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

Great. Good afternoon, everybody. Welcome back to the Bank of America Healthcare Conference day 2. I am Tazeen Ahmad. I'm one of the senior SMID biotech analysts here at the bank. It's my pleasure to have our next presenting company with us, Praxis Precision Medicines. Sitting next to me are Marcio Souza, who's President and CEO; as well as Bernard Ravina, who is the Chief Medical Officer. Gentlemen, good afternoon, and thank you for coming to Las Vegas.

Thank you. Thanks for having us.

So maybe, Marcio, can we start with a quick overview of the company, maybe a 2-minute summary of your platform and some of the recent developments, and then we can go into some more detailed questions?

Yes, absolutely. Actually, exciting to be here right on the heels of positive data that we had this week. Three areas of the company: epilepsy, movement disorders and psychiatry. We have assets in the clinic in each one of them. They're all based on genetics. That's the first pillar, the way we screen all the targets. We think that's fundamentally derisked the way we move. I'm going to use that as an example for what happened this week. Then from that point on, not driven by wish or by hope, but driven by really the translational science, the really strong biology. We look into first in the animal models that were already somewhat derisked by the genetics underpinnings. And from that point, once we are very certain for that stage of development that we should move to the clinic, we tend to design the first studies to the proof of concepts relatively short. And the relatively short what we call is like 1 to 3 months in duration. So then we can invest real capital, real like people capital and so on. So when you put out those pillars together, always with the wrap-up of what do patients need, not what we need as a company, but what patients need in the markets. It so far has been a very nice recipe for success. So when I try to use that example for what happened this week with 944, right, we are presented earlier in the week, on Monday, what I would call groundbreaking data for T-type-causing channel blockers in essential tremor. But that goes back to why we started that program, right? Cacna1g is a gene that is fairly prevalent in epilepsy. That's how we filter most of our programs. We use that data and ask what else is affected. It was one Sunday in genetics of essential tremor coming up. We saw that, that was one of the key things. And we asked ourselves, does it make sense from a physiology perspective? And we look into the network in the brain, the cortico-thalamic cerebellar network, specifically for this case. Those receptors, they kept 3.1, 3.2 and 3.3 are all very prevalent on that region. And we actually altered that. So we look into animal models that exists both on presence of the drug in the brain using biomarkers like EEG and on reduction of tremor. Both of them were positive. So the key question becomes, if we move to humans, can we get to exposures that are high enough, that are on this region that we need to be, in this case, the thalamus. So we run a translational study at that point in time. That showed very clear activity between 5 milligrams and 120 milligrams. This -- what we're seeing was a very clear change on the sigma bands. So now we know the drug works, right, from a mechanistic standpoint. So the next question is, does it reduce tremor? And what we're seeing this week is it reduced tremor, increased quality of life, activity of daily living for these patients. So it's the full circle. We did that once for essential tremor, just trying to use that as an example, as a company, but we're going to be doing this again and again, like soon enough. I'm sure we're going to be talking about this next month. We have a psychiatry trial reading out in major depressive disorder. We have 562 program for sodium channels coming up as well in the next few months, the first study, and many others throughout the year. We're very catalyst-rich, which we consider to be a really good thing on the market the way it is. This was 1 of 6 readouts we have this year. So we're going to have readouts like every 45, 60 days or so between now and the end of the year, which is quite phenomenal. So very proud of -- we have 150 people. Every one of them are incredibly engaged on what we do. So that's a little bit of update here out for the company.

Okay. Great. So let's go a little bit deeper into a couple of topics that you just discussed. So let's start with 944. So like you said, you provided an update on some data that you were looking at for essential tremor. Some of the questions that we've been getting relate to whether or not you saw a true dose effect. And I think in particular, some might be questioning whether the efficacy seem to plateau at around maybe the 60 mg dose. What's your view on this data? It's early, you might still be analyzing a lot of it, but just on the first path.

I think it's a very reasonable question, right? When we look into the translational data that I mentioned before, right, so quantitative EEG versus exposure. What you're seeing is it starts getting some effect around 5 milligrams. When you are in the region between 80 to 100, it does lapse up. So one could argue that at 80 milligrams or so, getting the max more effects. Now we wanted to push the dose higher to 120 to make sure the safety was okay, which was in this population, and that it's in the population that has an alteration or has an anomaly on the network, maybe for some patients in the future, you're going to need to push higher. But I think the system is quite valid actually. Right now, it's indiscernible. Once we cross that 60 milligrams, do we need more than that? It's unclear. We're going to the dose that patient tolerates, right? Some of them tolerate 60, some of them tolerate 80. Most of them tolerate 120. But where we're seeing the effect is what matters the most right now. After we cross that 40 milligrams, 60 milligrams, we're seeing pretty much everyone that would respond responding, which is quite amazing.

Yes. I think that's the key. We designed the study to understand how far we could get up. Could we get to the top of the dose range based on the pharmacodynamic marker, which is tapped out at that level of plateaus? We can. Most people tolerated the top dose, but we've never designed it to really understand the dose-ranging efficacy. So I think that's something we'll learn. But the questions we needed to address to get to an efficacy study, we have that.

Okay. So are you better able to predict which patients are going to respond to your therapy just after looking at this first cut of the data?

We are in a sense that we know they respond relatively early, they start responding relatively early. So as we move forward, as we think about other designs, eventually the drug in the market, it's not going to be the kind of drug that we give patients for like several months and then eventually you're like, oh, that did not work. So we can see a progressive -- it seems to be very sensitive the network to the drug itself. One thing that is dissociated with side effects, it's very good from a clinical development perspective because you don't want functional unblinding and things like that. So some patients respond but they don't tolerate. We had 2 patients that do not tolerate the drug on this. Actually, one of them was a good responder. So it's not necessarily related to that. I think as we expand the cohorts, as we have now a potential one, we understand better. There are probably going to be some indicators of patients for a response. But right now, we know how to titrate. We get these patients in very quickly, a week or 2, if they're going to respond or not. And if they are, as the vast majority of them will, then we can continue titrating down.

Yes. And we definitely learned that people need to have a certain level of tremor monitored to severe in order to be able to reliably measure tremor. So that's definitely -- we addressed it. What's the dose range? Who's the patient population? If they don't have about a 10 on their upper limb score, which is just a few items, you really can't measure a change. So they may respond to the drug, but they're not good people in the trial because they're too unstable to measure.

Okay. Now what about for the patients that needed to down-titrate? Is there a particular profile of patients that can only tolerate a certain level of drug? And how does that make you feel about the overall adverse event profile going forward?

So on the data set we showed right now, right, 8 out of the 11 made to 120 and stayed there. So it's pretty high. One was 100 milligrams. So had to down-titrate to 100. Normally, what -- the most prevalent side effects here that people down-titrate is around dizziness and manifestations of dizziness can be called a little bit different. It does resolve very quickly if you remove the drug. But for some of them resolved once you push down the dose a little bit. So we actually think that it just shows the drug is very active in the brain. And they still respond to the previous dose, which is very good. This is a refractory population, right? Patients who enrolled on this study, they had years of essential tremor. The 80% of them fail on prior medications. So they were not the ones, the easiest patients to treat. The fact we can get that to respond with a fairly positive, I would say, safety profiles are never positive. Let's just be honest. But like with one that is manageable, we're pretty happy with it.

Yes. And the AEs that occurred that led to down-titration, they tend -- or discontinuation, they tend to occur early. And so one of the ways we've addressed that in Essential1, which is ongoing, is just having started off at a slightly lower dose and let them ramp up. Generally, it looks like those AEs accommodate. And as people go up in dose, they do fine. So it's probably just the beginning that is the real key to tolerability.

So you looked at 3 doses, right? You looked at the 20 and the 60 and the 120. So do you think you'd want to start lower than 20 in some patients?

Yes. So we're going to -- in the ongoing study, we're starting at 5 milligrams which, according to our EEG data is already an active dose, marking T-type calcium channels in the brain.

And is there a minimal threshold that the drug needs to achieve in order to be minimally efficacious? So is 5 mgs -- you're seeing efficacy, but is that enough?

Probably not. We probably have to go around 40 to 60 to see efficacy. That's our clearance hypothesis. So 20 milligrams might be very little, if any, effect. And then 60 and 100s, maybe they even are going to be distinguishable from an accuracy perspective. Just running the experiment as we understand right now. But 60 is the best prediction right now when you look into the data we have.

So how quickly do you think you'd be able to titrate patients that you start off at 5 to get to the application's level?

Yes. So for the current study we just reported this week, they could not change dose after day 36. So between 0 and 36, they went to 120. It's actually not any different if we start at 5, just go from 5 to either 10 or 20, depends on what -- and then from there. So it's one step really. It's mostly -- those are hyperactive networks and just wants to block the channel like sequentially.

So in terms of the dizziness that you mentioned the patient's experience, how do you think that would impact the ability to make it widely commercially attractive? And is this something that patients might end up needing to dose at night in order to reduce the prominence of the quality-of-life impact from having dizziness?

Yes. One of the nice things about having the ADL measures is it tells you how people are functioning, incorporating both sign tracks as well as efficacy. So they're functioning much better overall. And the main side effects we saw are CNS that would show up in their ADLs, right? So I think you always want to have as few side effects as possible. But overall, the benefit/risk was very favorable when you look at it from that point of view. And then the drug has a really nice flat Cmax because we use this modified-release profile. So I don't think like dosing at night. I think the way we do it in a trial is actually much faster than it will be done, the titration in clinical practice. In clinical practice, people are going to go up very slowly, see where there's benefit in trying to treat these people for many years since it is a chronic disease. So I actually think it will be less of an issue in real-world clinical setting that would be the trial where we're pushing the titration.

Okay. That makes sense. Let's move on to epilepsy for a couple of questions and then we can do MDD. So you recently held an R&D event where you profiled your epilepsy programs. You want to take sample shots on goal with epilepsy. And I guess why does epilepsy make sense right now given that you're already tackling 2 potentially large indications with essential tremor and MDD? And I think people would be happy to give the company credit once those are fully validated. So why also talk now about epilepsy and your desire to move into it? And also, can you just give us some color on, I guess, PRAX-222, which you did talk about on the R&D Day, which has since been put, I think, on a clinical hold?

That's right. That's right. So the epilepsy or the reasons for epilepsy as the franchise #3 here, maybe are 3 very, very important, right? One is the unmet need is incredible out there. Millions of patients in the U.S. like with nonresponsive refractory epilepsy is kind of every single day with drugs that are really not that great and have a lot of side effects. We'll talk about dizziness, for example, virtually all these drugs right now creates like insane levels of dizziness in these patients. So the patient level, that's how we guide pretty much everything in the company is there. It's overused in the industry, but it's very serious for us, right? But then strategically, every screening we're doing is based on epilepsy. And why? One, the genetics of epilepsy is better understood than anything else in CNS, so we can derisk. And the models are the most projected, the animal models. So when you put that together, every drug that's going through was once upon a time an epilepsy drug. Then the question we have to ask is, okay, do we go to movement disorders, psychiatry, or we stay in epilepsy? Why is the reason to stay in epilepsy for some of these compounds or to develop specifically for epilepsy like some of the ASOs that we are doing? The market is large enough. Number one, the unmet needs is very large as we just thought. But the regular framework is very straightforward, right? It doesn't sound like based on what we just talked about, a clinical hold. But once you get to the clinic, showing benefits, end points, the bar, getting an approval is very straightforward. So if you put that all together, we're talking about multibillion-dollar opportunity with what we have in our hands. We're not developing the company just for 1 stakeholder here, right? We're really developing the most comprehensive, prominent CNS company that is guided by genetics. So it just makes sense strategically for us. Now going back to your second question on the clinical hold for 222. It's obviously disappointing, right? I'm not going to say we're happy when we received the communication from the agency. It's our first ASO. Ionis is our partner here, as you probably know. The preclinical package, the IND-enabling package, was very extensive. And we had really extensive meeting as well with the agency in terms of a pre-IND meeting in terms of when should we start, the toxicities that exist in this molecule, the pharmacology, how to make sure it was safe to be given and so on. And the response were very good. It's very satisfactory to us. They did encourage us very heavily to look into a seamless design, but that might have been the most vague part of the advice. We took all of that. We designed a package. And if you might recall, when Bernard presented at Epilepsy Day that we had in New York a few weeks ago, we talked about that was probably the biggest area of risk for us. We don't know why the clinical hold right now. They're starting dose on the pre-IND meeting and today is the same. So we don't believe that, that's the issue. The tox program had been completed back then, so that shouldn't be the issue. So maybe modeling for the dose, maybe how fast we're going. Maybe they don't feel comfortable with the expansion cohorts. We are waiting to hear from the agency as soon as we're here. We're going to be updating as we said. We are built on transparency. We got the letter, the email from the FDA 4:30 p.m. Before the market opened the next day, all of you knew. We're going to continue to do so. We believe that that's the best way to interact with stakeholders. But right now, you know as much as I do.

Okay. So what is the normal process that FDA goes? So they'll let you know that you're on a hold? And then do they have a certain amount of time to give you color or more?

Yes, they do. They have 30 days to send us the letter. They reinforce that they might need to use the 30 days. Historically, they have not. In other holds, normally, 2, 3 weeks. It might be as early as next week. If it is clear, as we expect to be, we just make the notifications. We submit our complete response to the hold. And then 30 days later, we might be able to start the clinical trial. That's what we expect to happen. There is an alternative scenario here that is not clear, right, that they send us their comments. We have no idea. We still have no idea what's going on. We can request a Type A meeting, then 30 days for the meeting and that, then we resolve after that. By the color we got so far, it should be straightforward. So we're going to indirect color me like they say, wait for the comments, it should be clear. We're going to wait, and then we're going to move forward from there.

So is that why you didn't immediately request a Type A meeting because the letter said basically wait?

Yes. One of the alternatives that we always have to work on is that the agency is trying to do their best to help all of us, right, as in the U.S., they are just to a limit of their staff. So by pushing a meeting that is not necessary, you're not helping the agency. And we always try to work in collaboration with them. Before designs, you had 6 INDs cleared on the last 12 months, right? So we know the process. We have a very respectful relationship with the agency. We've been pushing off those. They all got cleared. Why push this? It didn't make any sense for us. We're going to wait. Hopefully, 1 or 2 more weeks, and then we're going to move from there.

Okay. So at the R&D Day, you did talk about different types of epilepsies that you might want to focus on. What's your view of rare epilepsies versus much of the more common ones?

The needs is very clear on both sides, right? Maybe a little bit more clear to quantify on the rare ones and a little bit more variable on focal, for example, or generalized epilepsies, but we see the need on both. We have compounds like 562 that we are in healthy volunteers right now. We dosed for 28 days. We have now exposures that are so high. They were pretty certain we're going to get an incredible margin in patients. That makes sense to go to both, to use that class of compounds to go to rare and to focal, for example. That's why we got a second drug on that family 628 to go to focal epilepsy. The market, the way to go to market, the number of trials is different. In a rare epilepsy. One registrational trial should suffice. Very high burden of seizures, very clear reduction on focal, for example, probably going to need two. So the rate of success as we start are much higher on the rare ones. So we're going to be balancing use of capital employment of success from the beginning to address these populations. But like if you look into companies that got either acquired recently or got to very good development in terms of the market, epilepsy is an open market in terms of the potential. So we're chasing that as well.

Right. So yes, so there was GW, there was Soligenix on the rare disease side. There's Xenon on the more clinical...

Which we respect and appreciate a lot.

Would you consider them, on a go-forward basis, a competitor? Would you be going for the same types of indications?

I think the overlap that we have is with the last one you mentioned, right? So very large market for focal like Xenon is developing a drug. Again, as I just said, I respect them. I think they're doing a phenomenal job. The drug seems to work really well. There's still going to be a space there to make it better, and we think we can get another part of the market. For the genetics, we tend not to overlap. We want to be first or best -- or first and best in that.

So you wouldn't do like Lennox-Gastaut or Dravet?

Not to the point that we thought we couldn't be much better than what exists right now. So we're only going to go if we really have such a clear view that we're going to be the best if we are second. If you're first, we want to be first and best. But if you are second, we want to be best.

And we have kind of a unique opportunity. So in our pipeline, we have at least directly targeted causal therapeutics like knocking down SCN2A sort of gain of function. But we also have drugs like 562, which are based on genetic mechanisms, and we're able to apply it to children BEEs but also extend that mechanism more broadly to focal epilepsy. So I think it's -- we can meet the needs in 2 ways for these children with BEEs. There's both the seizure need, but also the developmental needs, which are going to be met by truly targeted therapeutics, right?

So maybe let's stay on the topic of focal seizures. So you mentioned a second ago, Marcio, that it's a big market opportunity. And you have a good view of the data that's presented so far, but there could be room for improvement. Are you referring to a level of efficacy or safety? Or is there something else?

And I would say both, and they are normally more correlated than not. And epilepsy is how high can you push the exposure normally needs to better efficacy. When you look into 562, we entered the second Phase I study with about fourfold the EC50, which is very translatable in epilepsies, which is already way bigger than anyone else that we've seen before. We are now several folds higher than that on exposure on a drug that is incredibly effective on the animal models, which led us to believe that we can really resolve seizures in a good part of those patients and not lead to side effects. It's always a balance in general in drug development, but more so for epilepsy.

Yes, yes, absolutely.

How far do you think you are from actually demonstrating that with clinical data?

We're starting the persistent patients in the next few months in epilepsy. We expect that to recruit relatively well. You're going to hear about the design pretty soon. So we're not guiding quite yet because we like to start this study 2 guides. But I'll say in the next 12 months, you're going to see proof of concept for 562 in patients.

Okay. We look forward to that. So we have a few minutes left. I wanted to just go over a little bit of the top level, what to expect for your upcoming MDD update for 114. People will inevitably always compare your data set to Sage. So maybe try to help us set what the right expectation should be.

Yes, absolutely. So what patients with MDD needs are drugs that work faster, better, they can rely upon for months during the episodes and that are safer than the alternatives that exist right now. And we are set up to deliver on all of that. So Aria enrolled 216 patients. We had last patient last visits for the safety follow-up last week. So we are on track to deliver the top line next month. Incredibly bullish. Every single assumption we had on that trial pan out. We know the controls we put in place like using eligibility criteria that is stricter than others resulted in a more homogeneous population, a more stable population at baseline. So we're expecting to see 3 points or more separation of day 15 on HAM-D17 that are largely maintained at day 29. That's something that is -- does not exist currently on other programs. So we're not expecting to lose the effects.

But do you expect to be a [ side effect ] at day 29?

We're not powered for that, but we're pretty bullish about like what to expect from this drug and side effects that are around 10% of somnolence, that was our target. We believe we're going to be within the target. No sexual dysfunction, no weight gain, no needs to have an SSRI to rescue as others in this class are developing. This is not a co-initiation. It's not a quick hit and then you get better in 3 days or 4 days, and then you have to go back to the horrible drugs that exist from side effect profile. This is a drug being developed to treat the episode of depression 3 to 6 months. That's what the FDA believes. That's what we believe. Patients can stay on because there's no liability associated with that after 2 weeks. And they don't need to be on underlying SSRI. That's what you're going to see. You're going to see day 4, 8, 15, 21, 29 for the primary end points in a matter of weeks, right, in June. We are almost mid-May now.

So you're not powered for septic at day 29, but what about any time before that?

We are at day 15. Day 15 is the primary end point, and we are powered for 3 points. All the assumptions we had in terms of completeness of the data at day 15, standard deviation and other measures that we have that are all at or better than the original assumptions, which makes us very bullish about this setup.

Okay. Now you're saying that you feel really good that you'll see the durability of response going beyond your 15 days. But if you aren't powered for septic, how do we really feel comfortable with that trend?

Yes. I think the way to feel comfortable there and what to expect for drugs in general in depression that work is we're going to have a maximum effect between day 7 and 15 for this mechanism. After that, we should see a straight line. You should see a line that patient that's not moving, right? Now depression is not a chronic condition. Depression is an exotic condition. So some patients, they start feeling better over time. So we might see the placebo patients feeling a little bit better, reducing the -- so you might reduce the numerical difference. They might or might not be sufficient to show a statistical separation or statistical powering. But the numerical difference is going to be present, and that we're pretty certain of.

Okay. And so you've kind of walked us through what's going to be presented at the top line. What are you going to hold back just so that we're clear?

So we are trying to be as comprehensive as possible for this, as you can imagine, like it's going to be a race towards by getting all the stables already. We are very forthcoming of time lines. I just told you, last patient last visit May 5. June readouts that puts us on best-in-class in terms of cleaning the data reporting. But we're going to be reporting responder analysis for these patients, the subitems of anxiety because that became a little bit more important, I would say, in the real world, right? These patients with high HAM-D have high anxiety. So it's obviously expected to work, but we can't only see effect on high-anxiety patients. We want to see effect from the general population. So HAM-D at day 15 and 29 as key safety follow-up, what happens once you remove the drug, very important for safety and as many secondary end points as we can analyze in a matter of days.

Okay. So with that, I think we're out of time today. So thank you very much for coming and presenting in Las Vegas. And thanks, everybody, for listening to us.

Thanks for having us.

Thanks.