Third Bridge专家复盘Compass Pathways COMP360三期数据，认为疗效低于预期但仍可能获批

[[00:02:23]]
I'd like to begin by unpacking the COMP360 [psilocybin] phase 3 data. In Compass Pathways' COMP006 trial, which studied up to two initial doses of 25mg vs 1mg, COMP360 25mg demonstrated a negative 3.8-point delta vs the 1mg dose on the MADRS [Montgomery-Åsberg Depression Rating Scale] primary endpoint. Absolute change was minus 7.6 points at week six. The 10mg dose showed a negative 2.5-point delta vs the 1mg dose, for an absolute change of minus 6.3 points at week six. For added context, in the COMP005 study, there was a negative 3.6-point delta on the MADRS primary endpoint. In COMP006, there was a 39% response rate as defined by a greater-than-or-equal-to-25% improvement on the MADRS score at week six, and that compares to 25% in the COMP005 study, highlighting the benefit of a second fixed dose in the COMP006 part A portion. There was a statistically significant separation of the curves as soon as day two, so clearly rapid onset, and 28% of patients who achieved a response that had not remitted by six weeks went into remission after an additional dose in part B. With all that said, it seems the headline data are somewhat less compelling than originally hoped. Could you unpack the data further?

[[00:04:10]]
There are a couple of things to bear in mind, I think, that reflects favourably on the data and then also unfavourably. In terms of unfavourably, this was, as we know, a lower effect size, lower between condition effect size delta than was achieved in their phase 2 study. They reported a similar response rate, but that's because the definition of response was altered. Before, they were using a more conservative threshold for response being a change in at least 50% of participants' symptoms. In this study, they were using a 25% threshold. Favourably, I think we don't know as much about the variability in response. We need to take note that this is a treatment-resistant population, which is harder to treat and also only has two approved medications for TRD approved by the FDA. Sorry, the variability, I think, is important to understand as the data is released because it could be that because this mechanism of action associated with psilocybin is so different than the other approved medications and standard of care treatments for depression, it could be that this treatment can help individuals that can't be helped by others, therefore, releasing them. Even though we are finding that these are delivering response to, it looks like, 40% under two doses of participants, it could be uniquely helped by this particular medication.

Other than that, I think it will be interesting to see whether this 2.5 delta in the 10mg. The 10mg condition is really critical in order to demonstrate some degree of blinding integrity to the trial. It will be interesting to understand… Sorry, let me actually back up. In the previous trial in the phase 2, I believe that they found a statistically significant difference between the 10mg and the 25mg condition. In this, they haven't released that data yet, even though perhaps we could… It would be difficult to back into it statistically because I don't think they give us the standard deviations. In any case, if there is no statistically significant difference between the 10mg condition and the 25mg condition, then it could give the FDA pause about whether there is not functional unblinding that produced the greater response in the 25mg condition. I think my personal intuition is that the FDA would overlook that with preference for the data showing a delta between the 1mg and the 25mg condition. That would be also consistent with the precedent of Spravato being approved without actually any comparator in the placebo condition. There was actually placebo, passive placebo in the control for Spravato and it was approved. I think the reason for that and for why it's likely that COMP360 will be approved here as well is because FDA just does not have adequate solutions for TRD at this time.

I do believe that it's possible the bar will be lower for COMP360. I will say also that the treatment-resistant depression indication, that was done over much time speaking with FDA regulators. They believe that that is a significant competitive advantage for them.

[[00:09:30]]
Definium Therapeutics, formerly MindMed, is trialling its LSD therapy, DT120, in MDD [major depressive disorder]. COMP360 is in TRD [treatment-resistant depression]. In the former indication, there are 40-50-plus FDA-approved therapies, with payers requiring patients to try and fail on multiple generics before any branded drug can be used. One can argue there is less of an unmet need in the MDD population. In TRD, the opposite applies. All else being equal, does that give COMP360 a higher probability of regulatory approval?

[[00:10:17]]
Correct.

[[00:10:20]]
The 10mg COMP360 dose showed unexpected activity in the COMP006 study vs almost no activity in the phase 2b. While the efficacy is numerically lower than the 25mg dose at all time points, we don't know whether there's a statistically significant difference. Looking at the plot, the error bars are overlapping at multiple time points. Is that a significant concern here? Is it simply a statistical artefact? Do you think it's functional unblinding, or is there a third possible explanation?

[[00:11:16]]
Just one thing statistically to keep in mind that the overlapping error bars at some post-treatment time point, those don't necessarily indicate that there's no difference between… I actually don't know exactly which plot you're referring to, but I would say that that doesn't necessarily indicate that there's no difference between pre-post in one condition vs another. Just putting that out there. This might not be helpful, but it's not necessarily an issue that the 10mg is producing some form of efficacy. I think what Compass and others are attempting to achieve or attempting to demonstrate or would like to demonstrate if it exists, is a dose-response effect. That 1mg would be less than 10mg, which would be less than 25mg. To be honest, if 10mg was producing a substantive decrease in symptoms, that would essentially mean that this drug could be effective at a lower dose, which is not necessarily a negative or a huge disadvantage for the product and the medication, but where there could be an issue with FDA is that, like I said before, if there is not a statistically significant difference between 10mg and 25mg, there could be a concern about functional unblinding. There will be more data. There's very little data that's been released so far. I know that within this particular trial, they did measure blind breaking.

They asked other participants, "What condition did you think you were in?" If it was the case that all the 10mg participants knew they were in the medium or lower dose and all the 25mg participants thought that they were in the high dose, then there would be concerns, I think, associated with this difference between the small difference because we are seeing a substantive… The 2.5 is a substantive delta relative to the 1mg. I suppose this is telling us that there's a pretty minimal delta. We do know the delta actually between the 10mg and the 25mg. We just don't know whether it's statistically significant. Actually, that can probably be statistically deduced. Essentially, what we're finding here is a 1.3 point difference between those. That would actually be good for your audience to go into, and I can look into that as well, but whether that 1.3 point difference between them is statistically significant given their sample size. If it's not and there's basically blind breaking, then that might give the FDA pause and might point to this being a product of functional unblinding. That being said, I still think personally that the FDA would approve it, but there obviously is some risk associated with this.

[[00:15:14]]
What's the likelihood there was significant blind-breaking in the COMP006 study? Compass has pointed out that north of 95% of patients enrolled in the COMP006 study were psychedelic-naive. Presumably, that would limit the likelihood of the participants correctly guessing which dosage group they're in.

[[00:15:51]]
It absolutely would. It's an important and conservative practice that they've done. The problem with giving a good answer here is that the previous research has been fairly derelict in adhering to higher, better methodological standards when it's come to blinding integrity and measurement of blind breaking. I think some of these companies have omitted it by design because they didn't necessarily want to raise those types of questions. It's really hard to say whether this 10mg dose would be blind. I will say that in their phase 2, they did show that people in the 10mg condition were experiencing or reporting substantive psychoactive effects on the 10mg. I know from my experience that 10mg does deliver psychoactivity. I think it is it is reasonable to conjecture that this 10mg when administered to psychedelic-naive individuals could be convincing to them that they received a substantive or perhaps high dose and blind them effectively within the trial.

[[00:17:23]]
It sounds as if running a large well-controlled study for any psychedelic in TRD is extremely challenging. All things considered, Compass has designed as good a trial as has ever been run. Is that fair to say?

[[00:17:45]]
Yes, that is fair to say, and that's a good point. I will go further than that and invoke what actually was released yesterday, which was FDA's formal guidance to researchers, including drug companies who are conducting psychedelic research. They essentially outlined this exact trial design, the comparison of these three different conditions in order to characterise the dose response curve. I think they also stated that they would prefer there to be something like this phase 3 safety-focused trial comparing 25mg to placebo for safety. What this suggests to me is that Compass Pathways when designing these trials was in close contact and collaboration with the FDA. Given that they seem to be adhering to standards that are explicitly stated by the FDA, I think that probably will give them some latitude as the FDA is evaluating their results.

[[00:19:15]]
As you commented, the consensus seems to be that the COMP360 data isn't looking as good as initially hoped, at least based on the phase 2 data. The absolute improvement on the primary was fairly small – less than 10 points – whereas Spravato [esketamine] has demonstrated 13-14 points at four weeks. I know there are always caveats with cross-trial comparisons, but what do you think happened with the pivotal phase 3 programme? What explains the efficacy drop vs the prior phase 2?

[[00:20:04]]
In the wake of the phase 2 data and I think before that as well, there was confusion over what type of psychosocial approach would be used. That's one candidate reason, involves the psychosocial approach.

[[00:20:21]]
What do you mean by that?

[[00:20:23]]
Psychedelic therapy is unique when it comes to pharmacotherapy because the psychoactive effects are extremely psychological and they bring up mystical experiences, but also emotional processing that is more akin to psychotherapy, psychotherapeutic processes. Following a dosing session, participants are often moving through and navigating new emotional material with support facilitators or psychotherapists afterwards. That time, that space that is taken up by individuals, by patients exploring their life history and the roots of trauma or depression or these sorts of things, that may be quite pivotal to the ultimate efficacy of the therapy holistically. I know that psychosocially, there has been movement in the approach that Compass Pathways has taken and standardised. They believe that a more minimal psychosocial approach that essentially takes out psychotherapy is not quite as directive psychologically as a lot of psychotherapeutic modalities are that that is more conducive to approval because it doesn't complicate the inference that the FDA has to make about what is driving the therapeutic effect.

[[00:22:24]]
Essentially, the drop in efficacy could be explained by a more standardised, lighter-touch, adjunctive psychotherapy vs what was used in the phase 2.

[[00:22:38]]
Yes.

[[00:22:42]]
My understanding is that the phase 3 still required adjunctive psychotherapy. What exactly changed there?

[[00:22:52]]
Different researchers have different definitions of what psychotherapy constitutes. They would argue that they're not using psychotherapy. I'll say that when you look at… They've codified actually their approach in what they call the Compass Psychological Support Model. They define this by three principles, trust and psychological safety, present moment focus and self-direction and autonomy. The self-direction and autonomy referring to the patient directing their own ideation and expression. That is to say that there are elements of psychotherapy, but it's much more minimal within this model. What's directive and some researchers would say psychotherapeutics still within this Compass model is that they're providing space for individuals to explore the meaning of their psychoactive experiences. Beyond that, I think that they likely have really encouraged the clinicians that have worked with them to maintain that light touch. That could have impacted report, it can impact the efficacy of the psychosocial component.

[[00:24:16]]
In essence, this change was designed to help the FDA delineate between the drug effect and efficacy that is attributable to adjunctive psychotherapy. Presumably, this would also help with real-world uptake, because a 6-8-hour monitoring session plus additional adjunctive psychotherapy after the fact is rather burdensome on the clinics?

[[00:24:45]]
Yes, true.

[[00:24:48]]
The fact remains that the absolute improvement in MADRS was fairly small – minus 7.6 points in the COMP006 study and minus-6.3 points in the COMP005 study. Spravato was minus 13-14 points. On balance, one can argue that control-adjusted efficacy is comparable. It was minus 4 points for the COMP360 25mg arm, which was broadly in line with Spravato. However, the former used a 1mg active control vs an inactive placebo with the latter. Once you make all the adjustments to draw a more accurate comparison, how are you thinking about the treatment effect of COMP360 vs Spravato?

[[00:25:55]]
It's a good point. I'm just looking at this data on my own end. I think it is very interesting. For example, in the esketamine, ASPIRE II, the Spravato, ASPIRE II phase 3 trial, for example, their control condition delivered a 12.4 unit reduction in the MADRS, the depression scale. Whereas if we do the math in the COMP5 and the COMP6 [sic], the control effect was about 2.7 and 3.8 units. These are very, very small effects. Part of this actually is consistent with what's been observed by other researchers about the unique circumstances of these psychedelic or psilocybin trials in producing very low placebo effects. It's interesting because both the Spravato and the COMP360, they both should be affected by the same dynamic where people are aware of the obvious psychoactive effects of the experimental treatment and they can easily deduce whether they're in the control, and therefore, they might not report and experience the same antidepressant effects. One explanation for this could be just focusing now on the control is that, again, in the Compass Pathways, there was a more minimal psychotherapeutic design, whereas with the Spravato trials, there was treatment as usual, I believe, which suggests that there is some psychosocial aspect to it. I always imagine that I don't know the specifics of the psychosocial approach, but that could be delivering that 12.4 unit reduction, much greater reduction in MADRS than what is experienced in the Compass trial.

This is maybe honestly an issue because not only do you find a small delta between the control and the experimental, but I think it's reasonable to state that the placebo effect, the control effect within the psilocybin trials is more muted than really a robust placebo effect should be. Focusing now on the actual experimental condition and comparing between Spravato and psilocybin, it does show that a six- or seven-point reduction is just not nearly as substantial, obviously. I'm not sure exactly what is contributing to this. I do think that in the real world, if it is the case of the psychosocial condition is delivering a lot of the efficacy synergistically potentially or additively. I do think that this therapy in practice could actually be much more powerful because people won't have their hands tied. I don't know about the ability of a primary care physician who's administering this, but others, they might want to extend themselves to try to address some of the mental health concerns that people have. If that is the reason, it could be better.

The other consideration is that, again, it could be the case that there's a lot of variability. In my work, I want to do a study on our centre data looking at variability in the effect size that people experience, the treatment effect size. So far, what I've seen suggests that there is substantial variability. That would mean that there are treatment-resistant people in this population that are experiencing extremely large effect size changes. Then there are people who are experiencing more minimal. Actually, I will say this, too, in my work with psilocybin, what has been apparent is people differ so much in the intensity of the drug effects at 25mg. That suggests to me that there could be in the future, the importance of dose escalation or basically elevation should someone in the first session not experience a psychoactive intensity that could be therapeutic. That's another thing for, I think, investors in your audience to bear in mind is that in practice, physicians will have a lot more latitude to actually elevate the dose, I believe. That's something that you would need to evaluate and probably talk to physician or regulator. I believe that physicians will be able to do that.

[[00:32:12]]
You certainly can for Spravato.

[[00:32:19]]
I think that if this low average rate is a consequence of… We do see, obviously, in the data a high non-responder rate. If that non-response rate, that substantial non-response rate is a product of underdosing or potentially selecting patients that just aren't conducive to the sensitivity to the drug, then that could be ameliorated quite easily by merely elevating the dose, which is not possible within this structured clinical trial.

[[00:32:59]]
Would the FDA not want a better sense for what defines a likely responder, assuming that the drop in efficacy across phase 2 to phase 3 is a function of the variability in response?

[[00:33:22]]
They could be favourably and unfavourably disposed to that. Unfavourably, to your point, they could say this is not a very substantive effect size change. Favourably, they could say this is a condition where there are only two medications and people are suffering and nothing can be done. Even the small amount of efficacy is incredibly valuable just to make available to people. Also, we are aware of the variability in response, and there are some people who are responding incredibly well to this. We want to authorise making this available in prediction of some subset of individuals who can't be helped by other medications actually immensely improving and then a tolerable safety profile for others who don't necessarily.

[[00:34:27]]
Circling back to comparisons with Spravato, I know that in the COMP360 pivotal phase 3 studies, baseline MADRS scores were severe. 61% of patients had MADRS scores of greater than or equal to 31. In the 25mg arm, the length of the current depressive episode was more than or equal to two years in 57% of patients. How do these baseline characteristics compare to Spravato's pivotals? Are there any major differences in baseline characteristics? Could that inform the delta we see across these two therapies?

[[00:35:06]]
I think it could. I don't have all of that information in front of me. This was the case also with the comparison between the two phase 2. There's the Usona trial and also the Compass phase 2. Even though the Usona phase 2 recruited people with MDD and Compass phase 2 recruited people with TRD, the people in the Usona trial, they actually had higher moderate severity on average. These types of things happen. I also think that there's many people, like you said, in the Spravato. Also, what makes it interesting in the Spravato is that they were recruiting people with suicidal ideation as well. I need to look at the numbers, but with the treatment-resistant depression, I would expect, on average, them to have had the condition for longer, to have tried more treatments, obviously, on average, though I wouldn't expect their severity scores to necessarily be higher. It's really the chronicity of the disorder that's being targeted and that makes it really difficult to treat. I think it's meaningful, but I would need to look at every individual characteristic, which I just don't have in front of me.

[[00:36:46]]
As I think about the real-world utility of COMP360 vs esketamine, I've read that esketamine takes multiple administrations to reach the aforementioned efficacy threshold. Compass's management has noted that patients need up to 10 administrations of Spravato to achieve an equivalent treatment effect at week six vs 1-2 doses of COMP360. Do you subscribe to that view? Once we adjust for the different dosing, the convenience, the adjunctive psychotherapy and everything else, COMP360 appears to be, at worst, comparable to Spravato on efficacy and, at best, a better drug.

[[00:37:41]]
My reading of the literature is that a drug like psilocybin, like COMP360 is likely to be superior, particularly on the basis of the durability of the effect. Let me just read this. I will say that there is some data that is suggestive if you look at the ASPIRE I, ASPIRE II, the two phase 3 trials for Spravato, there is some data to suggest that there's a pretty rapid onset. People are improving following the second dose of ketamine. You can also see that there is this linear reduction over time that continues with the doses. It really does seem like according to the data that we have for the phase 3, it seems like Spravato is delivering about the same effect size as the COMP360 after just a couple of these doses.

[[00:39:48]]
So the earlier point that a patient needs up to 10 doses of Spravato to get a comparable effect size to COMP360 is perhaps somewhat exaggerated?

[[00:40:02]]
Yes, I think it could be exaggerated.

[[00:40:06]]
On the point of durability, we recently got the COMP006 study part B data, which was basically a six-month durability extension. 30% of participants who achieved a clinically meaningful response at week six maintained a response and went into remission – ie, a MADRS score of less than or equal to 12 following retreatment in part B. If you look at the COMP005 study, which was a single 25mg dose vs placebo, 25% of the 25mg patients achieved a MADRS reduction of greater than or equal to 25 in week six, maintaining the response through to week 26, with 10% going into remission. In the COMP005 study, 70% of the 25mg and 53% of placebo patients received a second administration after week six. In the COMP006 study, 58% of patients received retreatment after week nine. The majority of retreatment was between weeks nine and 14. How are you thinking about the durability of effect here, given this part B data?

[[00:41:34]]
I think that this is consistent with data that has come from, for example, Rosenblat, that the re-dosing does make a difference depending on when it's done. I think in some ways, this could be cost-effective and economical for this treatment where you have some portion of participants who respond to the first dose and then depending on how their symptoms progress, there is a re-dosing that then can effectively maintain or address relapse, which still might be less labour-intensive and costly than the default eight dosing sessions of the Spravato. I see it as consistent with some durability, I think. I again need to just look at these numbers I don't have immediately in front of me.

[[00:42:47]]
It seems Compass is pushing for an initial two doses vs an initial one dose, based on the COMP006 vs COMP005 data. After those two initial doses, how many additional doses do you think patients will need on average per year?

[[00:43:14]]
It's a good question. Let me just look at this really quick. I guess what I'll say about that is I do think it's possible that there's some unknown… That response is both a reliable consequence of a drug session, but it's also therapeutic response is probabilistic. That's because one doesn't exactly know what kind of experience will emerge or what kind of intensity even will emerge from a given dosing session. Because it does appear that the efficacy could be tied to the psychological nature of the psychoactive session, it could be that the more attempts as a participant and clinician understand the appropriate dose that is needed for treatment engagement that as those attempts increase over time, there's more likelihood of treatment engagement, if that makes sense. I think it's an individual nature, but I think the probability may increase over time as there's more re-dosing attempts. I'll just put that out there. I want to make just a conjecture. Obviously, we don't have this data. There is good data to suggest that people who dose three months later, who relapse after three months post the first dose, that they can experience symptom remediation that goes beyond what they experienced prior. I don't know at what point re-dosing is no longer needed and what that dosing schedule on average would look like in order to maintain someone's symptoms. I would probably say something like maybe four times a year, maybe three times a year.

[[00:46:05]]
In the Compass trials, the maximum number of doses on an annualised basis was four. Do you think the FDA may include labelling language setting this as the upper limit for the number of doses per year? Or, given the precedent that the company set with Spravato, would the FDA probably not want be so prescriptive with its labelling language?

[[00:46:55]]
I know that there's also concerns over the serotonin 2B receptor engagement and this concern over heart valve stiffening in relation to that. I'm not sure what schedule of repetition the FDA would have an issue with. I'm not exactly sure, to be honest, whether… Are you saying that Compass Pathways has set four as a limit and you're asking whether FDA…

[[00:47:28]]
Forgive me if I wasn't very clear. If you look at the COMP006 and COMP005 trials, the maximum that any patient was dosed on an annualised basis was four times. I don't think that's necessarily Compass setting that hard limit. That's just a function of how the trials were run and how the patients were treated by the investigators at the different sites. Does the FDA use that data to inform the label, or is it unlikely that it would be so dogmatic? As you say, there's a huge unmet need, a huge variability of response and a clear deepening of responses. My understanding of the Spravato label is that there's a wide range of doses and frequency you can administer this drug at, given the severity of the patient. Which is the more likely scenario – physicians will be held to a very tight and strictly defined dosing protocol for COMP360, or it will be rather loose and up to the discretion of the treating physician?

[[00:49:05]]
I would think the latter. Unless it goes into the REMS, but I think in general, there's a lot of discretion and latitude that's given to the physicians in what they're dosing. I think it's interesting actually that, implicitly, there was no limit on Compass Pathways in the phase 3 trial. That's actually where I would expect the FDA to be more conservative. I think that might be also suggestive of them not necessarily placing a limit in real-world application if they didn't in the phase 3 trial.

[[00:49:43]]
Because it designed the phase 3 in conjunction with the FDA.

[[00:49:48]]
Right. They essentially cleared that phase 3 trial, and it would have been within their right to essentially specify a clinical hold on the trial if they felt like that was dangerous to administer it more than four times. Again, I think the only rationale for them to limit the re-dosing within one-year period would be this 2B activation and heart valve stiffening risk. I think there's no real data to suggest that at that schedule, there would be a substantive risk of harm.

[[00:50:39]]
On the point of safety, I think suicide ideation was only seen in 1.4% of patients in the COMP006 25mg arm vs 2.8% in the 1mg arm. There was one suspected suicide, but that was deemed unrelated to the treatment. Beyond that, does anything jump out to you in the data? Do you have any concerns?

[[00:51:06]]
No. It is interesting though that particular one that you stated, where the suicide was ruled out as being related. That's something maybe to double-click on. I think one would have to be speculative to make any causal association or really adjudicate whether this was related or not, but that is something that may come up with the FDA. I know that…

[[00:51:42]]
If my memory serves, I believe the suicide event happened more than eight weeks after the last treatment dose, and the patient was very close to the physician or psychiatrist in question. Compass seems to believe it was definitely non-treatment-related.

[[00:52:14]]
That's fair. I think that's consistent with what I remember, too. An interesting yellow flag was when they stated that the patient was very close to the clinician. Obviously, I'm running these trials with psilocybin. If it was the case that the clinician at that point was very close to the patient at that point in time, it could have meant that this was a patient that was undergoing psychological difficulty or emotional difficulty for some longer period of time. That's a relationship and added support that might have been applied at an earlier point closer to the actual treatment and that the patient was being supported for that longer period of time, starting at an earlier time point that might be more costly for the treatment itself is the only concern that I would see there being some probability of the FDA wanting to examine more closely.

[[00:53:26]]
I want to talk about DT120, which is Definium's LSD therapy. There, we saw a placebo-adjusted improvement of minus 8.1 in MADRS. There was a rapid and durable reduction in MADRS starting as early as week one, with a negative 14.2 placebo-adjusted reduction that was sustained until week 12 with a negative 7.3 placebo-adjusted reduction with a single 100mcg dose. There was clearly some waning of efficacy over the 12 weeks. We know 24% of patients achieved remission, ie, a MADRS of less than or equal to 12. 35% of patients demonstrated a response, according to the more stringent metric of more than or equal to 50% improvement vs the 25% used in the Compass trials. Could you unpack these data? How does this compare to COMP360?

[[00:54:26]]
I think it's likely in this case, within this clinical trial that there may have… One explanation for this pattern of results could be that there was a heavier psychosocial approach used, and that could account for the higher placebo-related efficacy.

[[00:54:52]]
I believe there was no psychotherapy in the DT-120 study. There was a monitoring session, but no adjunctive psychotherapy.

[[00:55:04]]
Let me actually look at that.

[[00:55:08]]
I want to say that's one of the key differentiators for DT-120. Yes, you've got a rather long monitoring session, which doesn't really solve the logistical issues associated with COMP360, but it does do away with the psychotherapy component, which means it's easier to attribute efficacy to the drug.

[[00:55:29]]
Interesting. Fair enough. Where did you get that information about it? I know that you've been doing your own investigation, but is that not printed within the press release, Definium's?

[[00:55:46]]
I assume so. It's certainly disclosed in Definium's corporate material more broadly.

[[00:55:54]]
That's interesting. I suppose one has to be a little bit critical also about the degree to which psychotherapy or these psychosocial aspects are completely omitted when they say that only monitoring visits are used, but I don't actually have a lot of insight into that. That could be the case. The other thing that I would say is that LSD obviously delivers a longer psychoactive duration, up to nine hours. I did look at some point at their plot for the percentage of people who were released or who were discharged from the session at different hours. They were satisfied that a substantive percentage were discharged at five hours. I would say that I think it's the view of people within my centre that it's likely they were discharging people who might have still been undergoing some psychoactivity just based on the discharge criteria that they used and the distribution of that. I think that's to say that that's one thing to keep in mind when it comes to Definium and LSD is that this could pose an issue in terms of cost.

Then the flip side of that is it is possible that this is, in some sense, a pound for pound more intense experience and a higher dose or higher, I would say, receptor engagement or neurological engagement associated with the 100mg vs the 25mg psilocybin. Something to consider, too, which could bear on the treatment response and could be buoyed in the case of psilocybin by a higher dose potentially.

[[00:58:12]]
Definium is studying DT120 in MDD, whereas Compass is studying COMP360 in TRD. Looking at the baseline characteristics in Definium's study, 80% of patients treated with DT120 had severe MDD with MADRS scores of greater than or equal to 31, and 55% had two-plus past antidepressants. Is there more overlap than originally anticipated between the MDD and TRD populations in the respective studies?

[[00:59:22]]
55%, it's a little over half of the sample. I would say that that still is reflective of a pretty substantial difference between the two in terms of treatment resistance, the two samples.

[[00:59:36]]
Do you remember what the prior treatment mix was in the Compass study?

[[00:59:42]]
In the Compass Pathways in the TRD population, I believe a core property of treatment resistance is that a patient has tried a full dosing regime or a full regime of at least two treatments.

[[01:00:02]]
I'm looking at the COMP006 data. 64.2% of patients had tried and failed two treatments in the current depressive episode, 25% had failed three, 10% four and 1% five. In the Definium data, 55% of patients had failed two therapies. If TRD is defined by a failure of three antidepressants, then having 64% of patients only failing two antidepressants in COMP006 seems a bit odd. How much overlap is there between the two companies' respective definitions of MDD and TRD?

[[01:01:14]]
I was fairly confident that TRD involved at least two attempted treatments.

[[01:01:24]]
Exactly, my understanding of TRD was exactly 2-3 failed antidepressants.

[[01:01:31]]
I guess I'm confused about why that figure would be 65%. I don't know. That might warrant some… I'm going to be curious to deep dive into that, honestly, because I assume that it was 100% in that sample.

[[01:01:52]]
In the 25mg arm of COMP360, 64% had failed two treatments, 25% three, 10% four and 1% five. In the 10mg arm, 74% had failed two, 23% three and 3% four.

[[01:02:10]]
It adds up to 100%.

[[01:02:13]]
Yes, but I'm not questioning the total as much as I'm questioning why 64% only failed two therapies but were defined as TRD. In Definium's study, 55% had failed two antidepressants and this was called an MDD population. Looking at DT120, 80% had MADRS scores of greater than or equal to 31, which is in the same ballpark as COMP006. Is Definium actually enrolling more of a TRD population than MDD?

[[01:02:33]]
Absolutely. I think this is consistent with what I was thinking, which is that it sounds like 100% of Compass Pathways sample had tried and failed at least two treatments. Then in Definium's sample, it was 55% that had tried and failed at least two. That is typical of an MDD population where you're doing your recruitment and some are going to be exhibiting that TRD profile. What they can do essentially is they can say to FDA, "Look, we think that we should get dispensation in order to prescribe this to a TRD population," but because it's only 55% of the sample… My understanding based on is that they would need to do a formal study and perhaps also write and submit a formal application for their product to be prescribed to the TRD indication. 100% of their sample within that particular trial would have to meet criteria for TRD. Obviously, to your point, they do have some data for individuals who have a TRD condition.

[[01:04:05]]
Unfortunately, we're out of time, so we have to end there. Thank you very much for your input. Clients, as always, for those listening or reading this after the fact, if you'd like to speak with Brandon in a private call or meeting, please let your relationship manager know. Thank you again.

[[01:04:24]]
Thank you so much, sir.

[[01:04:26]]
Cheers. Bye-bye.

Transcription ends at 01:04:27 of the recorded material