All right, we are live. Good afternoon, everyone. My name is Yatin Suneja. I'm one of the biotech analysts here at Guggenheim. Welcome to our inaugural Healthcare Innovation Conference.
Hooray.
Our next presenting company is Praxis. We have a few executives from the company here in the audience, but I'm going to have a discussion with the President and CEO, Marcio. Marcio, why don't you maybe, there's a lot that has happened at the company in the last, let's say, 12 months only. I know you've been building it for a while, but just talk about it in five minutes. What are some of the upcoming catalysts for the company? And then we'll go into the Q&A session.
Yeah, absolutely. I appreciate being here. As you said, in the inaugural, I think there's going to be many more. I appreciate the invitation and all the success and the partnership we've been having for all the years, particularly the last 12 months. Praxis is a very, I would argue, unique situation right now, but a quite unique company as well. When you look into developing drugs for CNS, probably like several years ago, people would say, you just lost your marbles. CNS, very high bar, very difficult, very unpredictable. And things moved quite a bit. We learn more about the disease to go after, the mechanisms to go after, the networks in the brain, the things not to go, the ones with higher or lower variability, and so on.
But from the get-go, we have an idea that if you look into some fundamental networks and channels in the brain, we would be able to modulate a number of conditions. And we had our ups and downs, as you mentioned, but I'm going to talk mostly about the ups in the last 12 months, where several of the hypotheses that we laid out seem to be correct, quite positive, developing quite nicely, and generating not only a number of readouts we had already this year, but many to come. We have four drugs in the clinic right now.
I couldn't pick which one of them has the highest rate of success, which is a luxury in this business. And I don't mean that just by sitting here and to give any of you or myself lip service. It's quite true.
We can go one by one on each one of that. A number of other drugs coming behind or potential candidates coming behind to continue to deliver on this promise that at the end of the day, I know very deep in my heart, and I'm sure you would agree with me, that we're only doing this because there are patients out there that really need us to do this really well. We all like the returns, and we all like the successes, but fundamentally is what we are doing. Ulixacaltamide hydrochloride is our lead program. I'm sure we're going to talk a lot about that in terms of what is to come. Then a number of assets in epilepsy after that as well. Incredibly excited to be here today.
Thank you. I mean, I still think that drug development in CNS is still pretty difficult. I think a same asset or a similar asset could be successful in one company versus the other. I'm sure you follow what happened with AbbVie today.
Just today.
Similar asset, successful, but not here. So, I think there is a similar situation with your molecule as well.
Agreed.
Especially the T-type calcium channel. So help us orient why is your molecule differentiated, and what are sort of the next step? Because there have been some change of events or there are a lot of, I would say, nuance around the development and trial design. So if you can just simplify that for our audience.
Yeah. I think today's and the little we know, you just mentioned the muscarinic readouts and kind of a little bit unexpected, I would argue, for this space. And it's not that different than Essential Tremor in a sense. So I appreciate you bringing that up as well. When we look into drugs, I think for the most part, people tend to make binary decisions about whether the drug would be able to deliver on what is needed to patients.
And forget that around that drug, there's a number of quite key elements on execution, patient selection, how the trial is organized, even the speed of the study. So there's no cohort effects created. There's no different pools of patients and so on and so forth. We tend to spend a lot of time upfront talking about all those elements. Does that mean it's completely fail-proof? Absolutely not.
I think we still have the biology to fight against, but definitively, we're removing some of those elements. When you look into essential tremor, for example, is the largest untreated right now or primarily untreated movement disorders affecting Americans? Over seven million of us, almost certainly people in this room at this given moment, either today or in the near future, are going to manifest symptoms and then eventually be diagnosed with essential tremor. So it's very prevalent, very debilitating for patients.
So before we started this phase III program, we had to ask why. Why are we the first ones on the phase III? The molecule itself is very, very clear. T-type calcium channels are fundamental for essential tremor. CaV3.1 is, one could even argue, pathognomonic in terms of how the disease manifests in general. So why is it being so difficult to develop drugs in this space?
And then going back to the actual characterization of the patient population, the endpoint selection, the visits, everything else, and we checked all those boxes, and we decided to design two studies. They're fairly different. They come from the same general idea, but they're fairly different, but they're complementary. One to show in a more traditional way what happens when you give 12 weeks of this drug to a patient, drug or placebo, and we can estimate the overall change there on the point estimate and so on, and the other is what happens when a physician prescribes this drug, a more,
I'm going to call, natural type of treatment, so stable responder randomized withdrawal. Give the drug for eight weeks, use the responders, the responders stay on drug or get randomized to placebo, and that data is very, very complementary for the overall understanding.
The second part on the operational is choosing the endpoints. Not what people who never developed drugs are going to tell you you should use, but what we're observing, whereas the most base rates. Behavioral economics is a field now of what, 50 or 60 years, and we're still ignoring base rates, still ignoring what we know about drugs, what we know about the population when we're developing drugs. We didn't. We're looking to the population and ask what happens.
Tremor is variable, but what is not variable is the impact of tremor on patient lives. And that's why you had to, and we got a little push from the FDA. I'm going to give them a lot of credit for that, that the activities of daily living were far more important than anything else. So you pick that one. Then how frequently?
We know that about every other week is the ideal time, as short as one week if you want a little bit more definition. When you put that all together with the drug being right, because ulixacaltamide is still the right drug here, we believe that the probability of success in general increased quite a bit.
OK. With regard to the, so you're running two studies, right? One is a randomized withdrawal. One is a traditional placebo-controlled study. You recently introduced the concept of interim analysis. Recently, I mean like two, three months or four months.
Yep.
What led you to that? How much of that is informed by what competitors are seeing? And then now you're going to have that interim in Q1. What are certain scenarios if you can lay down three scenarios for us?
Yeah. So I'll go back to the concept that we are the first ones to run a phase III program in Essential Tremor. And while there's a lot that we know at this point in time, a lot of information that we collected throughout the years, there's a lot that we might not know. Maybe there are some unknown unknowns here, not to quote one of our favorite politicians of all times, but that might be a lot that we simply couldn't control.
We don't believe there are many, if any, but it would be silly to be in our position right now where demands for this study by patients being overwhelming. Last time we updated the numbers, we had over 75,000 patients that reached out to us to participate in this study. I can tell you that that number is significantly bigger today.
So there was never an issue on whether or not we would be able to recruit, which many other companies deal with. That was not the case here. So then the question we have to ask is, if we need to add a few more patients, because one of our assumptions were wrong, the variability, maybe the spacing like on some of these visits create a little bit more variability. We don't believe so, but just to illustrate, the effect size is a little bit reduced, but it's still quite meaningful for these patients. Why not? And it became very hard to justify the why not. There was another shot into making a drug that we strongly believe to be positive showed that it's positive from a regulatory perspective. So that's why we designed.
I think others in the space, reporting data incompletely as well, gave us a little bit of pause. What do they know that we don't? So we took the conservative approach, I would say, of conducting an interim analysis, which is now going to be completed in Q1, as we announced last week. Feel very confident about the procedures, about the assumptions going in, and about the potential scenarios out of the interim analysis. The way we design, there are multiple ways that many, many ways you can design an interim analysis.
But we really wanted to test whether or not, from a general likelihood, whether or not we were wrong in terms of a little bit. So by increasing the sample size, if we would really bring back to full power and restore the ability of the study to be likely positive. So that's how we designed, and that's why I believe that is still the highest probability coming out of that.
What will trigger the interim? Is it a certain percentage of patients? Is it like 50%, 60%, 70% of the patients when they have completed it that will trigger it? Why would you not increase the size, even if things are trending in the right direction?
Yeah. So the information fraction, as we call the number of patients that completed the final assessment, we gave the guidance before that we weren't going to be anywhere between 50% and 70% or so, 75% or so. Also always said that we wanted to be on the lower ends of that range because that gives the most flexibility for all of us operationally and otherwise. There is still a chance on this study, things are going to go really, really positive, and we want to take advantage of that as well.
So we haven't changed the guidance in terms of the information fraction. And what you're trying to learn, of course, there is how much effects exist. Why wouldn't we increase the sample size? I think there are different zones that are going to be given the option, and non-binding, by the way, by the IDMC.
And one of them, and the highest probability a priori here is to increase. But let's say the other one might be continue the study as is on the upper end. And I think at that point in time, we're very, very confident just by over-enrolling by a little bit would be already there. So that might be a reason not to formally increase the sample size, but to some extent informally already have increased the sample size at that point in time.
OK, so when you increase the sample size, I think you have said recently that it could be up to 200 patients.
Yep.
Let's say you go to 200, how much additional time you would need?
Yeah. The 200 is kind of our internal view right now in terms of what would still be feasible, meaningful, like a combination of a number of things. At our, I would say, steady state for recruitment, we believe we can randomize in the low end about 20 patients per week, in the high end about 30 patients per week randomized. So if you have 200 patients, you know the number of weeks you need. With 20 or with 30, it's not that long. It's a little unusual as well. I think other companies, when we're doing interim, is either because they believe they can stop very early, or they believe they need multiple looks to maximize the chance, which is not the case here, or because they really can't enroll.
So if they have to enroll, the second part of the study is going to take a significant amount of time. It's not the case here. It's pretty quick. There is, again, very high interest. So even in the worst case scenario, you would still be reading out next year. We still would be submitting the NDA next year. So it does not fundamentally change the timelines.
Do you even need to get to 200? Because that would sort of imply that you need a lot of patients. There was something not right earlier.
Yeah. Well, the thing that is not right is our assumptions. So if you need that, but even if there are so many scenarios that we could end up with needing 150, that we're going to be giving like as 200, for example, as the option, since they're not going to be so specific on the number, that could be just a little bit more variable, could be a discontinuation at a given point in time. There are so many other options there. At 200, still would be beyond 0.2 on the Cohen's d standardized effect size, which is quite meaningful for this population.
OK. Maybe two or three more questions quickly in the interest of time. The randomized withdrawal study, actually, no, before this, so the DMC, the IDMC has an unblinded look.
Correct.
How about you? How much of an unblinded look do you guys get?
Zero. We have zero. At the moment, we have an unblinded look, then the study is unblinded, and then all the alpha was spent.
So spent. OK.
They are the only ones.
For the randomized withdrawal study, is there anything that you're going to do before you, are there any interims there? How is that structured?
So the studies are intrinsically connected. And I think that is a quite important feature of this study that is often misunderstood. Patients are randomized to study one or study two. That is the first time it's being done in neurology. It is a feature to reduce potential placebo, which I actually think to be, of course, I think to be very smart, but I think others are going to follow in the future as well. So we continually monitor that. It's an events-based study. We know the patients who failed on the second phase. So it's an easier study to monitor, let's put it this way. And I think everyone here in the room and outside would agree that probability of success is extremely high for that study.
But there are no interims there?
No interim.
Based on the decision that you get or the decision you take on the study one, you might change study two?
Unlikely, but yes.
But both will read out together, or will you still read out?
Very closely.
Very closely. OK. What about the filing package? Are these two studies enough for you to file, or there is any additional requirement that you foresee?
No. No different requirements for an efficacy. From a safety perspective, when we met with the FDA, we agree on what we call the minimum criteria under ICH. So 100 patients, one year, 300 six-month. So that's not going to be a limitation for us either. And everything else is either completed or going to be completed before the NDA submission. So very good shape.
Very good. Let's move on to the other assets. So six to eight, obviously, you have generated some very interesting PPR data. The RADIANT is up and running or is going to be reporting out soon. So how translatable are the PPR data, if you can put that perspective? And what really you need to show in RADIANT study for people who sort of have some doubt, can get more de-risking or more comfort with the efficacy?
Yeah. So there are two levels of evidence here that are quite important to the risk in epilepsy. And the first one is unlike other indications, animal models translate really, really well in epilepsy. And we have a pristine package on our website. You can take a look. It really is best in class when you look into that. So that checked the box pretty big. But then when you go for the patients, the PPR study was on patients that have an epileptogenic discharge when you actually expose them to stroboscopic light at different frequencies. There's never been a study where it didn't work on the PPR. It didn't work on focal or generalized. A couple of studies were on generalized as well. So I think it's very hard. Everyone doubts 100%.
So I think it's fair that we all think, oh, maybe if you run 100 more studies, you're going to find one that doesn't work. But as of today, there are more than 30 studies done. And if they didn't work later on, it was because of safety, it was because of exposure, or because idiosyncratic issues, but never because there was not an effect. So I think it makes us feel extremely good about it.
So they are translatable. What about the RADIANT study?
RADIANT, a very good study, fascinating, I would say, for us, because no one ever attempted to get an early read on both focal.
And generalized.
Generalized epilepsy in one package and gain a lot more pharmacology data and a lot more safety data and have that early readouts to inform late-stage studies, as you mentioned, up and running. We intend to report data in the H1 of next year. We believe the mix is going to be 70% focal, 30% generalized, as the general population out there is the same. Really excited about it. If what happens there is what you're seeing in preclinical PPR projections, benchmark studies, it's going to be quite meaningful. While that study is not going to influence POWER 1, that is our parallel group placebo-controlled study, it's likely that could influence POWER 2. It starts with 30 milligrams per day.
30 milligrams per day in focal or any other epilepsy for these patients would be about 20 to 25 folds, depending on the patient, of EC50 on the animal models. The maximum any other drug in history ever got to was five, so going to be, let's say, double to four times the potential therapeutic values here, so it's going to be a quite interesting hypothesis to be tested what happens when you do that, and I think what is going to happen is what you're seeing on our other assets. They're going to have an unprecedented reduction in seizures, increasing in days without seizures, and a very safe profile.
Got it. So it seems like in the POWER 2 study, you might have a lower dose, 10 milligram. So you are not evaluating that in RADIANT. So how will you make that decision?
Yeah. I think POWER 2, which is pretty much. I wouldn't call up in the air because there is a lot of thinking there. But POWER 2 is probably the most fluid of the designs right now. I think the more we walk through the design on that and the more we learn, we need to ask the question if we really need that 10 milligrams or if we actually should push on the other direction, so I think that is the internal debate right now.
OK. Can you talk about seizure freedom, how important that is, both in generalized and focal? What do you hope to achieve? What can you actually show in RADIANT?
Yeah. There are three million Americans living with the constant fear of having to stop whatever they are doing because they had a seizure right now. I bet one of us here in this room had it during infancy or right now. It's not only terrifying.
If not right now.
If not right now, meaning contemporaneous, or it could be a known morbidity. There are so many things you cannot do when you have a seizure, like driving, staying alone, going to a hotel, even as some of us are this week, and many, many other things increase your risk of death by several folds. It's very early death, terminal death is 100% for all of us. So removing that is the ultimate goal on the treatments. I don't think anyone, if you call 1,000 physicians that treat epilepsy patients, all of them are going to agree that that's the ultimate goal, is to not have seizures at all.
No drug right now can do that really well. Cenobamate does that really well. I'm going to give credit to. It's just very, very difficult to titrate, very difficult to continuously give, but it's a drug that does a really phenomenal job, and I think it's the one to beat in the market right now.
OK. So just one more question on six to eight is on the commercial positioning. So I think one of the things that we have heard from. Maybe just explain to me how you are thinking about commercial positioning, because a lot of patients have already been on sodium channel. So would they start on another sodium channel? So how do you see it fitting?
Yeah. So the Human Epilepsy Project did a really beautiful publication a few years ago and said, if you could just be in one thing, what should that be? And it's very clear sodium channel modulators, sodium channel blockers. So then you have to ask the question, why are people not necessarily staying on it? And it's safety and lower dose that they have to go down to lower dose because of safety. So it's safety and safety at the end of the day.
But given a chance, everyone would be using it because it is the most effective. Actually, first to GABA, when you think a priori, you would think that GABA is actually more, but it's actually sodium channel blockers. So what six to eight or relutrigine brings to the table is actually you eliminate the issue on safety, and you increase the margin for efficacy.
And it does not need to be dosed as the only sodium channel mechanism. We had the same questions about our relutrigine program that we're not even getting there. And we dose on top of one, two sodium channel blockers right now. And we're seeing not only additional efficacy, but no real new safety concern and completely unprecedented seizure-free days. So I think that concern probably is unnecessary right now.
Got it. Maybe in the last, let's say, 30 seconds or one minute. So when you produce the data on relutrigine, I think there was some confusion. What do you think community or investors don't understand? Because I thought the drug was active when you put out that data. And obviously, you're getting some credit for it, but help me understand how is that differentiated?
Can't possibly begin to think that I know what investors didn't understand. What happened on the data would be so pretentious of me. The patients are incredibly severe. So we designed that study in a way that would preserve the blind on drug or placebo, but maximize the chance of being on drug during the study. And going back to the very beginning of this conversation about risk on CNS, everyone repeated the past. And no one thinks about the patient needs and what the drug can do. We did. And therefore, it requires a little bit more thinking than just copying another study. And I would say that's what happened.
But what we've seen was a seizure freedom that has never been seen before in this condition. I think that's what we should be actually talking about, is how hard it is to understand 33% seizure freedom in this patient population. That's not that hard, even for some of us.
I think everybody keeps talking about certain DEs are contraindicated for this. How do you answer those questions? I mean, it is easy potentially not to enroll those populations. How are you thinking about developing it across all DEs?
Yeah, so we have an all-DE study starting pretty soon. We're aligning with the FDA on the bells and whistles and what is required and criteria. We're going to be talking a lot more about that. But we think there's very little restrictions for a drug like relutrigine there.
Perfect. Very good. Marcio, thank you so much.
Yes. Always a pleasure.
Thank you.