Ladies and gentlemen, thank you for standing by, and welcome to the COMPASS Pathways' Third Quarter 2021 Investor Call. [Operator Instructions] I would now like to hand the conference over to your speaker today, Steve Schultz. Please go ahead.

Thank you, operator. Welcome all of you, and thank you for joining us today for our conference call today, which will combine both the third quarter 2021 results and the much anticipated COMP360 Phase IIb top line results. We're able to combine these 2 events as they coincided from this timing standpoint. Again, my name is Steve Schultz. I'm the Senior Vice President of Investor Relations at COMPASS Pathways. And today, I'm joined by George Goldsmith, Chairman and Chief Executive Officer; Dr. Guy Goodwin, our Chief Medical Officer; and Piers Morgan, our Chief Financial Officer. The call is being recorded, and it will be available on the COMPASS Pathways' Investor Relations website shortly after the conclusion of the call. We hope you've had a chance to view the 2 press releases issued earlier today, one on the Phase IIb clinical result and the other covering our third quarter results. We've also posted a presentation to the Investors section of our website in the Events section that supports the top line Phase IIb data presented in this webcast, so you can follow along. There's also an excellent COMP 001 Phase IIb hub on our website in the Clinical section, and I encourage you to check that out as well. Before we begin, let me remind everyone that during the call today, the team will be making forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995 as amended. You should not place undue reliance on these forward-looking statements. Actual events or results could differ materially from those expressed or implied by any forward-looking statements as a result of various risks, uncertainties and other factors, including those risks and uncertainties described under the heading Risk Factors in our annual report on Form 20-F filed with the U.S. Securities and Exchange Commission earlier today and in the subsequent filings made by COMPASS with the SEC. Additionally, these forward-looking statements represent our views only as of today and should not be relied upon as representing our views at any subsequent date. We specifically disclaim any obligation to update or revise any forward-looking statements. With that, I'll now hand the call over to our Chairman and CEO, George Goldsmith.

Thank you, Steve, and welcome, everyone. I'm pleased to be here today to present both the top line results of our Phase IIb clinical trial of investigational COMP360 psilocybin therapy and our third quarter results in one webcast, which I hope makes it more convenient for all of you. Before going into the Phase IIb result, which I know is the star of the show today, let me first begin today's call with a business update on our recent progress. I will then ask Guy Goodwin, our Chief Medical Officer, to discuss the top line results of the COMP360 trial; and Piers Morgan, our Chief Financial Officer, to provide a financial review. We will then open the call to questions. This has been an exciting quarter. So we've made great progress on numerous fronts. Last month, we presented positive data from an open-label investigator-sponsored study from the Maryland Oncology Hematology Group at the Aquilino Cancer Center in Rockville, Maryland. In this study, the team also pioneered simultaneous group administration with 2 to 4 participants being given psilocybin at the same time with 1:1 therapist support. This tested the value of group support for cancer patients with moderate major depressive disorder as well as the potential for increased scalability in providing psilocybin therapy in real-world settings. Most importantly, some participants benefited from this study, whilst expanding our knowledge on how to improve future research for this population of special interest in MDD. In our goal to continue to expand our development pipeline, last week, we announced the launch of a study in posttraumatic stress disorder, or PTSD. This will be a COMP-sponsored Phase II program and should begin in the next month or so. Over 350 million people worldwide suffer with PTSD with an individual lifetime prevalence of 7.3%. In addition…

Thank you, George. It's a pleasure to speak to everyone today and review the positive data generated from the COMP360 Phase IIb clinical trial, COMP 001. I'll be referring to the slides on our website that Steve referred to earlier, and Slide 2 is where I'll start, our COMP360 psilocybin therapy. Our COMP360 psilocybin therapy consists of 2 components. The first is the investigational drug itself, which is presented in oral capsules. And the second is the psychological support which consists of the preparation phase, support during psilocybin administration and the subsequent integration. This, in total, is the psilocybin therapy that is being tested in this study. It is important to point out that this approach is developed as an integrated therapy and the results from this trial can only be interpreted in this context. Other forms of psilocybin on their own or with some alternative support therapy protocol are unlikely to be comparable. Slide 3 is COMP 001 study design and endpoints. Looking at the trial design and end points, you will see that all patients were screened and subsequently entered into the study within 3 to 6 weeks. This preparatory period was to allow the managed withdrawal of antidepressants which the patients may have been taking. Bear in mind that these patients were all resistant to treatment with previously administered antidepressants and, therefore, had to be washed out before COMP360 could be administered. The randomization was to 1 of 3 doses, which were either 1 milligram, 10 milligram or 25 milligram of COMP360 and were administered as you can see here on day 1. The effects of the treatment intervention were measured on day 2 at week 2, week 3, week 6, week 9 and week 12. The primary end point was designated to change in the MADRS…

Thank you, Guy. The company continues to maintain a strong financial position with cash and cash equivalents of $294 million at September 30, 2021, compared with $190.3 million at December 31, 2020. This is expected to fund our operations into 2024. I will now recap our financial results for the 3 and 9 months ended September 30, 2021. The loss from operations for the 3 months ended September 30, 2021, amounted to $21.8 million compared with $13.5 million for the prior year period. The loss from operations included noncash share-based compensation expense of $2.3 million for the 3 months ended September 30, 2021, compared with $5.2 million noncash share-based compensation in the prior period. The net loss was $15.8 million or $0.38 loss per share for the 3 months ended September 30, 2021, compared with a net loss of $16.7 million or $1.30 loss per share during the same period in 2020. The loss from operations for the 9 months ended September 30, 2021, amounted to $54.9 million compared with $39.9 million for the prior year period. The loss from operations included noncash share-based compensation expense of $5.9 million for the 9 months ended September 30, 2021, compared with $16.6 million noncash share-based compensation in the prior period. Research and development expenses for the 9 months ended September 30, 2021, amounted to $30.4 million, an increase of $11.6 million compared with the 9 months ended September 30, 2020. This increase in research and development expenses compared with the prior year period was primarily attributable to an increase in development, personnel and consulting expenses as the company continues to expand its activities, including launching a new trial in PTSD, preparing for Phase III trials in TRD and expanding our digital activities. General and administrative expenses were $24.5 million for the 9 months ended September 30, 2021, compared with $21.1 million for the same period in 2020. The increase of $3.4 million in general and administrative expenses compared to the prior year period was primarily attributable to increases in personnel costs to support our growth initiatives, including our transition to a public company, increased insurance and other costs of operating as a listed company, partially offset by a decrease in noncash share-based payments. The net loss was $46.1 million or $1.17 loss per share for the 9 months ended September 30, 2021, compared with a net loss of $41.5 million or $3.90 loss per share during the corresponding period in 2020. Thank you, and I'll now hand the call back to George.

Thank you, both. We're now working to schedule an end of Phase II meeting with the FDA as quickly as possible. I remind you that we have breakthrough therapy designation, which can provide us with a more rapid and open dialogue with the FDA. Our immediate goal is to present these data to the FDA and gain agreement on a Phase III program design. We will report on our progress in future communications. In addition, we expect the SSRI drug interaction study to be completed by the end of this year, which will provide important information about the potential influence of concomitant SSRI use with our COMP360 psilocybin therapy. You will recall that our Phase IIb study did wean patients off of their SSRI therapy prior to the COMP360 psilocybin administration session. Overall, this trial's result is consistent with the data we saw coming from the Imperial College London study in April and the recent Aquilino Cancer Center data, which is to say the patients saw a rapid and sustained effect with a generally well-tolerated safety profile. We believe this consistently supports our assertion that COMP360 therapy can potentially bring a new approach on how we may be able to treat patients for whom a few options exist today, and today's result continues to bolster our leadership in this field. To be clear, all of our efforts are aligned with our commitment to think differently about mental health and to follow up those thoughts with well-executed actions. The need is great. Patients are suffering. We have to provide them with the new tools in the most responsible way. That means we must understand who can benefit from our work and who cannot at this point. We are working day and night to ensure we expand our understanding. Thank you for your time today. We will now open the line for questions. Operator?

[Operator Instructions] Your first question comes from Ritu Baral from Cowen.

Congratulations on the entire data set, not just the top line, but especially that remitters analysis. I have -- my first question, and I'll keep the second one really tight. My first question is Slide 11, the TEAEs, the venture preferred terms, can you go into a little more detail about what was seen around the individual behaviors like when did they occur? Were these patients responders or remitters? And were they on rescue meds, especially suicidal behavior, self-injury and ideation, I think, are the most concerning to investors?

Great, Guy?

So the suicidal behaviors were really reported at least 1 month after the administration of treatment, and they occurred in patients who were essentially nonresponding. And the suicidal ideation was a little different, and that was at least in one case was quite early on in the treatment process. Of course, these numbers are very small. And our plan is to look at them in as much detail as we can. But in addition, in the secondary analysis to look at the Columbia suicide scales that were collected on every patient at every visit, that will give us, I think, a much more accurate understanding of the relationship between treatments in suicidal -- in both experience -- suicidal ideation and suicidal action.

The intention of self-injury, is that, I guess, an aspect that is overlapping with the suicidal behavior and suicidal ideation?

We actually don't believe it does. It's classified in that way, but it represents really quite trivial self-harm. It means scratching or banging the head or very simple things that really don't carry any suicidal threat or intent, but they are quite common in depressive populations. They're well described and I'm afraid they're just part of the picture for these patients in part of that suffering.

Got it. And a very quick follow-up, George and Guy, do you have your dose now? And do you think that a Phase III with the dose or powering a Phase III for durability, whether it's week 8 or week 12 would make sense for Phase III from a reimbursement perspective?

Yes, we think we conclusively have our dose. And in speaking with payers and others, we have included even in our Phase IIb trial health economic-related measures, quality of life measures. And we'll be discussing all of this as we go into our Phase III side with those who advise us on payer metrics, et cetera. We've been working with payers and health technology assessors here in Europe since 2016. So we're well aware of their requirements, and we expect to be able to address that in a well-powered study. But obviously, our Phase III program will have to be agreed and discussed, which we've been discussing already with regulators on both sides of the Atlantic. And we look forward now to bringing these data to then agree that program and swiftly move into its execution.

Next question from Josh Schimmer from Evercore ISI.

How are you thinking about ways to preserve effect beyond the first few weeks, either with use of adjuvant therapies or patient selection to optimize for durable effect?

Josh, I'll start to answer and hand off to Guy. I think it's a really important question. When we started this whole program, our whole goal was to identify who responds and who doesn't and who responds longer and who needs additional support. To that end, what we've done is actually recorded every therapy interaction. So we have machine learning out looking at transcripts across the site and look the trial participants really starting to understand the deeper level of patterns, including changes to patient narrative, therapist actions that are then correlated with particular outcomes. This is unprecedented data and we'll be digging deep into that to look at how we can, in fact, optimize therapy and develop, as we've said from the beginning, a more precise and predictive model of care with psilocybin therapy. So we're looking deeply at these data. But keep in mind, we just received the data about 48 hours ago. So there's a lot of work to do. We're just reporting top line today. Guy, do you have anything else to add?

I think I'd just underline that the way in which this study was designed is that the psychological support is essentially safety around the administration of the medication. So this is a test of the medication in isolation. It's pretty remarkable that we get these effects after a single dose after 12 weeks. Clearly, if we are working with colleagues within the company to look at digital support mechanisms and monitoring which may very well have to carry the promise that we can improve outcomes through that sort of approach, then we'll certainly be considering that in the future.

I should just also say that this is the data set that we developed in close collaboration with payers and regulators, again, on both sides of the Atlantic in over 10 countries to answer the fundamental questions, and we couldn't do that unless we really look at the dose finding and a single dose of its durability. Only then can we look at how we can identify the patients who have different experiences and look at how to optimize that experience to maximize the outcome and the duration.

Got it. And when do you expect to have some of the additional analysis completed in the full data presentation?

We'll be working on that part and doing it as soon as possible. We don't have a date, but you know us, we're going to be at an intensity and vigor.

Next question from Charles Duncan.

Okay. George and team, congratulations on a rigorously conducted trial. It's clearly seminal event for the field. My first question is regarding next steps on sizing and timing, and I know you need to discuss this with the agency. But could you imagine being able to outline a protocol or even operationalize a next step, next maybe a Phase III by roughly mid-'22?

We can certainly imagine many things, and that's one we're focused on imagining, to use your term. We are very heavily recruiting sites and looking at preparing our Phase III infrastructure to build on our really robust Phase II infrastructure. That all work is all in process. But again, we can't anticipate what the agency, the FDA and the EMA and HRA will say as we review these data. So we are ready to hit the go-fast button as we do whenever we get the word from the agencies. But we will be reviewing this very rapidly with them.

Okay. That's good, George. Second question is regarding any phenotypic factors. I understand that this is fresh data. So probably haven't done the analysis. But is there anything that you're in particular interested in such as prior lines or time since their -- since diagnosis that may have drove responses in terms of either the primary endpoint or durability? Anything that you're particularly interested in looking at where you've seen?

Again, I will start that and then hand off to Guy. I think a few things were really important to us. One was that this study actually used patients who were referred based on medical records. So we have a validated background and those background information -- or that background information will be used in future analyses, obviously. One of the things that was really important to us was to be able to demonstrate that these benefits could happen to people who had no prior experience or 94% of our sample had no prior psilocybin experience, quite odds with the prior studies. And the majority of our therapists in all of our centers also had no prior psychedelic experience. This was so important for us to generalize in all of the population, which is approximately the same in our study, about 90-plus percent who haven't had these experiences. So to be able to demonstrate that this works in that environment, it's really important for scalability and, more importantly, to help most patients possible. Guy?

Yes. Just to say that, obviously, we're aware of existing data that suggests that one can, to a limited extent, predict treatment response from baseline measures, and we have those baseline measures. And there is also, of course, emerging interest in biomarkers, and we'll be looking at the measures that we have in that category as well. But as you rightly said at the beginning, we're not yet in a position to comment on that. We're just going to be very excited to follow up and look at those questions and answer them for you.

Okay. Last quick question, then I'll hop back in the queue, referring to Slide 6 and 7 in terms of remitters or responders and remitters. It looks like weeks 9 -- or 6 and 9 numerically are a little bit less than week 12. And I guess I'm wondering if you think that's an artifact of the remote visit or if there's something else going on there with the drug. It's hard to imagine what that might be. But tell us about that and tell us how you would use those learnings in the future trial designs.

Well, I think in general, we feel that the data from 6 to 9 -- 6 to 12 weeks gets slightly more difficult to interpret because there is additional treatment going on in the 3 treatment arms with some of the patients. So that's a factor that we've had to consider, and we're going to have to analyze more closely. So I think we feel very confident about the statistical differences up to 6 weeks that we're encouraged by the group to receive no treatment. But we need to understand the whole data set. We're going to have to take into account the fact that antidepressants were available for use and in some cases were used in all 3 arms of the study.

I should also add that one other thing where we need to look at, this study was conducted through the very depths of the pandemic with people being in lockdown, and we'll be also looking at are there any effects of when the lockdown is occurring on patient populations, et cetera, because that's quite an unusual event, fairly have had many mental health sequelae for people. So we are going to be looking at that. It's an unusual thing to have to consider a trial that it's a very meaningful one. So standby.

Next question from Esther Hong.

Congratulations on data, really impressive. Wanted to ask about patients who had to discontinue antidepressants prior to the administration of COMP360. Any additional details on these patients and perhaps regarding washout periods and response rates and remitter rates? And then I've got a follow-up.

Yes. Well, we will -- all the patients were washed out from antidepressants, and we will be giving further in details as to how long it took and what the delays were between initiating that and the actual dosing day in due course. But you asked a very interesting question, of course. There's a lot of interest in the effects of discontinuation. But we think that discontinuing in patients who were highly symptomatic and treatment-resistant proved to be relatively straightforward to generalize rather grossly, but nevertheless, generally successful and acceptable to patients. And as you can see, many of them stayed off for the next 12 weeks as well.

Right. Got it. And then just quickly, can you remind us what biomarkers you're looking at?

We have polymorphism in the 5HT2A receptor that may be of interest to the drug action. And we're also interested in an inflammatory biomarker, which will be informative as well.

Great. Congratulations.

Next question from Patrick Trucchio.

Congrats on the [indiscernible]. I have a follow-up question on the suicidal behaviors. I'm just wondering, what was the protocol definition in terms of how far out post the exploratory session? Could this be considered attributable to the treatment? Is there a limit to that? Is it within the 12 weeks? And is that something that's consistent across TRD studies? Or was this unique to Phase IIb trial for COMP360?

It's -- I mean at any time in the 12 weeks that there was a suicidal ideation or behavior, it was recorded. And the timing, of course, is recorded. In terms of frequency, we -- although it's completely very low numbers, make the estimate very uncertain, they are comparable with, in fact, the majority of studies ever conducted in depression. The recent meta-analysis which suggested that the average rate of suicidal events at various times was about 3% in actively-treated patients. So we seem to be aligned more or less with that kind of rate. And of course, this is a group where you would expect to see emergence suicidality as a risk because of their history and so on.

Got it. That's helpful. And then I'm just wondering if there was any learnings that have heard from the study or some of the others that have read out recently in terms of feasibility of simultaneous treatment or therapy. I'm also wondering if you think you would need a separate Phase II trial to assess the feasibility of simultaneous therapy? Or could this potentially be a step for treatment arm in the Phase III program?

I don't think we're at a stage where we can answer that question. We do -- we will have experience of administering 25 milligrams of psilocybin therapy to patients who are still taking SSRIs in a small open-label study, which will report relatively soon. We'll then be in a position to make an informed judgment about the rest of your question.

Yes. That's makes sense. And then just one last one, if I may. Just in terms of the PTSD trial, if you can just kind of discuss again there, too, any learnings you could take from the TRD Phase IIb trial or the other trials for that program? Or even a program with MDMA and how should we kind of expect the pace of enrollment in the PTSD trial to progress?

Well, I'm not sure -- well, the learnings, I think, will relate to the therapy provision. We have an enormous amount of data that we will be able to analyze that can pick apart the interactions between the patients and the therapists in this study. Every single administration, in fact, all of the sessions with therapists were recorded and we have really sophisticated approaches ongoing to analyze the interactions and the key themes that come out and that's going to shape how we deliver our treatment program in the future. With the specific lessons for PTSD and the recruitment and so on, we will -- we're doing the study to find out that feasibility.

I'd just go and add -- I'd just say that as we pursue areas of high unmet need in multiple indication areas, our focus is to be a learning organization and continue to apply learnings in real time so that we can actually progress all of our clinical studies and benefit patients as much as possible. So we are very much looking at building these staircases with smaller studies into larger studies and drive learning.

Next question from Sumant Kulkarni.

Thank you for all the company's work on what must be an exciting day for patients who might see another treatment option for TRD. I have 3 questions. I'll ask them individually. First, based on what you now know with your Phase IIb data in hand, are there any specific variables that you instantly see in terms of needing adjustment for your upcoming Phase III program?

Guy, do you want to start with that?

I don't think instantly, we see anything we need to adjust. But obviously, we're reflecting on a whole series of issues that are raised particularly when we look at the secondary analysis. What is your reaction, George? That's mine.

Yes, it's mine as well. Again, these are very, very fresh results. We have the largest data set ever created in this area. We have an incredible team, statisticians and data scientists. And we're really looking at all of that and taking that into discussions with regulators and devising the path forward.

Got it. My second question is -- sorry, go ahead.

No, it's fine. No, I was just going to say that you appreciate the top line results are all driven by a very, very strict and conservative analysis plan. So that's a sort of limitation on how far one can immediately react to differences.

Understood. So second question. Given your work with patients, physicians, payers and other components of the supply chain so far, what do you think would be the sweet spot for durability and COMP360 that in the real world?

Well, I think that everyone's dream is to actually look at how we can predict and prevent relapse. And we're hard at work looking at new tools that might help with that and digital phenotyping. But I think that people understand today that this is an incredibly difficult patient population to support and serve. It was a single administration of monotherapy. We see great promise at 3 weeks as well as 12 weeks, considering the lives of these people. So we'll be in lots of discussions, and I think we'll come back to that question with more information in the not-too-distant future. So we start finalizing our Phase III plans.

And my last question is, should we expect to see any difference in response to COMP360 based on the specific antidepressant used or failed by patients in the past?

We have no idea what to expect because it's an empirical question. And fortunately, we have data to be able to answer it.

Let's continue, your next question from Bert Hazlett.

My congratulations on a well-conducted study and an important one. Just 1 or 2 questions to try to better understand the patient characterization of this study. Do you -- could you -- would you be able to provide baseline MADRS rates for either the study as a whole or any of the particular groups?

Yes. Yes. I think it's on...

Yes. It's on Slide 5, if you want to refer it. It's written rather small. It's very easy to miss. But it's on -- all the data is on Slide 5 that you requested there.

Okay. And then just with regard to geographical groups, are there any notable similarities or differences within the particular patient groups, either in North America or in the U.S? Or excuse me, on in U.S.?

We'll be looking at that with great interest. So we haven't been on that.

I think -- go ahead.

Please, go ahead.

I was just going to say what's really remarkable about this is we've trained therapists in 7 languages, 10 countries and demonstrated this results across the board. It's going to be fascinating to look at all sorts of cuts in terms of demographics, countries, et cetera. But we were amazed and I think also amazed at the levels of recruiting and engagement we have in the Netherlands from our sites. And so I think that was also a very interesting characteristic given the availability of other substances there. So this is clearly an unmet need and a high level of interest in the trial.

And then just one more for me. With regard to the serious treatment-emergent adverse events, do you have any more information at all with regard to the time course of those events just in general characterization? Were they closer to the administration and maybe waned further on? Or just, again, any anecdotal characterization you might be able to make with regard to those?

It wasn't simple. Every patient tend to have their own complexity, as you probably appreciate. The suicidal behaviors all occurred at least a month after the drug administration. So we can certainly say that from this data. Some of the ideation occurred closer a couple of cases. But I think -- I would emphasize again that we will get much finer grain data when we look at the ratings of suicidality that we will have in the secondary analysis. And we'll certainly let you have that information as soon as we have it.

Congratulations again.

Next question from François Brisebois from Oppenheimer. François Brisebois: Just wanted to ask, in terms of the placebo effect, was that based on literature and your expectations, mostly in line just based on the potential negative placebo effect here and the triple blind that you mentioned?

Well, we have nothing really to go on, but have not been -- there had been other studies, but they've usually been crossover studies. So we had to really think this through from the start when we designed the study. All of the patients were obviously prepared in such a way that their expectation was that they would have a psychedelic experience. And so our assumption was that there would be an appreciable nocebo effect in the group who got 1 milligram. But at the same time, we also knew that they were getting the wraparound psychological support that's extremely positive and could be regarded as placebo. So we did -- with the reason we -- you do experiments is to find out what happened. And this is what happened, and we will try and interpret it as best we can.

Next question from Elemer Piros.

Yes. George, you mentioned that you think you have divided the 25-milligram dose. Do you -- would you exclude maybe repeat dosing within the clinical trials?

Right now, what we're doing is trying to understand, as we said, each patient's experience. And I think that, that will help inform thoughts along that line. But I think what we really needed to do with this trial was to actually establish a level of understanding of durability up to 3 months of a single dose. That will help inform a tremendous amount of analytics moving forward. And then I think we'll step back and look at what's required. I do know that I think if we have in the 21st century the ability to predict relapse and to preempt it, that's really the goal for us. And so we're hard at work, looking at those kinds of things and aren't rolling anything in around at this point. Obviously, lots to do and lots to discuss with regulators.

Yes. And the last question is, how many sessions were integrated per data for this protocol, George?

There were 2.

2. And do you think that, that number is sufficient?

Again, we have an amazing team working on all the transcripts, all the interactions and creating an empirical response to that question. So I think that one of the things that's really important is that we look at a very fine range precision approach to this, patient by patient so that we really develop a plan for every one and anyone. And I think this is really where we're headed. So we're going to, again, be doing all those analyses and sharing what we learned and then informing our future plans.

Congratulations.

Next question from Neena Bitritto.

So I just wanted to ask about the use of rescue meds in the follow-up period. I guess can you talk a little bit more about when patients were allowed to start taking rescue meds? And then anything you can share at this point on the actual rates of use of rescue meds in the different arms, that would be great.

Sure. I just think that Guy, when you talk, you should probably make a clarification or start with the clarification of rescue versus other medications.

Yes. So our understanding of rescue would be the use of medications on the day of the study, the day of the dosing. And if my memory serves, I think there was only a single example where that was necessary. So that means that we were successful at delivering the actual dosing day without having to rescue patients from the experience. So we think that the therapist did a great job. The other sense in which there's rescue medications is that these patients have been withdrawn from antidepressants. So we had to expect in advance that there would be a pressure for them to go back on to antidepressants if they showed no response or limited response to our treatment. And that would apply in all 3 arms of the study, of course, However, we did encourage the clinicians entering the trial to give the treatment of fair trial after 3 weeks as far as they could. And then after that, they were free to use medication if they felt it was clinically indicated. So we've got a picture that we've not yet fully analyzed of a number of the patients returning to antidepressant treatment. And when we fully understand the patterns, we'll publish them. I don't know whether you want to say anything more about that, George, but...

No, I think that's sufficient.

Your next question from Michael Okunewitch.

So I wanted to ask if you could put the data in a bit of context because it seems like the magnitude of result is not as high as previous studies. Could this be somewhat owing to the fact this is a single-dose protocol? And considering that the patients and the clinicians were psychodelic-naive, should this be viewed as a more conservative study to establish that baseline for COMP360, which could be improved in a more optimized setting?

I think that's a really, really hard question to answer. So first of all, I think that we simply did this study to address all of the regulatory and payer concerns about whether this was actually a generalizable model or not. And I think we can feel confident in saying that this is a generalizable model. And, in fact, they'd be optimized in various ways that we're investigating. But I do think that what really distinguished the study is it was multicenter. It was not -- it was done by -- with a great deal of clinical and scientific equipoise, which is something we could do at this scale that's very difficult to do in a single site study environments. And we use this triple blinded model where we had an independent radar calling each patient in their native language completely blinded to the study. So this was the bit that said we can really count on these results just because of the independent assessment using the gold standard MADRS. So we think that this is a great place to start from, and you bet we're interested in optimizing.

All right. And then I wanted to ask about how clinically meaningful the absolute reductions in MADRS and MADRS score, which were around, I think, minus 12 at week 3 and that stayed around minus 10 at week 12? And how does that compare to SSRIs?

Guy, would you like to take that?

Yes. Well, I mean, I think just what you make of the average change in the population is obviously very much to do with how you interpret clinical trials. The experience of individual patients is something that I suppose more concerns me. And so I think when you look at just data and see the highlights of remitters after 12 weeks, that's something that I take away as particularly clinically significant. The comparison between different deltas as they're sometimes called, is quite interesting, but it's difficult when the trial [indiscernible] quite the same way. I do appreciate that our effects are seen from day 2. The effect of SSRIs as seen usually by day -- by week 6. So in a sense, they only look at the interval over which we're showing statistically significant results. And you can probably see that just eyeballing the data on Slide 5 that the difference between 25 milligrams and 1 milligram at that point is still around 5 or something on the MADRS 5 or 6 and the usual reported 6 weeks in SSRI trials on average is somewhere between 2 and 3. So I think in terms of comparison, as far as I'm prepared to make them, you -- that would be the comparison I would make. But of course, we're looking at the TRD population. There are various caveats around that population, not responding at all to SSRIs, but by definition, have failed to respond to SSRIs. So this gets a little -- it gets slightly academic after a while those sorts of comparisons.

All right. And then just one more quick one, and I'll jump in the queue. For the suicidality, considering that, that was largely a month after, could that be related to a sense of hopelessness in patients that nothing worked for them so far and now they try psilocybin and they didn't respond to that either?

Well, I think that's a consideration we'll certainly be taking into our deep dive into the data in the future. It's an interesting possibility, we agree.

Yes. And I think we will have to look also at the level of expectation that is set in the general press. I think this is something really important to take responsibility for that. There are no panaceas, and the study indicates that. Any other question?

No further questions at this time. Steve?

Sure. Let me hand it back to George for closing remarks. Please, George?

I just want to thank everyone for their interest, participation and support. This is something that is going to be the start of really developing what we view as a way to transform the patient experience and mental health. We're on it. We're working at how we can help reduce suffering at scale and appreciate all of your interest and support. Thank you very much.

That does conclude our conference for today. Thank you for participating. You may all disconnect.