AtaiBeckley Inc. (ATAI) Earnings Call Transcript & Summary

Thank you for standing by. My name is Amy, and I will be your operator for today. At this time, I would like to welcome everyone to the AtaiBeckley EMP-01 Phase IIa Topline Data Conference Call. [Operator Instructions] I would now like to turn the call over to Jason Awe, VP of Investor Relations with AtaiBeckley. Please go ahead.

All right. Thanks, Amy. Good morning, everyone, and thank you for joining us today. We're here to discuss the positive topline results from our exploratory Phase IIa trial of EMP-01 in social anxiety disorder. Joining me on the call are Dr. Srini Rao, Chief Executive Officer and Co-Founder; and Dr. Kevin Craig, Chief Medical Officer. Before we begin, I would like to remind everyone that during today's call, we may make certain forward-looking statements as set forth on this Page 2 of the presentation. These risks are detailed -- discussed in detail in the Risk Factors section of our most recent Form 10-K and in subsequent SEC filings. We undertake no obligation to update or revise any forward-looking statements, except as required by law. With that, I'll turn the call over to our CEO and Co-Founder, Dr. Srini Rao, to begin with background on social anxiety disorder and the context for today's results. Srini, over to you.

Thanks, Jason, and good morning, everyone. We're very excited to share these first inpatient data for EMP-01, our oral R-MDMA candidate, with all of you today. As the first controlled clinical trial of EMP-01 in social anxiety disorder, or SAD, the study gives us an important early look at this compound with its unique pharmacological profile. In short, we're encouraged by what we've seen across feasibility, safety and initial efficacy signals. On the next slide, I want to start with a brief overview of SAD because the scale and impact of this condition often go underappreciated. SAD is one of the most common psychiatric disorders globally, affecting hundreds of millions of people, and it remains a chronic and highly impairing condition for many. Despite that, nearly 60% of people living with this condition received no treatment at all. Even among those who do seek care, many continue to struggle with persistent fear, avoidance and significant functional limitations in everyday life situations. Despite its prevalence and impact, there have been virtually no pharmacological innovations in this indication in over 20 years. Today's standard pharmacotherapies, typically SSRIs or SNRIs, require daily dosing, can take months to achieve a meaningful clinical benefit and over half of patients do not respond adequately. And even for those who do, the improvement is often partial and tolerability can be a limiting factor. So we're left with a very large, very well-characterized unmet need for treatments that act more rapidly, more robustly and ideally with intermittent dosing rather than continuous daily exposure. That's the context in which we designed this exploratory Phase IIa study of EMP-01. Our goals for this first inpatient study were twofold. First, to understand the safety, tolerability and feasibility of 2 in-clinic administrations of this novel agent in a real-world SAD population. And second, to determine whether we can see early clinically meaningful signals on accepted regulatory endpoints, all without psychotherapy. I'm very pleased to say that the study delivered on these -- on both objectives. The team enrolled a highly severe patient population, executed with strong adherence and achieved exceptional retention across the trial. Importantly, across the data set, we see a consistent pattern of improvement that supports further development of EMP-01 in SAD. Kevin will now walk you through the fourth study design, safety outcomes and the key efficacy signals we observed. Kevin, over to you.

Thank you, Srini. So on the next couple of slides, I'm going to talk through a bit about the study design and the baseline characteristics to really frame the results. So moving just to the next slide, I'll talk about the study design. As Srini mentioned, this is an exploratory Phase IIa randomized placebo-controlled study that was primarily designed to assess safety, tolerability, feasibility and efficacy of 2 doses of EMP-01 in 70 patients with social anxiety disorder. So as you can see in the graphic, participants are randomized to either EMP-01 225 milligrams or to placebo, and they received 2 doses, one at Day 1, one at Day 29. The endpoints were taken at Day 43, both the safety and efficacy, and we had a safety follow-up 2 weeks later. Importantly, this study was not powered for statistical significance and efficacy, but rather to characterize the safety and determine whether we could detect clinically meaningful signals worth advancing in this or other indications. On the efficacy side, the main secondary endpoint was change from baseline in the Liebowitz Social Anxiety Scale, the LSAS, total score between baseline and Day 43. And we had a number of other exploratory endpoints, including other clinical scales such as the CGI-I, the CGI improvement scale, which we'll talk about today, and others which are not part of the top line, but will come at a later date. An important aspect of the study was the fact that we went to make sure that we used central raters for all the clinical measures, including the LSAS and CGI, and this is intentional and adds rigor and improved consistency. And as Srini mentioned, there was no psychotherapy. We had a typical psychedelic prep and in-study support as participant support, but no psychotherapy was delivered. And in terms of enrollment, we randomized 71 participants, of which 70 received a dose and are part of our modified intention to treat analysis set. We had a very high retention rate with 69 participants completing the Day 43 rate. So I think a very high retention rate. And we're very pleased actually with the -- both the results, but also how the study was run. It was run, as Srini mentioned, kind of effectively enrolled well. We had good baseline characteristics. So maybe on the next slide, I'll move across and talk a little bit about what the patient group looked like. So you can see here the 2 middle columns are the baseline demographics and other characteristics of the placebo group on the left and EMP-01 on the right. And you can see that it was very well balanced across the demographics in terms of age, sex, ethnicity. There was a slight imbalance in terms of the number of prior medications that people have used for their social anxiety disorder. But this is very small, and I think it didn't affect the results and it's typical of a small study of this kind. Just to draw your attention to the middle of that, which is the baseline LSAS score, which in this case is very high. So we had a mean baseline LSAS of 108. And just to put that into context, and we'll show a slide a bit later, many of the registration trials historically have had LSASs of around 80 to 90 in -- at the baseline. So again, this represents quite an unwell group with severe disorder going into the trial. I think in general, as I said, we're very pleased with the results and how this has balanced out. So on the next slide, we're going to talk a bit about the safety. So I think the general takeaway was that this was very tolerable. It was -- it had adverse event rates, which were kind of expected. We had, as you can see at the top line, a number of treatment emergent adverse events, both in the placebo group and the EMP group, but a slightly higher proportion of treatment-related treatment adverse events. What I would say, though, is that we had only mild or moderate adverse events throughout the whole study with no severe adverse events and no serious adverse events. We had -- towards the bottom of that table, you can see that we had 3 participants who were withdrawn from study drug due to adverse events. Importantly, to note these were all mild or moderate, and a number of them were predefined due to vital sign. So I think in summary, we had a very well tolerated safety profile. We had no severe or serious adverse events. There was -- we looked very carefully at suicidality, and we had no suicidal intent or behavior reported during the study. We also had -- we had a measure of the CEQ, that's the challenging experiences questionnaire, which really sort of measures how difficult the experience was, and we had very low scores here as well. So I think generally well tolerated during dosing with the expected adverse events. And on the next slide, I'll talk you through some of those adverse events. So on the left, you can see the adverse events that occurred. These are drug-related treatment adverse events that occurred at more than 15% in the study. And we've separated these out into those that are adverse events of special interest and not typical of the psychedelic experience and those that are not part of that list. So on the left, you can see the most common there, things like nausea, headache, fatigue, dizziness, decreased appetite, palpitations, vomiting, hyperhidrosis, bruxism and blurred vision. And again, these are very typical of the class. And then we also had a number of these more psychedelic-related adverse events such as feeling of relaxation, sensory disturbance, hallucinations, euphoric mood, anxiety and somnolence. So again, very much what we expected and what we saw in Phase I, really no surprises, very consistent across the Phase I and Phase II studies. On the next slide. Here, I just wanted to talk a little bit about some of the difficulties in treating patients with social anxiety disorder. And Srini alluded to this a little bit as well. So the way that we measure social anxiety is twofold. One is the sort of a symptom that is the fear of the context. And the other is the -- whether or not participants or patients avoid those contexts. And so this could be anything from talking to a person that you don't know very well all the way up to speaking in public or being in a group setting in small group setting that kind of thing. And importantly -- it's important that people have both a reduction in their fear but that then has to translate into a behavioral change. So really, we're looking for people to then go out and experience these difficult experiences, and they then need to not avoid those. So in order to get an improvement on the LSAS scale, you really have to have a reduction in fear, but also a reduction in avoidance of those stimuli. And that does take some time. So if you look at the scale, there's something that happen very commonly like speaking to someone you don't know very well, but also there are things like holding a party, which takes a bit of time for those things and most of us don't have -- can hold parties too often. So there really is a time factor to the sort of how quickly the treatment trajectory for this illness, which is, I think is different to some of the other conditions, which are very much based on symptoms like depression or anxiety writ large. And I do think that EMP-01 does show great potential here for changing both the experienced fear in these situations and also the behavioral change, that change in avoiding those situations and really with a dose that is really twice in 1 month, so 1 month apart rather than a daily dose, which often takes some time to achieve. So moving on to the next slide, we'll talk a little bit about the efficacy. So on the LSAS, what we had was a least squared mean treatment difference between drug and placebo of 11.85 points, and that translated to a standardized effect size of 0.45. And this is very much in line with treatment of standard of care at the moment, both pharmacotherapies like SSRIs and SNRIs, but also for psychotherapy. So what we got here really after 2 doses and 6 weeks was a treatment effect that is in line directionally with treatments that usually take 6 to 12 weeks to have an effect. Next slide. Just looking further into the LSAS scale, as I mentioned, there is both this fear axis and an avoidance axis. And we were very encouraged to see that both of these pretty consistently moved between baseline and the end of treatment. And in every case, they were higher in the EMP-01 arm. So both the experience of fear in these situations and people were less avoidant of the situations because the fear was lower. And we looked at that in terms of subscales within that, like fear of social interactions, avoidance of social interactions, fear of performance -- this might be giving a speech or speaking in public even on the telephone -- and avoiding those public engagements. And again, consistently throughout, we saw greater improvements numerically in the EMP arm. Moving on to the next slide. We also looked at the CGI-I. This is the Clinical Global Impression of Improvement. And again, people coming into this trial, as I said, a high LSAS score, they also had a high CGI severity score coming in. So we had a significant disease burden in this group. And here, we saw some quite significant differences at Day 43 in terms of CGI-I rating. So here, really, nearly 50% of people were deemed responders in the EMP group, and that is a CGI-I score of 2 or less, as opposed to 14% in the placebo group. And this really translates to a number needed to treat of less than 3. And I think number needed to treat is one of the ways that we really want to assess the clinical meaningfulness of these scores. And this really means that in order to obtain one participant who responds and is in remission at the end of treatment, you need to treat 3 patients. And just to put that into context, I mean, that's a very low number. So in social anxiety disorder with other pharmacotherapies, the range of number needed to treat is around 3 to 5. And again, that's quite low in comparison to other psychiatric conditions like depression, where the number needed to treat is often 8 to 10. So we had a very good response rate in this group with EMP-01. Next slide. And just to put this in a bit more context here, you can see in the green row at the top of this table, these are the numbers that we got from our trial. Baseline LSAS, as I say, was higher than the kind of combined, if you take SSRI registrational trials as a whole and SNRI registration trials as a whole. And also we compared this to psychotherapy where there's less data, but we have some ideas about effect size. And what you can see is that we compare very favorably both in terms of the severity of the condition, but also then the change from baseline, which is very much in line with standard of care for SSRIs and SNRIs with a number needed to treat around 3 to 4 in the case of other pharmacotherapies and about 3 in our case. And I think importantly, these trials often take 6 to 12 weeks and really does take some time on daily dosing to get there, whereas we've achieved this with 2 doses over a 6-week period. Next slide. I also wanted to just talk a little bit about the experience of this compound relative to other psychedelics and relative to racemic MDMA. So on the left-hand side, you can see from this trial, the 5D-ASC, that's the 5 dimensions of altered states of conscious. It's a very commonly used psychedelic scale. And you can see the comparison between EMP-01 and placebo. And across the scale, we see quite high numbers for a number of these dimensions such as reduction in vigilance and oceanic boundlessness, but across the scale. And on the right-hand side, if you compare that to the literature, and again, this is directional. I think it gives us an idea. It's difficult to compare study-to-study. But if you take this as directionally correct, what we're seeing here is psychedelic ratings that are kind of in line or similar to psilocybin and actually quite significantly higher than racemic MDMA. And that's very much expected given the preponderance of 5-HT2A activity with R-MDMA EMP-01 relative to racemic MDMA. And again, we saw less of the monoaminergic, dopaminergic, noradrenergic effects with this. So we had less bruxism, less muscle tightness and things like that, but certainly higher rates of hallucinations, visual disturbance, et cetera. So again, I think this is a unique profile that we see with R-MDMA, but very much in line with the pharmacology that we would expect. And with that, I'm going to hand back to Srini.

All right. Thanks, Kevin. Let me close on the next slide by stepping back just to summarize what makes these findings and EMP truly meaningful. As Kevin kind of alluded, what stands out for me is in this data set is a unique profile of EMP-01. It's a compound that delivers a predictable psychedelic like therapeutic experience without the stimulant driven liabilities racemic MDMA and with a tolerability profile that aligns well within our patient interventional use scenario. That's exactly the differentiation our advisors highlighted in the press release, and it's encouraging to see it reproduced. At the same time, what we reviewed here are only the top line results, as Kevin indicated, for this exploratory Phase IIa study. We will still have additional analyses that are underway. We do have additional analyses and some additional data sets will be forthcoming. Those will help us -- they help inform us how we think about the next phase of development, and we want to make sure we give that process the rigor it deserves. But overall, we're genuinely encouraged by what we've seen so far with the strong feasibility, a clean and manageable safety foundation and early but consistently positive signals of clinical benefit across multiple measures. All that gives us great confidence in the path forward for this product. Before I close, I want to thank the patients, the investigator and the site teams who made the study possible. Operator, I will pass it back to you for questions.

[Operator Instructions] Your first question comes from the line of Andrew Tsai with Jefferies.

Thanks for this interesting data set. Congrats.

Thank you.

So it looks like your study completely stopped at week 6, maybe no open-label follow-up. If that's true, can you maybe somehow find these patients again and maybe track their LSAS scores just to see how efficacy could be trending over time? Or is that not possible?

Yes. I mean, again, this is an early stage trial. So it's common not to do any kind of extension. We certainly talked about following up with these patients. Kevin, I don't know if you want to jump in on that.

Yes. No, absolutely. I mean I think with the sort of data in hand, that would be great, and we had exactly that discussion. But I do think this was really framed as a very exploratory study. It's been a long time since people have studied social anxiety disorder. So I think for us, this is really about feasibility, safety in this patient population and some idea of efficacy. And certainly, in the next study, we want to look at that durability in a more focused way.

Right. And that actually goes to my second question, the next step for EMP-01. When do you think you could start like a Phase IIb or even a Phase III next? And in this next trial, what kind of time point for the efficacy primary endpoint would you be choosing? And would you be choosing also the same 225 mg dose?

Yes. No, I mean, this is really a discussion of the top line readouts. As I mentioned, we are taking these data, and we're still awaiting some additional data. We'll be doing those analyses and then come back with a more formal plan. So this is really to relay the top line results and our excitement for them, and then we will obviously have to do our own analysis to understand how we want to take this forward. I mean I can say sort of in general, the other trials that the pivotal studies from a long time ago that Kevin alluded to did have endpoints in the 8- to 12-week range. So that seems to be fairly -- that's something that's come up previously. And of course, durability is going to be a key one here, but those are very general comments.

Your next question comes from the line of Ritu Baral with TD Cowen.

First question is around time to discharge readiness. Did you guys assess that in these patients in the 2-dosing period? And then my follow-up was on the 3 discontinuations. If you could give sort of time for resolution both for those, I think they were cardiac. And then just a little more detail on the other dropouts, please.

Yes, sure. So maybe address the first question. The patients that dropped out, starting there maybe. So we had 3 participants. We had some criteria around blood pressure, which were predefined. And we had 2 participants who had increased blood pressure, one just very marginally above the cutoff. But we were very keen, again, in this early study not to redose people who had raised blood pressure. That the one participant had a blood pressure that just tipped the number and then came back down. Another person had more of a sustained -- I'm talking a few hours blood pressure rise. And so I think it's early on, probably most cautious to not redose those subjects. That was 2 of them. And the third subject was somebody who had some suicidal ideation that was moderate in severity a few days after dosing. This is very much in the context of premorbid pre-study suicidal ideation as well. So it was there. But again, out of the abundance of caution, we felt that it was best not to redose that subject. Those are the 3. In terms of discharge...

There was no intent certainly and no behavior. So that's the usual -- that's what is typically exclusionary in these kind of trials, including in depression. Suicidal ideation per say is not normally an exclusionary criteria. Sorry.

And days later, you said?

It was some days later, and it recovered without any hospitalization. It was -- as I say, it was sort of a resurfacing perhaps of some premorbid things. And I think it's someone who was sort of well-known and managed within the system. So not really a concern, but I think our decision was best not to redose that subject. And maybe just then moving to time to discharge. So we don't have the discharge readiness data yet, but what we have got a sort of -- we had a sort of time course for a measure called the SIRs or the subjective intensity rating. So this really ask participants on a scale of 1 to 10, how intense is this experience right now. And our sort of rubric is that by the time you get down to a 1 or a 2, you're kind of back down to baseline. And the average time to baseline was back to go back to baseline was less than 6 hours. So this is a longer duration compound, and we knew that relative to some of the short-acting compounds that we're more familiar with. But I think it's going to be in that 6-hour range when we do get the readiness for discharge assessment, which was conducted, but it's not part of the top line.

Your next question comes from the line of Sumant Kulkarni with Canaccord Genuity.

Nice to see these data. On a later stage program, do you think LSAS is the best primary endpoint at the same time points used in the trial? Or could we see a move to a public speaking test? That's my first question.

It's a really good question. I think there are difficulties with the LSAS. And as I say, part of that is that it's sort of event-driven. So a number of items on the scale, people have to have experienced those events in order for them to rate it on the scale. So it is difficult in that way. But I do think it is more valid than public speaking. So I think in this group, actually, we actually excluded participants who only had performance anxiety. We're really looking for true social anxiety. And I think the thing about performance anxiety is that it is a very -- it's a subtype and it's a very context-specific place. So those public speaking challenges are, I think, very good for an experimental medicine studies that give you an idea. But really, if we're trying to treat this condition, our preference was to use the scale that regulators are familiar with and on which other drugs have been approved.

And just one point there, clarification. So the other companies that have been developing things for SAD or social anxiety disorder have really been focused on acute interventions, right? So they are PRN drugs for when there's a situation you're going to go into that you know is going to trigger your social anxiety. That's the reason that they use this measure of performance anxiety, the social speaking or the social stress test. That is different than the previous approvals, which are based -- they're basically based on the totality of social anxiety disorder. And that is really what we were pursuing here. So it is a bit of an apples and oranges comparison.

Yes. Got it. And then what drove the choice of 2 doses to start with in this trial versus 1 dose?

Yes. I think we had some prior data with racemic MDMA that suggested -- and again, if you look at the treatment trajectory in PTSD that you get further improvement with doses. And in fact, the separation from placebo really begins at dose 2 and perhaps continues to dose 3. So it was a balance between an exploratory trial that was trying to get a sense of whether or not this would work. But I think one dose we felt would be less effective. And I think that's been borne out. So we were keen to give the drug the best chance that it could get while also trying to make sure that this is a fast trial that gave us a readout in short order.

Yes. I think one of the things here is because of this behavioral element to the LSAS, there's this question of whether it's time or doses, right? Do you need more time to actually have some of the symptoms start to be reflected in behaviors? Or do you need more doses to kind of work through these issues? And of course, we don't know that from this one. But yes, going back to this trial, it was really around what's feasible, let's understand this drug a little bit better. We did change 2 variables, so to speak, with this trial. We had a compound that was unique. Again, it's got these properties that seem to split the divide of an pathogen and taxogen and a psychedelic. So that was the first new element. And the second one was the indication, right? So really, no one's looked at a psychedelic in SAD before. So those are the 2 things. We're trying -- we wanted to take a look at both, and I think we did that quite well. But that's why there are some of these questions that are sort of open here.

The next question comes from the line of Patrick Trucchio with H.C. Wainwright.

This is [ Arabella ] on for Patrick. The placebo arm showed a 16.67 LSAS reduction. And so how do you contextualize this relative to historical SAD trials? And then what design features would you consider in later studies to mitigate placebo response? And have you collected any subjective assessments or guess rates from participants or riders?

Yes, good question. So I think certainly in terms of the change from baseline in the placebo group, it's very much in line with other studies. And in some ways, the LSAS is interesting in that it has a relative to, say, major depression, it has a relatively lower placebo response rate. But I think especially in the context of psychedelic trials, we don't want to completely remove the placebo rates because I think that really questions some of the validity of the trial. You would expect that there's an effective trial. It's just more patients get seen, they get seen by health care professionals that are coming in, and you would expect to see that. So I think the question is more about how do we separate from placebo rather than how do we have no placebo response. So I think that's the way we think about this. And so we sort of live with that. The central raters clearly help with that. So I think the central raters are blinded to the visits that the patient is going to. And that means that we have a lower variability with central raters, and we can kind of mitigate some of those potential unblinding issues. In terms of whether or not we would -- we certainly didn't take assessments of unblinding or expectation in the study. And I think it is difficult to really contextualize some of that data. So because we have psychedelics, we do work in this and we think about this a little bit more. But of course, most CNS drugs have some effect that help people guess, whether that's sedation or weight gain or whatever it might be. So I think it's certainly common that people taking CNS medication have some idea of what they're on. But I think, again, the separation here from placebo with a reasonable placebo response. And we did have some patients who had quite high psychedelic rating scales, which also suggests that they were well blinded. And I think that in the context of the way that we designed the trial feels sufficient rather than specifically picking up these or asking patients to guess and then it's always a difficulty to sort of really interpret that data in light of the results.

Great. And then just one follow-up question. The baseline LSAS was quite high. How do you see this translating into more moderate severity patients?

That's a good question. And I think the way that we were thinking about the study going in was that this is a new therapy. It's quite intensive, and it's likely that genericized drugs will be first-line treatment or CBT or something like that, at least initially. And so in order to sort of think about where we are in the treatment line, which would be, say, second line, we probably want a group that is either resistant or more severe and has tried these therapies and have failed. There aren't clear definitions for a treatment-resistant social anxiety disorder, but we can certainly think about making sure that patients both on the CGI severity and the LSAS baseline was sort of unwell enough that they are sort of effectively treated. And I think that's probably where we continue to think about positioning the treatment of the sort of second or third-line drug. So I think that is why we got to this higher baseline as a start.

Yes. And it's an interesting point. I don't think we touched on it in the core presentation, but we noted about just under 50% of folks had been on drugs previously for their social anxiety disorder. Interestingly, almost no one needed to wash off, just speaking to the lack of efficacy or the lack of tolerability or some combination thereof for this patient population of those drugs. So basically, no one was taking anything. No one needed to wash off. And again, so that's, again, speaking to this whole notion that this is probably a treatment-resistant group as Kevin was alluding to. The higher severity is an interesting one because, again, we've been talking about the LSAS as a behavioral scale. Behaviors are hard to change. And if there are severe behaviors, there may be more entrenched. So that is something that we are thinking about and trying to work through and have some discussions or have had some discussions with the KOLs as well. So would it be -- are behaviors that maybe are less entrenched more movable? I mean, conceptually, sure. But it's something that we obviously need to think about as we plan for subsequent studies.

Yes. And maybe just to sort of follow-up on that. I think this is a condition that really begins often in the second decade in people's teens. It's very trait like. And I think the idea of taking a daily therapy for the rest of your life is a daunting prospect for many. And I think actually many people have a sort of an ecological approach to this. In other words, they -- over time, they find a way to live with the condition rather than take daily treatment for it. So I think that's one of the difficulties, in fact, is that in many cases, people have spent years or even decades managing this by avoiding certain situations, and it does take a lot of time for that to change now that they've got a different sort of fear response to the same community.

Your next question comes from the line of Elemer Piros with Lucid Capital Markets.

What I was wondering if whether you have looked at the LSAS scores between the 2 doses or after the first dose, please?

It's preliminary. So yes, I mean, this is our top line. So we will be looking at the totality of the data, including the split between the 2 in subsequent weeks and months.

Okay. And second question I had is on what were the response rates when you look at the LSAS scores?

So the response rates, I have to say, again, we're -- I'd have to go back and have an exact look, but I think they were about 50% higher in the EMP group. And so sort of really tracked, I would say, with the CGI-I. But again, the sort of definition of response with LSAS does take a bit of time. So we're still looking at that data to really understand it better.

Yes. And I think the CGI-I, the way it was represented to us as we discussed with KOLs, was that this is maybe a -- this an early warning, so to speak. The clinicians do get a sense before some of these behaviors are changing. So I think that was also interesting in terms of contextualizing these results. And there's -- I believe there's also some literature out there where the CGI was, in fact, used as a responder criteria as opposed -- like with MADRS or depression, you have a 50%, right? That's the response criteria here. I have seen some literature as well that is actually focused on the CGI as opposed to the LSAS per se.

Your next question comes from the line of Pete Stavropoulos with Cantor Fitzgerald.

Congratulations on the data. I'm not sure if I missed, but you did say 70 participants received at least one dose of study drug, 69 completed Day 43 efficacy assessment. I don't know if you can give any color on how many patients received a single dose. I'm curious to know what the level of response after the first dose and durability after Day 43. Were there any signals that would suggest a single dose would suffice for some patients?

Yes. So as you said, we had 70 dosed and 3 did not receive the second dose. So 67 went through, received the second dose. And again, those -- and remarkably, the 3 that didn't receive the second dose actually continue to the end. So again, really good retention despite not being able to be dosed, we actually got the LSAS CGI scores. So we're still, as Srini mentioned, looking at the Day 29 versus Day 43. The difficulty here is that the scale is one that does take time. People have to go through these experiences. So it's actually difficult to disentangle the -- how rapidly the drug works versus how to change the scale because people have to go out and experience speaking to strangers, having small group meetings, speaking on the phone in public, et cetera, et cetera. So I think there is a time element to the scale, which is unique relative to some of the scales that we're more familiar with, which are very symptom based.

And unfortunately, we don't have data beyond the 6-week primary. That was consistent with the early exploratory study. So we didn't really focus on durability, and we don't have data beyond that.

Okay. And a follow-up question. You reported simultaneous improvement in both fear and avoidance subdomains. So how does standard of care affect these subdomains? And if both are improved, are they improved simultaneously? And do you view this as evidence that EMP-01 is influencing core learning or behavioral distinction mechanisms rather than simply anxiety?

Yes. So I think, again, the effect of approved pharmacotherapies are similar. And I think the idea is that the fear experience goes down first, people then experience these things, which they are typically fearful of and avoidance of, and it isn't as bad as it was before. It's less threatening. And so that they then gradually learn and they do more of those sorts of things. And that's actually very similar to the way people think potentially SSRIs might work that the environment is less threatening and you can then go out into that environment. And gradually over time, you interact with that environment differently. So I think that's exactly what we are seeing here, and I think how we think this is going to work People perceive the world as a less threatening place, and that changes behavior over time. But it takes time for that behavior to consolidate and people to have repeat experiences that suggest that it's actually not so bad and that you can go out and do these things again. And there are clearly within the scale, some low hurdles, as I say, like speaking on the phone or speaking to people you don't know very well. And then there are sort of higher hurdles like speaking in public, arranging parties, things like that, which I do think would take longer to really shift.

So here with this trial, I mean, again, it was a short study. Typically, with an SSRI, SNRI, you're seeing fear move first, as Kevin said, and then the behavioral stuff, the avoidance actually catches up over time. With this small study at 6 weeks, we saw both move sort of in parallel, kind of more in keeping with something like a therapy, but again, small study, something that we need to replicate.

Your next question comes from the line of Eddie Hickman with Guggenheim Securities.

Congrats on the data. Just 2 for me. How confident do you feel about this dose and dosing interval? Would you think you need to try others in the subsequent study? And given the light dose experience, how are you thinking about therapy versus no therapy in the Phase III? Have you had any preliminary FDA dialogue on this?

Thank you. So I think on the first one, the dosing interval is something that we may want to explore. I think this is a -- again, there's some strong priors with racemic MDMA that suggests this is not a bad separation, but it could potentially be expanded or contracted. It is a longer stay in the clinic. And I do think that that -- if you push it if it's every 2 weeks, there's a sort of an increased burden and scheduling issue there. So there is a sort of sweet spot to find on that. So I think that's something that we're still exploring. And forgive me, the second part of your question, could you repeat it again?

Just given the light dose experience, how are you thinking about therapy versus no therapy in a pivotal? Have you had any preliminary FDA dialogue here?

Very good. Yes. So I think we have not had the preliminary FDA dialogue. But clearly, we are keen to think about this in a no-therapy context, if possible. And that is an access issue as well. So I think as soon as it becomes something that is bound to therapy, we are also bound to the availability of therapists. That being said, I think it's an interesting -- with all psychedelics, I have no doubt that in the clinical setting, there will be people who are treated with both pharmacotherapy and psychotherapy, and I'm sure that there will be an additive effect there. But I think our current intent is to think about this as much as possible without a therapy component. And there are various gradations of that. There are certainly drugs like weight loss drugs or drugs in the treatment of substance [indiscernible], which are given in the context of underlying therapy. But I wouldn't imagine that we are pairing this very intentionally with the kind of therapy as part of a trial.

Your next question comes from the line of Ami Fadia with Needham & Company.

Congrats on the progress and thanks for sharing the data. My question is just sort of a follow-up on how to think about the evolution of sort of the response over the course of the 43 days. Did you see separation after the first dose? And just based on this data, how would you think about a future trial in terms of the number of doses you would explore? Could there be benefit from additional dosing after the 2 doses here? And then secondly, just on the statistical significance, I think you indicated it was a one-tailed p-value. Why did you pick a one-tailed test for this study? And then what would a two-tailed test have looked like?

Sure. Maybe I can address the first part about the number of doses. I mean I think the next phase would definitely require some dose ranging. And I could imagine dose ranging both in terms of the amount per dose and the number of doses. And I think we can think about those 2 axes. And there's both a follow-up period that would -- is 3 doses better than 2. There's clearly some aim also to achieve a meaningful benefit without dosing for 6 months. So I think there's obviously some guardrails around that. But I think those are the thoughts that we're having right now about what does the dose ranging look like? Is that number of doses and/or the amount per dose. So that's the way we're thinking about that at the moment. And again, forgive me the second part of your question?

The second part was on the p-value ones. Why did you take a one-tailed test? And how should we interpret that? And perhaps if you could tell us what a two-tailed would have looked like?

Yes. Sure. Yes. And I think we were again quite intentional here about this being an exploratory study that wasn't powered. We took the view that if it was something that was going to get worse, we would not move this forward. So hence, the choice for a one-tailed test, we're only looking for improvements really because a worsening would have clearly led to no further development of the compound. So we're really focused on that improvement tail, if you will. And again, because it's unpowered, it was a small study without any sense of what the treatment effect might be, we chose that as a level. And again, I mean, one-sided p-value is essentially half of a two-sided p-value. So if you multiply that by 2, you would get a 2-test pitch, which was, I think, 0.07.

Yes. I think it's kind of -- let's go back to that first part of your question in terms of number of doses, et cetera. I mean, we are aware, obviously, that this has some -- this is a behavioral kind of endpoint. It's -- the question is whether it's time or doses again, right? So it's not symptomatic improvement like depression or what is measured with the MADRS, for example, where it does move quickly. So that's something that we really do need to spend some time thinking through. And again, that kind of leads into the exploratory nature of the trial and the fact that we are only looking at improvement essentially. So not that unusual in small studies where they are unpowered because there's no real data to properly power them to do a one-sided test, particularly in that context of the time element that we weren't fully -- we didn't fully understand going into it.

Your next question comes from the line of Jay Olson with Oppenheimer.

Congrats on these impressive results. It's really exceptional to hit stat sig on such a small study and the effect size and especially the NNT is super impressive.

Yes, yes, thank you.

Yes, thank you.

We had a couple of quick questions. Was there a psychotherapy given to the patients in the study? And were any patients enrolled in the study psychedelic experienced at baseline? And then maybe a bigger picture question. Can you please remind us the differential diagnosis for SAD versus GAD? And any read across from these results in SAD to a GAD population?

Yes. Sure. And I'm going to start backwards and maybe we can kind of go from there. But clearly, there is some comorbidity. And one can make a specific diagnosis of social anxiety disorder. It typically occurs earlier. It has a very kind of social context. So it's context-specific. So it does look different to GAD. But of course, like many psychiatric conditions, there is comorbidity and people are anxious and they can have episodes of anxiety disorder. They can also have episodes of major depression. So I think the definition of the case is important here. We certainly required medical records that were consistent with the social anxiety disorder diagnosis. But of course, it is underdiagnosed in the area. So a lot of people seldom actually make it to tertiary care with the sort of treatment seeking for social anxiety disorder. Many cases, they have an episode of something that presents and they also -- the social anxiety is discovered during that intake. So the way that we approached that was that we required medical records and a treatment history that was consistent with SAD and we did a thorough assessment using standardized structured interviews at screening to make sure that they did indeed meet the criteria. But I think that's a key piece going forward as well is how we define this group and the evidence that we need. I think for the validity of this, we really do need some sort of documentary medical records.

Yes. And I think that going back to your point around how does it differentiate from GAD. You can think of it as -- I mean, I sort of conceptualize GAD as almost an extension of MDD, right? I mean in both cases, there's a prominent rumination component. There's particularly a negative rumination or a negative -- the self-referential thoughts Depression, you're kind of looking backwards. Anxiety, you're kind of looking forwards. And these 2 are very comorbid. But then there's the other -- there's a whole bunch of other things that are more phobias, right? They're kind of more specific to a single thing. I mean where everybody think knows about arachnophobia or something. Yes, you can sort of live with that. I mean you avoid spiders and you're probably okay. But this is a phobia almost against social interactions, right? That is different than GAD where it's just this unnerving kind of sensation that something bad is going to happen. This is more specific, you're avoidance of social interaction. So it's got that phobia-like element to it, but it is kind of clearly differentiated. And that's what the LSAS and all these structured interviews are also getting at. Like is it this nonspecific thing that the sense of impending doom, which is kind of GAD? Or is it you really don't want to do these things like talk to someone you don't know, interact with other people, blah, blah, blah. That's what differentiates these 2. I know you had a question also about therapy. No, we didn't do therapy in this trial. We don't wish to it. Again, as Kevin outlined, that isn't something that we really want to touch on going forward either.

And just to elaborate, we took the standard participant support approach. So we had some preparation with participants. There was a facilitator who was in the room during dosing with the second one monitoring from outside the room. And then we had a follow-up assessment that was done after the dosing day. So it's a very standard, very focused on education support during the dosing. And again, what we found was that this is a very different compound with a lot less of the challenging experiences. So some people were sort of more talkative. Many were quite internal in their experience, and they described it almost as dream like. So that part, in fact, was particularly easy relative to some of the other drugs that we've experienced.

Yes. I think we didn't touch upon it in the core presentation again, but there were some elements that were kind of unique around the anxiety or anxiolysis that seems to be -- that's occurring with this compound. Again, we need to do follow-up studies and particularly comparative studies to really flush this out. But feeling of relaxation is not something you normally see with other psychedelics, in my opinion. I mean that's just not something that's normally seen. It was prominent with this -- in this patient population. Is it the patient population? Or is it the drug? That's something that we need to assess out. But that was something that was really interesting. And again, there was normally -- I mean, there was a subset of individuals that was quite talkative during this process, kind of like MDMA and not typically like a normal psychedelic. I mean people that are taking psilocybin, which is similar in duration to this compound are not normally talkative unless things are going sideways, but that wasn't necessarily the case here. So it was different in these elements. And then there was this vigilance reduction, which I think was very prominent here. I mean, again, think about this, we have someone who has social anxiety that is stuck in a room with someone they don't really know very well, and they're comfortable with that. And some of them are actually talking to that person, which is kind of antithetical to their nature essentially.

Your next question comes from the line of Harry Gillis with Berenberg.

Congrats on the data. Could you help us contextualize the 12-point placebo-adjusted improvement in LSAS at 6 weeks, I guess, in terms of what this means for patients in the real world and perhaps how it compares to what you were hoping for heading into the trial? And then you've also clearly described this as clinically meaningful and supporting further development. Just curious what would have been the minimum threshold that you thought would have supported further development or broadly what that level might have been?

Sure. So I think in terms of the contextualizing of the clinical response, again, this sits very squarely in the -- similar to existing pharmacotherapies. And actually, I think the CGI in conjunction with the LSAS was most compelling. So both the response rates and a CGI of 2 is very much improved. It's quite a high bar. And I think that sort of gave us comfort that it wasn't just the LSAS, but these numbers are very good. I mean if you look at the numbers needed to treat for CGI relative to the existing pharmacotherapies, they are better or comparable depending on the drug in question. So I do think that together, we're sort of encouraged by that. And I think it is a clinically meaningful benefit. I think in terms of what might be a no-go for this, it would have been twofold, and we actually had safety criteria and efficacy criteria. But clearly, something that is worse than the standard of care would have been a problem and I think would have led us to sort of stop the program. We had -- there's a range for LSAS, there's a range for CGI. And they were very focused on those things, both in terms of responders and absolute change. And the absolute change in LSAS across the sort of pharmacotherapies is about 10 to 15 placebo subjective, so 10 to 15 points placebo subtractive difference. So we were in the middle of that quite squarely. I think if we found a 5-point difference is something that wouldn't have been viable, especially if the CGI matched that. And I think that was for us the kind of the thing that tied it up really. And similarly for safety, we had that early Phase I study that suggested it was fairly well tolerated. It was a fairly pleasant experience, reliably psychedelic, but not aversive. But of course, in a social anxiety population, we're taking this into a very different group. So Phase I participants tend to be fairly okay with risk in general, whereas I think this group is a sort of an anxious and worried population. But again, we saw a very consistent sort of signal there that looked like the Phase I data. So I think both of those potentially could have been reasons to not regret.

Yes. And just contextualizing further, I mean, we saw a similar efficacy at 6 weeks of 2 doses, right, as opposed to 8 to 12 weeks with daily dosing. That is something that really resonated with the KOLs that we talked to. And the other element here is because it's behavior, there was a lot of conjecture what would have happened if we had run the trial out longer, right? Even if with these 2 doses, if you just left them alone afterwards, that allowed them to experience these new things. And we got anecdotes on that, like the folks are saying, Oh, this happened to me yesterday, I would not normally have been able to do it. And that kind of success builds upon itself, and it just takes time. So there was a lot of discussion about if we had -- if we had just looked at another endpoint, not even dosing, but looked at an endpoint, say, 8 weeks or 12 weeks out, could this have continued? In other words, is this a plateau? Or is this kind of the start of something? So these are some important questions. And that's what goes along with the CGI are all reason to believe here.

And just one follow-up. I mean, one of the things that we think about in terms of the safety is really the prevalence of the side effects. So in our case, we have side effects like hallucinations, feelings of relaxation, headache, but they are limited to 2 days out of 6 weeks, whereas others may have maybe fewer nausea, weight gain, concentration difficulties, but they do persist for the duration of that 8 to 12 weeks. So I do think from a safety perspective, we see this as sort of if you look at the prevalence, how long are you going to suffer with this adverse event or side effect. We have a real differentiated here. People have mostly drug-free with no side effects in between those dosing days.

That's really clear. And could I just squeeze in one more. Just based on the data that we've seen today, do you think 225 milligrams is the highest dose you anticipate giving sort of at a single point in time? And then it's -- I know you then discussed it's about the amount per dose. Is it then lower than that? Are the other options and then the frequency? Or could you even go higher than 225? Or should we think of that as a ceiling?

Yes, that's a good question, one that we've been thinking about ourselves. I mean R-MDMA has been studied up to 300 milligrams. I think there are changes in vital signs like with all of these drugs, you get increased blood pressure. And there is a balance and especially if we want to redose. So I think 225 is approaching a limit. I mean we'll certainly think about higher doses. But I mean, we've got a good effect here. And I think that it was tolerable. So I'm not sure that we would necessarily go up higher because I think there's -- the higher you go, the more patients may not be able to receive further doses.

Our next question comes from the line of Michael Okunewitch with Maxim Group.

Congrats on what looks like really great data.

Thank you.

I think just to start off, I would like to expand a little bit on some of the prior questions on schedule. And I guess, how you're thinking about how the drug could be used in the clinic because there are quite a few considerations here. So is this something you're thinking more along the lines of an induction regimen followed by an untreated follow-up or potentially a more Spravato like schedule where there is some maintenance dosing over a study period? I'd like to get your thoughts on that.

Yes, it's a really good question. And I do think for this, we really need some durability data.

Yes.

So the next step would be, say, 2 doses and then a follow-up that goes out for 10 to 12 weeks. So we can have a look at that response trajectory, both in terms of the average. But of course, what we're really interested there is the subgroups. There's typically subgroups who get well and stay well, and subgroups who get well and then gradually begin to sort of get worse again. And that really helps us inform how frequently. I don't expect this will ever be Spravato like and that it's once every 2 weeks, even once every month, I think would be challenging. But if this is 2 doses and then it's a quarterly follow-up, maybe a sort of second or a third dose, one dose quarterly, that I think is manageable. So to be determined once we've got some more data, but that's where -- I mean, again, looking at the other compounds in the psychedelic class, I think that's what we sort of expect to see.

I think that does lead into my follow-up question a little bit just on the clinical strategy going forward. There has been a lot of difficulty in SAD. But as you mentioned, there are still a lot of answers that you kind of want to understand. So is the strategy here more to go into a Phase IIb for a larger data set, longer follow-up and to really understand these questions rather than moving potentially faster given the unmet need in SAD?

I mean we'll provide more guidance on that. This is -- again, this is top line results and very encouraging top line results. We have some additional data sets coming from this. I mean this is -- the topline results show a subset of the -- all the analyses that need to be done. So we'll make some decisions once we have all that data in hand.

And our last question comes from the line of Justin Walsh with Jones.

This is [indiscernible] on for Justin. Congrats on today's update.

Thank you.

We have a further down the line sort of question. MDMA is a Schedule I controlled substance. So we are curious about the company's current strategy for EMP-01's scheduling and how are you preparing for it.

Yes, that's a good question. And of course, we -- it's for us, a bit of a well-trodden path now with BPL-003 and VLS-01. So all ends...

And for the specter, right, psilocybin and LSD also in development. Yes. Yes.

So I think we're approaching this at the moment by collecting adverse events of abuse potential. And you saw that list there, which is really a list that's -- although we haven't had this discussion with this compound of the regulators, we certainly have on others. And there's a clear list that I think we need to characterize the abuse potential of this drug. So currently, it's Schedule I, and that really means that there is no therapeutic potential. But of course, should we get ourselves through pivotal trials and this goes to drug approval, it then, by definition, has a treatment or it has a therapeutic use, and it will move out of Schedule I to Schedule II or Schedule III. So I think there's no way that this drug has no abuse potential. I don't think it's realistic to say that we're going to come out with no scheduling, but it will certainly be in a reasonable scheduling along the lines of Spravato, which had a similar course to the approval process. And that process really means that you present your dossier of abuse potential to the regulators. And in the U.S. that then following approval move to the DEA. The DEA make a decision on scheduling and you move through from there. So I think that's an approach that we will take, and we're beginning to gather that data even through the study.

Okay. For our follow-up question, the FDA stated recently that they will now require only one pivotal trial plus confirmatory evidence for new drug approvals. How does this change your late stage development plan due to EMP-01 drug category? Would you now consider only one pivotal trial for a faster path? Or would you plan to do 2 pivotals to provide a more robust MDMA package?

Yes. So this has come up a fair amount. I mean the issue typically with one trial is you still need safety data, right? You need a database in a chronic indication, right? So depression, social anxiety disorder, generalized anxiety disorder, these are all chronic indications that will require redosing as we've been discussing. So in that context, you typically do need more exposures. For an example here, if you want to find a 1% adverse event rate, you need to dose 300 people and 300 placebos, I mean broadly speaking, to have a comparison -- a comparative data set. With acute studies, that's not -- I mean, with acute indications, that's not necessarily the case. It's a little bit different in that context. One trial approval is certainly a possibility there. And indeed, one of the other companies in this space that's looking at PPD recently announced that trial approval was a possibility for them makes sense because PPD or postpartum depression, sorry, is an acute indication. It only occurs in that postpartum period, you're treated and it's done. So that's acute. So -- and as a corollary to this, you don't normally want to do everything in one very large study. You'd rather split it into 2 smaller studies to get different data from those 2 smaller studies. So that's a general fact that I think most of the companies have been taking. Our drugs are a little bit more different. I mean, we'd like to think of a second generation. The -- in the case of BPL and VLS, you can't really get those pharmacokinetics in any kind of normal approach. There's really no source of R-MDMA either. It's very -- it actually is very difficult to produce. You can't just entirely separate it like you normally can with other compounds. So they are different. So I could easily see the requirement for more safety database because of that.

There are no further questions at this time. On behalf of AtaiBeckley, we thank you for joining today. That concludes today's conference call. You may now disconnect.