and the parameters, in particular, that everybody can contribute both safety and efficacy data. There are other suggestions in there, but we submitted the IND last month. And so we think here, there's been some dialogue. Optimistic, but we're still waiting to hear. But we know we'll be able to engage on whatever points they raise because they seem extremely engaged in this particular program. And then in terms of where will we be able to start to see efficacy data which showed you that dosing paradigm, right, we start randomizing people and they go into that escalation of dosing. That's the first part of the study. So our initial target is 35%. It would take about 6 months of dosing to get there on what we've laid out so far.

A quick follow-up for Dr. Friedman. As you've laid out the unmet need and as Dr. French laid out the unmet need, she talked about a small percentage of patients getting to seizure-free, but maybe 75% to 90% seizure reduction being like, okay. Is there an absolute threshold for seizure rate where you feel that even if a patient hasn't gone seizure-free, that maybe you're stopping some of the neurological damage, maybe they'll start -- especially children sort of gain function, gain milestones? Is there like a number?

I mean there's certainly not a number. I think in many of the epilepsies, once you sort of reach 50% or greater seizure reduction, you begin to see add-on benefits on cognition, social interactions with the developmental encephalopathies and other things. But -- and there's incremental benefit. And it also depends on the kinds of seizures. So if you -- and that's why sometimes looking at absolute seizure reduction is not always the right way to go because we know that convulsive seizures, tonic-clonic seizures are the most dangerous seizures, the ones associated with the greatest morbidity and mortality, drop seizures in Lennox-Gastaut seizures. And if you can eliminate those seizures, you can make a big impact in people's lives without necessarily making them 75% -- without achieving a 75% overall seizure reduction.

Okay. And last question, sorry. As you mentioned, the tolerability burden of the current options that you have, what are the things that patients complain about the most? And what are you watching these new therapies for? What are the potential additive tolerability issues that you don't want to see?

So the ones that on a day-to-day basis in clinic I hear the most are fatigue, irritability, mood changes, sleeplessness, double vision, unsteadiness. Those are the most common ones and often directly related to the mechanism of actually phase of GABA activation -- inhibition or blockade and voltage sodium channel effects.

Myles Minter from William Blair. Just on 222, have you disclosed your sort of target dose that you're looking to see with the intrathecal administration there to achieve that 35% knockdown. I'm curious whether it's up towards the 80 milligrams or above, which are -- might raise those tominersen issues that we've seen in HTT lowering.

I'll take that one. I'll take the bullets. No, just kidding, there's no bullets there. So we know where to start. And I think that's what Bernard was guiding, right? So most of the dialogue with the agents has been where we would start so far. It's fairly conservative. I would say, right now before we disclose, we do expect they're going to agree with us, but we just don't know, and I don't think we'll be responsible. Now it is, I would say, lower than what one would expect where we're going to end up, not where we're going to start, to reach 35 to 50 where we think is the sweet spot there. There is general accumulation that happens and then there's an expectation that we'll be able to move from a number of weeks between those to a much larger number later on, on the treatment as well. So lower than the number we quoted, Myles, there. But I think right now, it should still like hold a little bit what it is. But it's starting pretty low, let's put it this way.

And then maybe on 114, we saw that in the slides. Why haven't we seen a trial started in epilepsy comorbid with depression? Are you waiting for Aria to read out on that one?

It's -- no, that's a great question. We were joking about this a little bit the other day. It's discipline. It's probably the best word. It's overused a little bit as well. The -- it makes a lot of sense to use 114 in epilepsies, as you saw, and a few other things as well. But it makes even more sense in [ fear extinction ], on mood modulation, on anxiety, like major depressive disorders at this point in time. It's not the only compounds that we have on the class. You might take a look at our chart. They are on disclosure targets there that we might be looking into different things. There is a unique proposition for us that we really don't believe it exists in the market that those drugs can be combinable. So while we don't like using a number of drugs, you could imagine for a movement is orders for psychiatry, for epilepsy, that when you look into the entire patient population, we could really treat a much larger number of patients by combining them. So that's kind of half of your answer there, Myles. As soon as the positive results come in, that we expect 5 or 6 between now and the end of the year. The entire market has changed a little bit as well. Ability to have more capital available, more for the company, then we might change the strategy a little bit. Right now, we're going to stay focused, keep the burn the way it is, keep the size of the company the way it is so we can deliver the promise instead of just waste resources.

It's Doug Tsao from H.C. Wainwright. Maybe just to start, Dr. Friedman, you sort of mentioned obviously the goal is to get a patient seizure-free, and that does not often happen. Most patients are obviously on polypharmacy, taking numerous medications. How do you think about at what point do you sort of take your foot off the gas pedal and you sort of say, like, hey, this is enough, even though they're not necessarily seizure-free, but they're just sort of reaching that tolerability, and sort of how do you mix and match and go through that process of picking through the different options? Because as you noted, for many patients, the list sort of starts to get whittled down very quickly.

Yes, I mean it's an important question that I have and I talk about with patients all the time. Obviously, the initial goal of treatment is seizure freedom. And the paradigm is usually try one medication in monotherapy, try a second medication in monotherapy if that first one fails, and then proceed on to combination therapy. And at some point, I have a conversation with the patient who's still having seizures and say, well, we have to weigh the benefits of becoming seizure-free with the risks of additional therapies or side effects. And then I start focusing my strategy on, well, let's work on the most disabling or the most dangerous seizure types. And if I can make somebody not have tonic-clonic seizures and maybe have focal awareness seizures or focal impaired awareness seizures, I may not make them seizure-free and I may not get them where they want to be in terms of being able to drive or work in the jobs they want to work, but at least I reduce their risk of mortality and morbidity. And so that's -- it seems like a consolation prize for patients, but after many years of struggling with medication trials, I think people are willing to accept it.

Great. That's helpful. And then Bernard, you mentioned the hyperexcitability that you often see in between seizures. I'm just curious if you sort of have an assessment of the -- how important is that clinically? And is that something that would be sort of just a downstream benefit in effect of medication? Or is that something that you need to specifically think about?

That's the way we're presenting or think about it as kind of leading edge of an efficacy market, right? If you're dampening excitability as you see on the EEG, that we would expect to then turn into seizure reduction. But it may relate, and Daniel alluded to this, like there's shared pathophysiology that may be driving the seizures and the comorbidities. It just shows that pathophysiology is active even when seizures aren't. So I would say, we don't know what the clinical significance will be beyond seizures, but it should sure tell us that you're dampening excitability, you're on the right track and would expect to have efficacy on seizures.

I mean there is good evidence from other types of childhood epilepsies that the burden of these interictal discharges translates directly -- especially during sleep, translates directly into cognition. And reducing the spikes, what colloquially is called cleaning up the EEG, can improve learning and school performance and attention.

I think we have 2 indirect measures already of that, Doug. One, right, we're seeing today, "normal" kids or non-seizing kids don't have them. So we directly should ask like what's really going on here, right? But the second was on the case KCNC1 model that Steve showed. Those spikes are always present there. They're really not seizing. But then when you remove them, the combination, which we are calling combination and that's our building behavior, comes back. Once you get them back, remove the drug, they don't form masks either. So it's a connection here. It's indirect and hence it's not a direct evidence, but it's pretty strong indirect evidence right now. Quite importantly, and maybe we haven't, and I'll take a second to explain why we are doing this run in 3 months, observational study for the 222 prior to the 222 [ in situ ] drug who wants to very objectively measure those things. So when you go to the study, we understand on those patients instead of on the cohorts. And then we can have a conversation with the agents since they were incredibly positive on the dialogue with us. I think we hear a lot about the FDA not being positive or competent. We don't hear enough about when they are. And I want to say how pleased we are with the agency conversations we've have so far in general across the board on trying to extract as much data from every patient. And that's another way that we are doing this. And I think it's -- we're going to see, but we're bullish about it. Yes. I know you've been waiting for a while, Yas, so [indiscernible].

It's okay. It's all good. First of all, thanks for bringing us all together. I know you could have done this virtually, but it's nice to see everybody and all our colleagues together, so thank you for doing it. A couple of questions for Bernard and a few for Dr. Friedman. Bernard, I don't know if you commented on what the age range will be for the PRAX-222 study, but it would be helpful to help us provide a range or how you get to that. Second question is, is there a correlation between those [ IAD ] burden to seizure frequency and what that would be, and then whether that correlation is the same when you go from the different subtypes of epilepsies, so you could be using that for other studies as well? And then I have 1 or 2 for Dr. Friedman.

Yes, sure. So thanks, Yas. So it's a pediatric study, so we're going 2 to 18. And then we hope over time, we'll be able to lower the age range and get the children as soon as they start -- as soon as they're diagnosed, and I mentioned they have seizures within the first few days. But broad age range there, as you saw, kind of stable clinical characteristics, so that works. To your second question about is there a relationship between seizures and interictal discharges, looking at my colleagues and have we looked at that kind of thing.

Because the seizures are reported and they may not line with the EEGs, but with the observations certainly we'll be able to answer in a more structured fashion.

And then Dr. Friedman, so we as analysts, we try to always look at clinical studies when they get posted. And we try to figure out if there are clues in there that helps us identify that the sponsor is capturing a homogeneous population. So now as you now practice this approach and see [ SCNA to an A ] in focal epilepsy, so what type of patients or inclusion/exclusion criteria could one use to really create the most homogeneous population to be tested in a clinical study. So if you could just give us some maybe [ strokes ] or comments there, it could be really helpful.

I mean I think with some of these targeted therapies, obviously, having a -- the actual monogenic disorder and not only -- and perhaps not just the gene, the pathologic variant, but a functional consequence. So a monogenic disorder that is a gain of function or loss of function, when you have, is probably the most homogeneous population you can get. We don't know what the effect of genetic background is on the severity of the expression of the disease, and the SCN1A story is very important. We have people with the same gene mutation with a whole range of disease severity and phenotype, and that probably is related to the -- part related to the genetic background. But I think -- I'm not worried about capturing a homogenous population for the monogenic epilepsies. For common epilepsies, I think that may require some thought. And depending on the mechanism of the drug, it may require excluding certain folks, like maybe patients with posttraumatic epilepsy may behave differently than patients with focal cortical dysplasias because of the underlying biology of their epileptogenic process. But that remains to be seen. All of our symptomatic therapy trials have not done that yet. But as we get more precision in our mechanism more targeted, we may need to start to think about how to do that. And I think that's going to require a lot of input from the community.

This is [ Eddie ] from Guggenheim. I have a question on these DEEs, which are obviously names because they're developmental in nature. And are there aspects of the disease or the particular diseases that are sort of untreatable or undruggable because there are permanent structural or cellular changes that happen during development such that just changing the state of the channel or the level of channel isn't impactful? And sort of how are you thinking about that when designing some of these rare studies?

So I think for some of these disorders, if you've had a brain that's been seizing for 18 years, there might be changes that have occurred that are irreversible. And the field doesn't know. And [indiscernible] touched on that a little bit about the things that we did see that were reversible, and it was quite remarkable how reversible they were. Some things might -- some aspects of the disease will always be reversible. Correcting the genetic lesion is going to be beneficial because that's a continual change in the function of that brain. So I think it's going to be that we'll see over time. We know in some cases -- and it's not common that in the KCNC1 cases, some of the children get microcephaly. But that's not a common finding. And there's a whole group of children that we think that have normal brains, that goes on. So we're not seeing those signs. And in PCDH19, which is you'd think would have some structural elements to it because it's a cell-to-cell contact disorder, when you look at MRI, there are no structural changes in those brains. So I think we're not seeing in general, with DEE's, wholesale morphological changes, which is a really good sign of a irreversible developmental issue. So I think we're quietly confident that that's going to have something we're going to be able to contain, but early evidence is very positive as well.

And maybe even more important there and without giving too much on the tips on things we're working on is, there are changes that happen within the channel, like migration or density of channels on different types or place in the neuron life when you're talking about like [ 1 2, 1 6 ] for [indiscernible] or when you're talking about interims that don't get retained or can't retain throughout the development. So we touch a little bit on the surface today, but when you look into this disease, we're looking a lot deeper in terms of, is this going to work at say, 2 months old? Is this going to work on the 5-year-olds equally? And maybe that's not as explored as it should, but it's something we take incredibly seriously once we're declaring candidates.

This reversibility question, it's always a question until you have a therapeutic. But it's also one of the nice things about having a broad age range. Like we said, you do get -- start to get a sense of is there a window, do we need to go earlier, and that helps us make the case for if we need to go earlier. So I think to be determined, but the logic is as Steve laid out.

And just a follow-up on the seizure freedom conversation we've having, for Dr. Friedman. Now that there is a drug out there that sort of is pretty good about doing that with Cenobamate, is the clinical bars are changing with your colleagues that like that's becoming important when your looking at these novel therapies? And given that that's not the registrational end point for these studies, like is that something you're thinking about in trial design that like those would be important endpoints to look at because that where the field is going now that there is a drug?

I mean I think it's going to be a important endpoint of the market, right? So we're still doing placebo-controlled trials where the outcome is -- present seizure reduction for registration of common antiseizure medications. And I think most companies are including seizure -- significant amount of seizure reduction or seizure freedom as an early signal to -- whether or not to pursue later stage clinical development because I think they want to differentiate themselves from the 18 other drugs. And that right now is -- seems to be the best way to do it, better efficacy, but it could be one of the other aspects that we discussed. But I think it should be part of early clinical development, is really shoot for an improved efficacy over existing therapy. And unfortunately, you don't know that until you get that into at least Phase II trials with lots of people.

Ritu again. Dr. Friedman, as you think about the pathology of many of these epilepsies, how do you tease apart some of the clinical dysfunction and the sort of the known defects for those diseases where there are no defects, right? How much of the dysfunction is sort of upstream of the seizure and how much is downstream of the seizure? So as we think about the symptomatologies, I'm wondering like what can be decoupled from seizure reduction and what may not be?

I mean I think it's a good question, and it probably varies a lot by the DEE, and whereas some of them, there is -- the major impact of the development trajectory is the seizure burden. And if you could stop seizures early, then you may have a normal developmental trajectory. So infantile spasm is maybe one example where there's good evidence that early aggressive therapy can, in many cases, improve developmental trajectory. It's also a heterogeneous disorder. But I think for a lot of these, especially disorders of the synapse, there -- some of those comorbidities are related to other things that the synapse, so the comorbid autism and other cognitive dysfunction, it probably has less influence from the seizures.

Super helpful. And then just back to the 222 conversations with the FDA, and how you see that integrated trial design right now? You've mentioned at some point you were going to start looking at optimized dosing through more spread out dosing, but that wasn't going to be in that initial dose escalation, right, it was going to be after? Would you see that as like maybe post-registration, or is that something that you plan on having before any potential filing?

We're likely to have some of that information before filing, because some children will complete the randomized blinded period as we proposed now. We were just taking about what -- to complete that, and then they would roll into the longer term, essentially where we have less frequent dosing, what, yes.

Laura Chico, Wedbush Securities. Just one last one for Dr. Friedman. Could you talk about how a genetic screening has change in your practice? And maybe not just the pediatric setting, I think you still treat patients into early adulthood, so...

I treat [ kings and ] adults. So actually, this is the point I really wanted to make here is that these are not -- these are disorders that present in childhood. But if the children survive their epilepsy, they will become adults. And while most of our efforts for screening, so for instance, a lot of the sort of sponsored screening, subsidized genetic testing that many, many companies, including Praxis, support, and insures focuses on children. But we have a huge population of adults which we use to put in a bucket called symptomatic generalized epilepsies that have never had genetic testing. And it's still takes effort now to get it covered by insurance, but it's something that I do as part of my practice when I can. When I see patients with undiagnosed childhood onset, infantile onset epilepsies and intellectual disability, I will try to initiate a genetic workup because the same therapies that are being developed for children will one day be available for these patients if we have a molecular diagnosis for them.

For that case that you provided in that type of scenario, where you have somebody presenting in adulthood, how often are you screening those candidates?

For mostly for the common epilepsies, I don't because the yield is very low. Most of our screening looks at sort of monogenic pathologic variants, and whereas probably in the adults with common epilepsy, they have polygenic risks of more common variant -- attribute to more common variants, and that's just not part of the diagnostic [indiscernible] yet. I screen them when they have childhood onset epilepsy with intellectual disability. And I screen them if they have a family history of epilepsy. That's a unique population.

Yes, Myles from Blair again. You presented some pretty interesting data on the number of like medications that are used in these patients. It seems to be a lot, and I'm particularly interested, if you do get PRAX-222 to lower -- to 35% knockdown and you're on background like sodium channel blockers, how are you going to control for that use, like if that pushes you further down the functional knockdown, so the threshold? And especially if you start controlling for that use, what happens to the other 8 antiseizure medications that these patients are on. And I mean is that a problem that you see in the clinic, Dr. Friedman? And are you concerned about that? I'd like to hear comments from both the company and you.

I mean from a practical standpoint, we use medications with overlapping mechanisms all the time. And sometimes it works and we don't -- and they work in combination, but not alone, so -- to sodium channel blockers. So I think that -- I don't think that sort of 2 therapies targeting the same mechanism would present a problem for me in terms of interpreting the data as long as the groups are balanced in terms of overall agents used.

It could be great problem to have, right? So one way it could play out is that -- an ASO like 222 is kind of the basal therapy, you've reduced levels of the channel that you still need to modulate the channel on top of that a little bit. And so our hope would be that it could be a monotherapy, but maybe at lower levels. The way you can -- of any seizure medication, the way you run a clinical study is people come in on whatever they're in, and they're obviously going to be taking a number of different medications. Then you have 222 will be added in a blinded fashion. You'd like to keep things stable and only reduce them. And then Ritu asked about that open-label extension in the third part of the study. That would be the place where you start to ask, if they're doing really well, do they need all these toxic sodium channel blockers? So I think we've thought about this a lot. We just want to go about systematically so that Dan and colleagues have the data they need to interpret it. You don't have too many moving pieces unnecessarily.

I'll connect this back to your first question, Myles, as well, right? You could imagine without once again getting too much, that knowing that the channel is blocked, right, And we know sodium channel blockers do a pretty good job blocking the channel, maybe too good of a job and then you can't do anything else there. We just have naturally a safety valve. So as to your question about the -- is the drug in the brain, is that doing what's expected? If you hit non-toxicity, for example, you know what's happening there. Now we can start working with the baseline therapy because you know the additive therapy of 222 or any other that we might be developing is present and is doing what's expected, which other drugs didn't have the benefits, other ASOs didn't have the benefits of doing that. So we see actually as a potential feature that we can play with. And more to come on that, but safety was the #1 thing we built into this investigation by far, maybe why it's being relatively simple to dialogue and even to incorporate some of those things into the design. Okay. All right. It seems like that was it for questions. So thank you, everyone, again, here in the webcast. I appreciate the time, and more on the next Analyst Day a couple of months. Thank you.