AtaiBeckley Inc. (ATAI) Earnings Call Transcript & Summary

Good morning, everybody, and welcome to Atai's webinar focused on the Cognitive Impairment Associated with Schizophrenia. I'd like to thank everybody for joining both our speakers, Matt and Rich as well as the audience members. The goals of today's call or webinar are really twofold. The first is to provide some -- to provide a little more clarity, a little bit more color on the results that we announced last year -- late last year on RL-007 and the initial proof of mechanism trial of this compound in schizophrenia and cognitive impairment associated with schizophrenia in particular. We talked a little bit about the EEG results that we found there as well as some of the symptomatic benefits that we're seeing. We, again, want to give you a little bit more color there and also provide a touch upon some of the ERP results that were down there. What's equally important, however, is that we want to provide some context for these results and really give the audience a little bit more information on the background of this compound. So this is an interesting compound. There's a lot of human experience with this particular agent. There's a lot of data around cognitive improvements. However, there is no publicly available data. And I know that's been a source of frustration to several folks within the audience. So again, I hope to provide you a little bit more context there. Let's go ahead and advance the next slide. Perfect. These are our standard forward-looking statements. Let's go ahead and get to the next slide. So this is a quick overview of today's agenda. Again, I'll be providing a little bit more by way of introduction. Both of Atai as well as, of course, a little bit more on our speakers. Rich is going to jump in and give you a nice overview of CIAS broadly, the background as well as some of the therapeutic landscape, which, of course, at this moment is quite limited. Matt's going to be going through the background on RL-007, starting with the background, all of the existing data that's been generated and then go into some of our Phase IIa results and then, of course, we'll dive into a Q&A and answer any questions that any of you may have on the compound. Next slide, please. All right. So I think many of you are familiar with this. I mean clearly, mental health disorders really become one of the greatest global health challenges. There's about 1 billion people, that's with a B, impacted by mental health disorders. Lots of unmet patient need and lots of -- I mean, the treatments and innovations have been quite limited. And you'll see from Rich that this is particularly true for cognitive impairment associated with schizophrenia, with we are comfortable in many -- for the most part in treating of positive symptoms of schizophrenia. But you'll see that the impact of the cognitive impairments negative symptoms, et cetera, are really what drives a lot of the morbidity associated with this condition. We are really focused on compounds that have prior evidence in humans. And in some situations like DMT, et cetera, those are published data. In this case, of course, those data are not published. As I mentioned earlier, Matt's really going to be focusing on that and giving you the context around the -- why we are interested in this compound in the first place. I'm going to talk a little bit more -- I'm going to at least flash out the slide, I should say about our -- with our entire pipeline. We also have a number of enabling technologies broadly, these kind of fall into a couple of different buckets. We have digital technologies. We have formulation technologies, and we have now several drug discovery capabilities, house in different companies as well. So really interesting kind of orthogonal technologies. We have mentioned a number of times we're not interested in the excess [indiscernible] So we're really interested in things that truly move the needle for mental health treatment. And obviously, we are very excited about the RL-007. And as I mentioned, really, I do believe that the results that we've found today are significant derisking events for this particular therapy. Many of you are familiar with the very -- the unique model in many ways unique model over Atai. Obviously, we have central Atai itself or most of the capabilities in many different realms are housed. And then we have a number of subsidiary companies, of course, Recognify being one of those. And in general, all of these subsidiary companies leverage the capabilities of a tide, which, of course, are, as I mentioned, quite broad, obviously, R&D and again dragging and dropping, if you will, as necessary. But also broader capabilities from intellectual property to general legal to finance and accounting. We've had some recent successes in terms of validation of the model, of course, many -- I think all of you are going to be familiar with the Compass Pathways IPO and the Otsuka collaboration that was announced early last year. At this point, we're about 100 FTEs internally. We have many consultants beyond. And our cash balance, the last indicated cash balance is $430 million. Okay. So let's go to the next slide. So as I mentioned, this is just a quick overview of our pipeline. I'm not going to go into any details here. As you can see, RL-007 is one of the furthest assets, one of the most advanced assets that we have. Obviously, we'll be talking about the results of our Phase IIa proof of mechanism trial. And of course, as I've mentioned, as Matt will also highlight we will be moving into a double-blind placebo-controlled Phase II trial of this compound in the not-too-distant future. All right. So with that, let me jump -- go to the next slide. Let me just give you a little bit of background on our speakers. Rich is someone I've known for quite some time. He's a professor of psychology psychiatry and neuroscience at Duke. Salient to the current discussion, he is an expert on measuring symptoms and in fact, was the founder in 2004, I believe, of NeuroCog. And there was a sort of central mandate there around instruments and around validation of instruments for assessing symptoms, again, broadly, but certainly schizophrenia -- in fact, cognitive impairment in schizophrenia is where I met with in the previous program. So again, very happy to have him on board as I mean on this talking more broadly involved with our program. As I mentioned, Rich is going to give you a little bit more context around the indication itself. Matt is the CEO, President and Co-Founder, of Recognify. He was instrumental, along with a couple of other folks on the team in getting this very exciting, very interesting asset, essentially out of Allergan and into the hands of Recognify as Atai previously was the entrepreneur in residence at Foresight Ventures. Prior to that, CSO Diaxon health, and has a PhD out of Salk here in San Diego. All right. So next slide, I believe I'm going to be handing off with this point to Rich, who will be giving you an overview, as I mentioned, of CIAS.

Great. Thanks, Randy. I appreciate it, and it's a great opportunity to be here to talk to everybody about cognitive of impairment in schizophrenia going to talk a little bit about the background about why cognition is important why this is an unmet need and some -- a bit about the current programs in place to try to do something to address this really incredible devastating aspect of mental illness. Next. This is my disclosure. As Srini mentioned, in addition to being a professor, I have a company outside of Duke that is called VeraSci, and we support clinical trials, including some of the ones that I'm going to be discussing today. Next. So people probably know a bit about schizophrenia. They, like many psychiatrists, often think of schizophrenia in terms of the delusions and hallucinations. Paranoid ideas that martians are taking over their brains, hearing voices and these very striking aspects of the illness that really get a lot of attention from the media. And also, for obvious reasons, because they're dangerous, have been the primary target of treatment, the so-called antipsychotics. There have been antipsychotics available now for 70 years. And yet the disability associated with schizophrenia has really not changed much over the course of that 70 years. And schizophrenia is one of the most disabling illnesses of young people in the world, only behind alcohol use disorder and depression. Next. So if we look at some of the statistics associated with that disability 2/3 of people with schizophrenia never have any intimate relationships that leads to marriage. They -- often in young age, after psychosis going forward when everybody else is developing relationships, kind of planting their stake in the productivity that they're going to have, people with schizophrenia are unstable and don't get that kind of traction in developing the kind of things that make us all productive members of society, well, most of us. The other thing -- and as a result, by the way, they are very unlikely to work. Less than 15% of people in the United States with schizophrenia have a job and 20% of them are homeless at any time. So if you drive around the cities of America, you often will see people living under the bridge, living in terrible conditions, most of those folks have schizophrenia. So unbelievable devastation associated with this illness. And it turns out that this devastation actually isn't caused by the delusions and the crazy ideas and the hallucinations, actually caused by the cognitive impairment. So I'm going to talk a little bit about what that cognitive impairment is and how it leads to this dysfunction among people with schizophrenia. Next slide. Yes, first, the cost associated with schizophrenia. So these estimates are a bit old, about 8 years old. But at that time, and things have only gotten worse, the estimated cost of schizophrenia in the United States alone was $156 billion. Some of that was from direct health care costs, but a lot of it was from indirect costs, up to $120 billion or so due to the fact that people with schizophrenia for the entire course of their lives, right, onset is often 18 to 25 through to death. They don't work. People need to not work to take care of them. And they cause other -- they have other health care problems associated with their illness. Of course, if you're living under the bridge, you're probably not getting your annual check up, not taking the other medications, you need to treat heart conditions and cancer and things like that. So it's a tremendously devastating illness with real costs to the United States and to economies around the world. Next. So this is the severity of cognitive impairment in schizophrenia. And I need to apologize to those of you who are not psychologist, but there are going to be a couple of statistics that I need you to understand, very simple. First is T-score. So a T-score is a statistic that psychologists use, where the mean in the general population is 50 and the standard deviation is 10. So pretty simple. Average person in the general population on these cognitive tasks gets a 50. People who are doing really well, get a 70. People who are really doing really poorly, get a 30. When you look at these aspects of cognition that are particularly impaired in people with schizophrenia, they perform about 1 to 2 standard deviations below the healthy control means. So T-score is ranging between 40 and 30. But if you look at the overall sum of all of these aspects of cognitive impairment, the T-score is 25. So that's 2.5 standard deviations below the healthy control mean. And if you put that into metrics that might be a little bit more familiar, it's the lowest 1% of the population. In IQ score terms, it would be lower than 70, which is the cutoff for pervasive developmental disorder or intellectual disability. So these are really profound levels of cognitive impairment in people with schizophrenia. Next. And as I mentioned before, these cognitive deficits, more than any other aspect of the illness, account for the difficulties that patients have in terms of getting a job, keeping relationships, functioning in the community, being able to even benefit from the psychosocial rehabilitation programs that are available to them. Next. So important question, certainly, from the perspective of what is the unmet need? Is -- well, which patients with schizophrenia have this cognitive impairment, CIAS, cognitive impairment associated with schizophrenia. Our estimate is that almost all patients have some level of cognitive impairment compared to what they would have been if they had never developed illness. So if we look at these scores here, and again, these are a different metric. This is actually an IQ type score, where the mean is 100, standard deviation is 15. So you can see right in the middle there between the light blue 25% is 100. That's the average of the general population. You can see there's a nice inverted U-shaped curve from the healthy group. Same with the schizophrenia group. It's just shifted about 2 standard deviations down from the healthy control. So you think, well, what about those -- that 2.2% there that appears to have above average performance on cognitive tests? Okay, they're performing cross-sectionally. At the current time, they are performing in the range of what's considered normal in the healthy population. However, people have probably heard of somebody who is at an Ivy League School or who was in medical school and had a psychotic break, never recovered, developed schizophrenia and the rest of their life stayed in that same state, difficult to get a job, difficult to get married. Those folks almost certainly performed cognitively much better prior to the onset of their illness. So do they have cognitive impairment, I think so, right? Somebody who loses that amount of cognitive function is doing much worse than they would have otherwise. And they are just as likely to benefit from treatment as somebody who is at the lowest end of the spectrum. So when the question is, does this apply to a certain portion of patients with schizophrenia? Our answer is that it applies to almost everybody. So you can just count up the number of people with schizophrenia, and those are the people who could benefit from a new drug. Next. But yet there's nothing approved. So I believe the current total is 27 antipsychotic medications are approved for the treatment of schizophrenia. 27, all of them basically doing the same thing, blocking dopaminergic neurons in order to reduce hallucinations and delusions. And that's it. They do nothing for cognition. They do nothing for the other aspects of the illness like negative symptoms, not interest, lack of motivation, lack of engagement with the world. So why not. There are about 200 CIAS trials listed on clinicaltrials.gov. However, a lot of those are early phase single-site studies. A lot of those are academic studies. And some of those aren't even with a pharmacologic intervention. Some of them are with like psychological interventions. So there are no treatments. And you wonder why hasn't there been any success? And largely, it's from shots on goal. When you compare the clinical trial -- number of clinical trials, the investment from the government as well as the investment community in drugs to improve CIAS is very small compared to other aspects of medical illness like cancer and cardiovascular disease. NIH spends 25x more on cancer than schizophrenia. And of course, as we know, the industry investment in clinical trials dwarfs what the government does. There are over 1,000 ongoing clinical trials in cancer for everyone in CIAS. And so sometimes people see that and they think, "well, that's because it's easier to get a drug approved in cancer than in psychiatry". Not true at all. In fact, if you look at the latest statistics, the progress from Phase I to FDA approval is 5.1% for cancer indications and for psychiatry, It's 6.2%. So why? It's a little unclear to me. I think some of the reasons have to do with stigma, bias. I think everybody knows somebody with cancer with cardiovascular disease or relative, even Alzheimer's disease. And so it's much easier for investment to be placed in that area, easier for senators and congressmen to approve the allocation of resources there, of course, because a lot of them are in the age of risk of cardiovascular disease and Alzheimer's disease, but not so much for schizophrenia. So it's a stigmatized disorder. And I think that means that in a way, it's underinvested, right? Because if it's a stigma associated with it, and there's just as much likely of getting success, why wouldn't there be as much opportunity from the investment community. Next. So there had been some recent positive developments. I think that everybody should be very heartened by. FDA, the current psychiatry division in FDA, has clearly recognized the severity of the unmet need. The whole matrix program that started a number of years ago, has built up a methodology that is very clear. mechanisms of action that are successful, such as this one have been emerging and FDA is aware of that. And so they have been expediting some of the CIAS drug development programs. So the potential economic gain to all of us for a successful treatment in CIAS is enormous and we know it will be welcomed by the government folks who are involved in making decisions about whether a drug can come to market. Next. So here are a few of the programs that are currently in place to kind of set a little bit of the context for the Recognify compound. So the biggest program -- actually, the biggest program in history so far is the Boehringer Ingelheim GlyT1 inhibitor, a glutamatergic drug. They had -- I'll show the results in just a minute, some successful Phase II results and the Phase III program is now underway. Biogen has a positive allosteric modulator, an AMPA compound, which also has glutamatergic effects. And they are currently in the midst of a Phase II program. That was a compound by the way, that they licensed from Pfizer a number of years ago when Pfizer got out of CNS altogether. Neurocrine, Takeda, not much about this trial is publicly available. So what I can tell you is from the public press releases, they had the original Phase IIa clinical trial was focused on negative symptoms. They also measured cognition in two ways, which is important. They measure cognition as a performance outcome, cognitive tests that we do. They also asked questions to the relatives about, is this patient getting better on the drug? They got positive signals on both of those endpoints, even though they were both secondary endpoints. And so now they're going to go back and start a study, a Phase IIb trial that focuses on cognition as the primary endpoint. So they've changed the endpoint, whether that signal is real, is going to be determined later. Those are the major programs where cognition is the primary endpoint. The other aspect of this in terms of cognition is cognition as a secondary endpoint. So I mentioned that all 27 of those antipsychotics depend upon dopamine blockade as a way of reducing psychotic symptoms. There are some new programs that people are probably aware of, where dopamine blockade is not the mechanism of action, but other mechanisms of action for these antipsychotics. It may be that those also have benefits to cognition as a secondary endpoint. So that whole field is opening up a bit, but not much is known about the effects of those drugs on cognition. It's going to be a few years before we find out. The important thing about these mechanisms, though, to kind of set the stage a bit for what Matt is going to talk about is that many of these programs focus on very specific selective mechanisms of actions. What's interesting about Recognify is it has a much more complex mechanism of action that modulates a variety of different receptor systems. The reason -- one of the reasons that that's promising is that other drugs for schizophrenia have also had a complex receptor system activation. The best example is clozapine. Clozapine, while it has this untoward side effect, potentially a fatal side effect, it has greater efficacy than any other antipsychotic drug. And that's largely seen as being associated with the multi-receptor profile that clozapine has. Next. So final slide for me. This is just data from the Boehringer Ingelheim program, and I wanted to make one point about this. The blue dotted line, that's the dose that was the best compared to -- in the black and placebo. It was about a 2-point change on the matrix battery, the FDA endpoint. Patients with schizophrenia perform about 25 points lower than healthy control mean. So that is a very small bump in improvement in cognitive performance. This drug is not curing the cognitive impairment of schizophrenia. If this result continues on into the Phase III program, I believe it will be approved. If they also have their co-primary measure improves as well. I believe that will be approved whenever the program is done in about 3 years. But it is unlikely if this result is replicated and it's approved that this is curing cognitive impairment of schizophrenia. This is 10% of the improvement. And so there will be a need for other drugs to do something as good or better in the future. So thanks again, Srini, for the opportunity to talk to everybody about cognitive impairment in schizophrenia. And during the question-and-answer period, I'm happy to address any questions people may have. Thanks.

Rich, thank you so much for that amazing overview and really, again, kind of contextualizing what we're doing in the broader space of cognitive impairment associated with schizophrenia and the other therapies. At this point, I want to hand it over to Matt to kind of give you a little bit more about the drug itself and the results that we found. Matt?

Thank you, Rich, and thank you, Srini, for the introduction and the great context of the huge medical need that we're trying to address here. So next slide, please. So I'd like to just start by calling out our combined clinical team of both for Recognify members and Atai members that have really come together in a strong way and in a complementary way to really push this program forward. Particularly, I'd like to call out Gary Walker, who's been doing -- who's a co-founder with me at Recognify, who's really been doing the heavy lifting during this initial clinical study in CIAS. Next slide, please. So obviously, Rich and Srini touched on this and what I'll be telling you about today is a unique neuromodulator that has demonstrated clear clinical -- clinically significant changes in learning in memory and cognition in general. And what I want to try to do here and as Srini alluded to, not a lot of data has been published -- actually, very little close to none, around this compound and really give you some context of what some of the underlying data is and partly what really attracted Atai to this product and really its clinical benefit that's been demonstrated already. And how we're extending those observations into CIAS to demonstrate CNS activity with EEG ERP readouts, but also looking at and gaining clinical response on cognitive measures. Next slide, please. So we view this product as a de-risked pharma developed product. We were able to get it into Recognify with an impressive preclinical and clinical data package. It comes with multiple replications of statistically significant procognitive signals and learning and memory. This is really important because learning and memory defects are a central component of CIAS patients and how their lives are impacted by the disease that they're suffering from. And also this product is really well tolerated. And this is also important when you're taking a product into a fragile population that is already struggling with their morbidities from the disease. Next slide, please. So just to give you some context of why we're calling de-risked product, part of it consists from the -- has a very consistent PK-PD exposure response relationship. And why this is important is it gives us confidence in the active dose range that we're investigating and how we can translate that both from the preclinical space to the clinical space and ultimately to CIAS patients. It has a complete CMC package, so it's ready to move forward and further into clinical development. And as I mentioned, it has excellent safety and tolerability and multiple clinical cognitive signals. Next slide, please. Obviously, the product has been tested preclinically quite broadly, and this is mainly because of its core mechanism of action of enhancing neuroplasticity. And this neuroplastic effect that the compound has translates into procognitive effects, antilotic effects and other functional recovery effects have been demonstrated both in vitro and in vivo in the preclinical space. Also importantly, the product is nonsedating and there's no dose-limiting CNS tolerability effects. Next slide, please. So the product not only has a unique mechanism, but it was actually discovered in a somewhat unique way, and that was through an in vivo phenotypic screen. And the product has mainly been -- the mechanism of the product has mainly been characterized through electric physiological changes. And what's been demonstrated is that the product does enhance basal glutamatergic tone and enhanced LTP. And this has demonstrated -- some example data is on the upper right-hand panel, where you see an increase in basal activity in the presence of 007 and upon stimulation a significant enhancement of long-term potentiation. And you'll see that if you look closely, the blue dots actually represent an intermediate dose, which elicits the largest effect. And this is indicative of the bell-shaped dose response of the product across both in vitro and in vivo models. Another interesting aspect of the compound is that it appears to be able to enhance long-term potentiation in the hippocampus. And this is -- in combination with acetylcholinesterase inhibitors in a synergistic way. And this is a demonstration of this is in the lower right-hand panel. So this is actually in free-running rats looking at hippocampal LTP response. And you can see in the presence of acetylcholinesterase alone, you get a modest enhancement in LTP, a larger enhancement with 007 and then what appears to be a synergistic enhancement in combination. So obviously, this demonstrates its unique activation or activity in the cholinergic system and able to overcome and potentially enhance signaling in the hippocampus, which obviously the central brain organ or region responsible for learning and memory. Another important aspect that's been described is that both the electrophysiological and in vivo behavioral functions of the compound can be blocked by GABAB receptor antagonists and nicotinic acetylcholine receptors antagonist. But importantly, the product does not directly bind to these receptors. Next slide. So in the preclinical space, the product has a very broad pro-cognitive effect. I don't have time today to go through all of the data. But just to summarize, this effect has been seen across species in young animals and aged animals, in chemically challenged animals or cholinergic challenge induced by scopolamine. And again, the PK-PD response -- exposure response relationship is highly consistent across these models and across species. And I'm going to show you in a few slides how these effects translate to humans. Next slide, please. So obviously, Rich touched on this and really did a great job of explaining why cognition is such an important target for helping patients suffering from CIAS and schizophrenia in general. And really, the focus of Recognify and obviously, Atai is really to look for products that can address not only cause meaningful clinical changes, but also improve quality of life. And we really feel that CIAS has a huge medical need that we can address with these cognitive benefits and directly impact quality of life. Next slide, please. So we're very fortunate to have a product that already has replicated procognitive signals. And again, just to summarize, the top study was a PK study in healthy normal individuals. So this would be your young individuals. The second study was a scopolamine challenge study. So this is the chemically challenged cholinergic model that was also run preclinically. And there was also a diabetic pain study that was previously run. So these individuals had an average age of about 60. So this is a more elderly population. And in each of these studies, there was a clear pro-cognitive effect that was expressed in a way -- in a biphasic way that correlates with the preclinical dose response that was seen in a variety of model systems. Another aspect of these original studies is that these procognitive effects were seen at the lowest dose. So one of the aspects of the current CIAS study that we wanted to explore was moving even lower in the dose range to fully characterize the procognitive dose response in CIAS patients. Next slide, please. Now this is just data demonstrating the procognitive effect in healthy normal individuals. So this also demonstrates that the product has very acute responses and can listen in acute pro-cognitive effect. So in this study, patients were dosed with a single dose or 7 days of TID dosing. And then a verbal learning task was applied. And what you can see is that even after a single dose, so highlighted here in red, even after a single dose, you see a strong procognitive response on verbal learning and that persists over the 7-day treatment period. Another thing you'll notice is that when you push the dose extremely high, you actually start to see suppression of cognitive response. So this is something that we've seen preclinically, and it's something that has been replicated in these studies in the clinic. Next slide, please. Now this slide is a little busy, and I'll try to walk you through it. So on the top half of the slide is the scopolamine challenge or cholinergic challenge model system that was run on healthy young individuals. It's similar to what was seen in the preclinical model, is the compound is able to inhibit or enhance cognitive performance in the presence of a cholinergic challenge. So in the upper 2 panels on the left is a verbal learning task and a delayed word recall, so learning and memory, and statistically significant effects we're seeing in both of these tests. But again, only at the lowest dose in this case, a 30-milligram TID. And in this study, it was over a single day of dosing. So again, a very acute response. Again, similar to what I just showed you in the PK study, when you push the dose too high, you actually start to see suppression of recovery in this population. In this case, it was an extremely high dose of 750 milligrams TID. One thing to pay attention to also in this study is that to test for central activity, EEG changes were monitored because it's known that scopolamine has its primary effects on alpha and beta bands where you get suppression of alpha and increased beta. And what was tested was to see if there were dose-related changes in the presence of the compound, but also how the compound affected the changes that were induced by the scopolamine challenge. And on the -- so the scopolamine alone is in this panel on the left -- upper left, where you can see increased alpha -- or decreased alpha and increased beta in the middle section. And you can see on the right what the scans look like in the presence of the compound and its ability to inhibit these scopolamine induced changes. So this will be reflected and this data will be echoed a bit in the CIAS study data when I get to that. Finally, on the bottom side, this is the diabetic pain study, again, looking at verbal learning and delayed word recall. In this study, there were two dose regimens. One was a low dose and one was a high dose. The low dose range from 40 to 80, one week of dosing at 40, followed by 3 weeks of dosing at 80 and then a higher dose with at 150 escalating to 300. Again, you see at the lowest dose given in the study highlighted in red, a statistically significant improvement in verbal learning and delayed word recall, again, episodic memory improvement. Next slide, please. So when taken together, you can see -- you can start to see a pattern emerge from the data that's already been generated about pro-cognitive effects in the clinic. You can see that there are 2 pro-cognitive effects. These effects were seen in this column challenge study and the human PK study at 30 and 40 milligrams TID and a more prolonged treatment duration of a month showing, again, statistically significant improvement in learning and memory at 40 to 80 milligrams TID. Importantly, we did see -- there was an assessment made of cognitive and physical function. And this is a questionnaire that's given to the patients to assess whether they feel they're benefiting from treatment. And again, in the diabetic pain study, this questionnaire came back as statistically significant, having positive -- in a positive way. But again, only at the 40 and 80-milligram dose, where there was a pro-cognitive effect. So this is where we may initially see some correlation between the pro-cognitive changes and actually a functional quality of life outcomes. Next slide, please. So what were our primary objectives moving the product into CIAS? Really, it was as simple as extending prior clinical data to CIAS patients. Initially, obviously, looking at safety and tolerability, looking at CNS activity by EEG changes, defining the dose -- the pro-cognitive dose range in the CIS population and exploring lower doses and also by making an initial assessment of what pro-cognitive signals might we be able to detect in CIAS patients. Next slide, please. So this is a schematic of the study design for our Phase IIa. It was a single center, single-arm, single-blind multi-dose study. It added 1-day placebo run-in of 4 doses followed by 2 days of TID dosing of 007. This dosing regimen was structured so that we could meet approximate CSF steady state following 5 doses and then do the testing after the at the sixth dose on day 2 of dosing. Again, these were independent cohorts dosed at 10, 20, 40 and 80 milligrams. And this was a typical schizophrenia patient population, age 18 to 50 on unstable protocol allowed antipsychotics. And the study was run with David Walling at his site at the Collaborative Neuroscience Research Center. Next slide, please. So as I mentioned, our primary objective with the EEG ERP readings were really to look at dose-dependent -- consistent dose-dependent changes across the different brain waves. And I show this as an example. So this is eyes open condition looking at the different dosing cohorts from 10 to 80. And what you can see here is there's a clear difference between the 10-milligram dose and the higher doses given in the study. In the 10-milligram dose, you see a decrease in the beta bands, whereas in the higher doses, you see a consistent increase across brain regions of Alpha and more importantly, of Alpha Slow wave Index. And the Alpha Slow wave Index is an indicator of functions of vigilance and weight fullness. So somebody that might be more alert. Next slide, please. For the ERP -- so what we saw in this preliminary set of data is that there were acute differences between the lower and high doses also in the ERPs. These differences suggest potentially that their support to sensory memory at the lower to moderate doses with the compound, we did not see with this acute treatment a consistent response across all time points and this may just be due to a very acute treatment regimen and short duration of treatment. The most consistent signal that we did see was on MMN and particularly at the 20-milligram dose. And I'll show an example of that data down below. So mismatch negativity is a passive auditory odd ball event, and you're really looking for a passive response to a change in a tone. And the amplitude really reflects the size of the response and the latency measurement, the rate of response. And so mismatch negativity is something that's suppressed in patients suffering from schizophrenia. And it is one of the markers that you do want to see move in a positive way in this patient population. As far as P300, we didn't see any effects on P300 at the lower doses and very sporadic effects on P300 at the higher doses. We do know from the data I just showed you that at the 40 and 80-milligram dose where there were sporadic decreases in P300. After a month of dosing, we do see positive and statistically significant effects on learning and memory. So at this point, we just don't have enough data to really make a strong assessment of what these early P300 signals might mean. Next slide, please. So now moving to the cognitive measures. So for obvious reasons, we selected a subset of the components of the MATRICS battery. The MATRICS battery is the gold standard right now for assessing cognitive benefit in CIAS patients. So the first component that we selected was verbal learning. We've seen multiple positive outcomes on verbal learning in previous studies. This is basically learning a word list that's given in multiple trials. And it's testing, again, verbal learning and memory. Second, we selected the BACS, our symbol coding test. An example of what that looks like is on the right, where you're matching up symbols with numbers as quickly as possible. And this test processing speed attention and really complex aspects of functional working memory. So this is an important test in CIAS patients. And then finally, we tested category fluency, which tests process speed, verbal ability and executive control. Basically, naming as many words in a minute from a specific category, in this case, animals as you can. Next slide, please. So this is the summary data of what we saw expressed as T-score. So Rich explained what T-scores are. And these T scores are normalized for age, gender and education level. I want to draw your attention first to the symbol coding task. This is one of the more sensitive quote from Rich here on the side, one of the more sensitive measures in schizophrenic patients and also in -- it has the highest correlation with other outcomes of the MATRICS battery. And what we see here is a very acute response in a in biphasic bell-shaped dose response similar to what we've seen with other tests and other outcomes. But also we see here that we've kind of bracketed the upper and lower limits of the dose response, where at 10 milligrams, we don't really see any response at all in this acute treatment, and we see 20, 40 and 80 giving a response. In addition, we did see what appears to be a signal emerging in HVLT at the 20-milligram and in the 40-milligram signal in category fluency. Again, something that's clearly different from the higher doses is that on HVLT, we see what could be potentially some suppression of cognitive response in verbal learning. One other point just to quickly point out is that I've shown you Cohen's D scores and not T scores on the other slide. If you convert these to Cohen's D, these would be medium to high effect sizes across the doses that gave a response. Next slide, please. So what did we learn from the cognitive signals? So we have a new test that hadn't previously been run, I gave a very response in an acute fashion. It's a test that's important for CIAS patients and correlates well with movement on the MATRICS battery. The observed changes and we've run this data by multiple KOLs, including Rich. And it's -- their consensus that these changes are larger than what you would expect from simple practice. The effect sizes are in the range that you would expect after 6 months of cognitive training. And these effect sizes also would correlate -- would be expected to correlate with improved work school performance. So again, some measure that can -- that has shown correlation with improved quality of life outcomes. Obviously, we replicated the immediate word recall and verbal learning effects in the 20-milligram similar effect size to what we saw in the diabetic pain study. And we will need to test whether longer durations of treatment will extend and strengthen these initial findings. Next slide, please. So overall, from the Phase IIa, I think we were quite successful in extending the previous observations into the CIAS population. So obviously, the product was safe and well tolerated. We showed central CNS activity by relevant EEG changes. These changes may suggest a more wakeful-attentive state in a relaxed state. We did see the beginning of some ERP changes that may be dosed cognition-related and dose-dependent, again, replicating what appears to be a bell-shaped dose response. And having a response on these initial cognitive assessments that -- and this is really the most important aspect of what we're trying to do in positioning the product in CIAS is really look for measures that are going to be a relevant indicator of patient outcomes. And Keith Nuechterlein, another one of our advisers, states it this way, yes, the simple coding responses are at a level that would correlate with better work school performance, which is an important point. Next slide, please. So where are we going with this in the future? So obviously, we just completed and replicated clinical effects in a small CIAS study. Obviously, the next study will be a double-blind, placebo-controlled study to demonstrate statistically significant improvement in cognition. Right now, we're working diligently with our key advisers to assess key elements for the next study, which, of course, are dose effect size and treatment duration. The data I've shown you has highlighted what we believe is the optimum dose range for the cognitive effects, which fall between 20 and 80 milligrams TID. The pro-cognitive effect size that we've seen across all studies ranges from 0.2 to above 1. In the CIS study, it was 0.4 to 0.8 and the duration that's been tested previously ranges all the way from a single dose out to 4 weeks. So basically, we're using this data right now to formulate a study that will give us the highest probability of success to demonstrate a statistically significant improvement in cognition in CIAS patients. Next slide, please. So obviously, we're optimistic and we're encouraged by these initial results in this Phase IIa CIAS study. We do believe that with the other functional outcomes that 07 can support that this product could be a central pillar for Atai in the neuropsych space, particularly when you think about the pro-cognitive effects in combination with the antiolytic effects that have been demonstrated preclinically and there is potential in other high-value opportunities in CNS in general. And with that, I'd like to thank you for listening and open it up to Q&A.

Yes. Thank you, Matt. For all the detail. There's a lot of information. I hope the audience was able to digest that. Obviously, very, very excited about the results that were outlined both on the quantitative EEG as well as the behavioral signals that we saw, in particular. I think it's worth mentioning that Keith, that was mentioned as one of our KOLs has reviewed this data actually, indicated that the sort of changes that we're seeing on simple coating. We're actually comparable to approximately 6 months of cognitive behavioral therapy. So really exciting results. Obviously, looking forward to replicating and extending these results in a more traditional Phase II trial focused on cognitive endpoints. All right, with that, I do want to open this up. I know we're running a little over time. So apologies for that.

[Operator Instructions] Our first question comes from Charles Duncan from Cantor Fitzgerald.

Yes. I think I'm set to go. Happy new year to everyone. I had a question for Dr. Keefe, and that is relative to thinking about the patient heterogeneity with cognitive impairment associated with schizophrenia. I guess I'm wondering if you could compare and contrast that relative to, say, the negative symptoms associated with schizophrenia that you see in different patients, whether or not CIAS is degenerative. And then what is the minimum, call it, length of time of exposure that you would like to see the company evaluate in the next double-blind placebo-controlled study to give you confidence that this drug may have activity that could translate into efficacy?

Yes. Thanks, Charles. Great questions, by the way. The degeneration question, first of all, is a controversy in the field. It's been hard to do studies that are long enough to demonstrate whether schizophrenia is neurodegenerative over time. It's certainly not over a brief period of time, like Alzheimer's. You can look at Alzheimer's patients and the slow downward progression is very clear. Schizophrenia is hard. There's actually just a very recent paper, I don't think it's been published yet, suggesting that there are these sort of three phases of illness. One is the -- the early one where there's cognitive worsening, things are kind of stable for 20 years or so. And then at the end of life, things may be -- there may be some degenerative process over there. The thing about Alzheimer's too, of course, is we know the biological mechanisms of that degeneration. There's no such thing in schizophrenia patients. And when you go in and you do autopsies and you look at the brains of people with schizophrenia when they're older and have died, there's none of the plaques and tangles, none of the indications of neurodegeneration that you see in people with Alzheimer's disease. So, controversial. I think it's not a crucial issue either way. I think that overall, it's the cognitive impairment is there from the beginning, even before illness. But at the onset of illness and throughout adulthood, it's substantial enough that it doesn't need to get worse to be really a big problem. With regard to your question about duration, I think that for a Phase IIa study that 12 weeks is a good duration in a Phase III program. FDA and other regulatory agencies, particularly Europe, will ask for a 6-month duration. But I think that's a lot to ask of a Phase II program. So I recommend usually 12 weeks.

Next question comes from Anvita Gupta from Cowen.

This is Anvita from Cowen for you today. Have you guys approached the FDA with the Phase IIa CIAS data? And if not, like when do you hope for a plan for a meeting. And additionally, if you could provide some color on what are the endpoints that the FDA might be looking for in the Phase II randomized controlled trial? And how does your exploratory endpoint of symbol coding might tie into all of that?

Matt, I'm happy to jump in, but why don't you take it?

Yes. So we haven't met with the FDA yet on our -- these initial findings. Obviously, this was a short acute treatment and really an exploratory study to extend previous observations. We're still working through with our advisers first what type of study we want to run next and what the clinical outcomes and measures will be. Obviously, in schizophrenia, as far as regulatory endpoints that are approved. The MATRICS is really the go standard. And whether we do the full MATRICS battery or a subcomponent of those, obviously, the simple coding component is going to be very important in whatever next study we do because of the strong acute effects we saw there in its correlation with overall change in the MATRICS. But those are all things that we're currently working through with our advisers. I don't know, Srini, if you want to add anything to that?

No, thank you. That was a good overview.

Our next question comes from Neena Bitritto-Garg from Citi.

Yes. So my question is just around thoughts on narrowing down a dose range, just given that. Patients with schizophrenia, I think, are notoriously kind of noncompliant with therapy. And I'm just curious about how you're thinking about that in the context of the TID dosing and potential for use in an outpatient setting versus in of the study with in patients?

Yes. So the PK does support, in particular, when you look at CSF PK, it does support moving to TID dosing. That modeling has been done. And that's something, obviously, we've been discussing with the Atai, when to bridge that. Also, the chemical composition of the API is amenable to slow release formulation. So there is the possibility that we could eventually move to QD. So this is -- that's a great question because it is an area where we're actively discussing on how to address that.

Operator. I think we are out of time, but let's just maybe do one more question then I can wrap up.

Okay. Great. So our last question comes from Andrew Sai from Jefferies.

Okay. I guess this is a question for Dr. -- I mean, I guess what part of the Phase II data set gives you the most optimism that this could work in the Phase IIb study? And I guess secondary to that is just how would you describe this overall data set relative to the other CIAS data sets that you've seen? Maybe just talk about it qualitatively.

Yes, absolutely. Both great questions. So 0.8 standard deviation improvement or Cohen D of 0.8 is extraordinary, really. And so that I don't remember ever seeing that granted. It's 29 patients, and it's 7 or 8 patients within that particular group. So replication is obviously essential. But the magnitude is there to be groundbreaking. And so that is the thing that gives me the most hope. And as Matt said, if you look at the literature and talk to any informed clinical psychologist that does cognition, tell you, there's nothing that is more important than symbol coding. It's related to functioning. It's sensitive to both positive and negative things. Great review 10 years ago by Dwight Dickinson at the National Institute of Mental Health, with Jim Gold from University of Maryland, really going into great detail about why that was the most important endpoint in schizophrenia recognition because of its relation to actual functioning. What was the other question?

I guess how would this data set compare so far to the other data sets you've seen for development?

Yes. I guess I mentioned that I don't know that I've ever seen anything with a 0.8 standard deviation. Yes. Of course, we need to compare that to placebo. I would expect about a 0.2 standard deviation change in the placebo group. So that's a 0.6 benefit over placebo. I don't think I've seen that.

All right. Again, as I pointed out, we're at time here. So I want to start by thanking both Matt and Rich for all their insights and there answering all the questions. And of course, I'd like to thank the audience as well for their attention and they're great questions. All right. Thank you so much.