Good morning, everyone. Thanks for joining us for the Definium Therapeutics Inc. Fireside Chat at the Sixth Annual Novel Mechanisms in Neuropsychiatry TD Cowen Summit. With us from Definium, we have Chief Medical Officer Dan Karlin. Thank you, Dan, for the time this morning, especially to discuss your recent compelling top-line data from two trials, phase III VOYAGE and the PANORAMA studies of DT120 in GAD. Excellent efficacy and with very benign safety. Maybe could you walk us through the key efficacy, safety, and durability findings of that top-line data?
Yeah, absolutely. It's great to be here. Thanks so much for having us. We're incredibly excited to be in the position we're in now. Through this summer, over the course of just 10 weeks, we had three phase III readouts, two in GAD with VOYAGE and PANORAMA, and one in MDD with EMERGE, which read out first, of course. All three of these studies had very similar designs, which is that they had a primary observation period, which was a 12-week double-blind randomized period, and then they all had a Part B, which is ongoing for all these studies, which is an additional 40 weeks of observation with opportunities for open label treatment. So in total, these are year-long studies.
What we were able to do, though, was read out those primary periods and provide a little bit of data on the Part B as well for the folks who had gotten through six months in the Part Bs. What we saw across these studies is that we were treating people who were quite ill. We had people starting well into the severe range of the relevant scales. In GAD, we measured the Hamilton Anxiety Rating Scale. In MDD, we used the MADRS, the Montgomery-Åsberg Depression Rating Scale. We had folks starting on the HAM-A at around 28 in this study, which is again, well into the severe category. On the MADRS, we had folks starting at just about 35. So again, quite sick with MDD and GAD.
Notably, in each of those indications, the opposite scale, so in the anxiety patients, the MADRS, and in the depression patients, the HAM-A, we kept that pretty suppressed. We wanted to really demonstrate that we were able to treat specifically the disorder in question so that there would be no question that, oh, in the anxiety patients, did you just make their depression better, and the depression patients.
Right. The same about the two comorbidities.
Did you just make their anxiety better? By doing that, we actually raised the bar quite a bit for ourselves, because when you have people with comorbid symptoms, it can be a bit easier to move these scales because the scales, of course, have a lot of construct overlap. But we segregated out these patients into these discrete populations, I think very effectively. What we ended up being able to do was with a single treatment of DT120, with no adjuvant additional therapy, no psychotherapy, no other anxiolytics or antidepressants, we were able to move the primary endpoint in each of these populations between 5 and 8 points. We had in VOYAGE, a 5.4-point change versus placebo. In PANORAMA, a 5.1-point change versus placebo. In MDD, on the MADRS, we had an 8.1-point change versus placebo.
All of these measured, in the case of GAD, 12 weeks after the single treatment, and in the case of MDD, six weeks after the single treatment, which is pretty standard for MDD studies. All of those were highly statistically significant with a p-value of less than 0.0001. Again, we are talking about single treatment day, no additional treatment in the interval, and efficacy being demonstrated 6 to 12 weeks after that single treatment.
In VOYAGE, response and remission rates were 43% and 14%, respectively, at week 12. For PANORAMA, response and remission was 32% and 15%. How do these compare with the phase II-B, and how do you look at the remission rates in phase III versus phase II precedent?
Yeah. The phase II, we saw nearly 50% remission rates, which again, we weren't expecting to see.
Right. It was much higher than you thought.
It's incredibly unprecedented. While of course, we'd always like to see a higher remission rate, we'd like to see everyone who gets the drug be in remission with a single treatment, that would be great, that's not a realistic outcome for these disease states and is not something that we could even really hope for. What we see with the response and remission rates in phase III, though, is remarkably strong and more than can reasonably be expected from existing standards of care, particularly when you look at the gap between the placebo response and remission rates and the treatment arm response and remission rates.
When you're looking at odds ratios of nearly 4x across that gap between response and remission in VOYAGE, and 3x to 4x, 2.5x to 4x in PANORAMA, we're still quite pleased with this, and we think that this represents really clinically meaningful differentiation from the existing standards of care. When we look across these different studies, why is there some degree of wobble? That is just the nature of studies. There's the reason you do more than one study in phase III programs, and that's to make sure that you're accommodating the sort of artifacts of the individual study and the sites that you're using. The one big difference, though, of course, between VOYAGE and PANORAMA is that PANORAMA included a 50 microgram control arm.
Now, we enrolled that at half the rate that we enrolled the full dose in the placebo arms, and it wasn't an arm of analytic interest. But what we were able to use that arm to do was to say that if someone felt the effects of the drug, and of course, there are transient effects of the drug that people feel, they wouldn't know that they were in the arm of analytic interest or not. So it existed purely to confound the understanding of people as to which arm of the study they were in.
The functional unblinding.
Yeah. So, there is functional unblinding in these studies. Patients, participants feel the drug and can fairly reliably estimate which arm of the study they're in terms of did they get drug or did they get placebo. Now, we have about 30% of people in placebo allocation who don't accurately guess that they're in placebo, who either don't know or guess that they got drug. So that's a whole interesting other phenomena. It reveals something, which is that the idea of functional unblinding, which gets talked about a lot in our category, but of course, has been present in every psychiatric drug development program, though often in the history of psychiatric drug development, we just didn't ask, we didn't assess for functional unblinding. But that it gets quite a bit of conversation, and whether that's appropriate or not is up to the asker, I suppose.
Now across our program, phase II with our five-arm study, VOYAGE and EMERGE with two arms, PANORAMA with three arms, we've shown that we can reliably demonstrate a treatment effect in a number of different control conditions. So at the end of the day, the point of research is to try to learn something about the world, right? To try to make sure that what you're measuring as closely as possible represents some reality. By using these complementary study designs as advised by FDA and their guidance, we believe that we have demonstrated that independent of controlled condition and across complementary study designs, the measured effect of the drug represents a real effect of the drug.
Got it. Do you think that remission rates could improve during Part B as patients receive additional open label doses? When could we expect the full Part B data set from both studies?
Yeah, great question. So what we've shown already from Part B is that as we continue to offer open label treatments to people who are moderately ill or worse, so we have a triggered treatment condition. People have to be moderate on the scale or worse, that we're able to push the mild or remission. So because we treat it moderate, we're not actually treating to remission, but we're treating to mild or better. Across the study data that we've released so far out to week six, what we're seeing is a convergence of mild or remission at 60-ish %, so that with the availability of open label treatment, most people end up getting between one and three treatments in Part B, we're seeing that we can in fact induce mild remission in about 60% - 70% of the population we're treating.
Absolutely, we expect that trend to continue as we offer people more treatments, and as we get deeper into Part B, we fully expect that we will see that convergence on around 60%-70% continue. We haven't guided on exactly when we'll release the Part B data from these studies. You can reasonably well estimate from the data we've shown when most or all folks will have run through the full- year of observation. Of course, we fully intend to share a consolidated view of what this treatment can do, what DT120 is capable of doing with that full- year of the availability of open label treatments.
Have you discussed publicly what the retreatment criteria of the Part B is?
In terms of like, it's just the t hreshold between mild and moderate. So it's a MADRS of 20 or a HAM-A of 16, and both when we report mild or better, that's the threshold we use, and when we assess for retreatment, that's the threshold we use.
Is that part of the protocol? Is required retreatment part of the protocol, or is there sort of a physician and patient preference? Basically, can a patient stay in the study if they don't want retreatment, or will they de facto sort of drop out of the study if they.
Yeah, sure. The open-label treatment is protocol specified. Folks who remain in the study and qualify for treatment do receive that treatment. Of course, people can leave the study at any time. It is important for the sake of data integrity that we have that trigger treatment. Now, we give people a window so that if someone were to trigger the criterion for retreatment, they don't need to immediately go in and get retreated if things are going on in their lives that would prevent it. So they have some time in which to make that decision, and there's always PI discretion in these things. So if for some reason a PI decided that someone wasn't appropriate for open-label treatment, that would be a reason for discontinuation from the study.
But we wouldn't, of course, in the conduct of research, we never make anybody do anything. It's all voluntary, right? Folks are in the study because they want to, and they stay in it because they want to.
Based on, and of course, please do not discuss non-public information, but based on what you guys have seen from the phase IIs that you've conducted and the phase III experience and those curves on the primary endpoint, what is Definium planning on for real-world retreatment cadence?
Yeah. It's an interesting question, and we get asked things like this a lot, which is, what do we intend for the real world? I think we take a slightly different view of this, which is that we are developing a drug. We're developing it with generation of the best evidence we're possibly capable of generating. We design our studies to be maximally interpretable. We want FDA, of course, as a primary audience, but providers, payers, patients, all to be able to look at the work we've done and know as much as they can about how the drug performs and the conditions under which we tested it. What we expect is that ultimately, if we're approved, that our label will reflect the conditions under which we tested the drug, and that those conditions can be considered the sort of minimum needed for safe and effective use.
So in the open-label treatments that we administer in Part B, we say they can't be more frequent than monthly, and people can have up to a total of five treatments over the course of a year. We think that those conditions will likely represent what sort of encircles most people, that most people wouldn't need more than five treatments in a year. We see most people in the studies getting one to three treatments over the course of those six months, and the treatments seem to space out over time, that people need less frequent treatment over time. What we'd like to be able to do is give guidance to the prescribing community that says, "Here's how we tested it."
Then there'll be a lot of treater discretion. There'll be a lot of clinical judgment that goes into the use of the drug. Over the course of time, again, if approved and out in the real world, as clinicians gain experience with the drug, practice standards and patterns will be developed and will emerge from that practice. I'm ultimately hopeful that the sorts of things like psychotherapy that have been used historically alongside the drug will be additionally researched by the treatment community, and we'll start to learn a lot more about how to maximize the effects of the drug, which is not what we did in these studies. We weren't aiming for maximization. We were aiming for demonstration of the minimal conditions for safe and effective use.
Getting to the million-dollar question, and this is a broader field-wide question. What in the data, Dan, is emerging to identify optimal responders? Is there any emerging predictor of response that you're finding in the GAD studies or across your depression studies that identify a responder to DT120 or a responder to this class of drugs? Is there anything being done in, any signal emerging, even in the field, an independent study? Because I get this question all the time, and I have no data points to discuss.
The idea of a priori prediction of response in psychiatry is incredibly appealing. And I get it. In my history in work I did in big pharma, we were asking the same questions 15 years ago. And in the course of SRI use now of a 40-year course of the hundreds of millions of doses of SRIs prescribed, there's always been this question, can you predict who will respond? Even back into the history of psychotherapy and psychoanalysis, the question was, who will respond?
We forget biomarkers. We just want what is a responder shape.
Yeah. I wish I could tell you that we had the answer to that now. We have looked in our studies if we could identify anything, whether it be demographic or aspects of the MADRS or the HAM-A that might be predictive, and I can tell you as of today, we don't have the ability to a priori predict. We're hopeful, because what hasn't existed before is a drug that over the course of, say, six months of treatment, could induce mild or better scores in people who start very severe in more than half the population being treated. So we will keep looking. We will keep trying to figure that out. We absolutely expect that the community of folks engaged with psychedelics will be asking that question and trying to figure out answers.
As of today, there's nothing that we could identify that would a priori predict responder.
In both of these GAD studies, average time to clearance was about six hours. The majority of patients cleared by hour eight. What did the end of session clearance checklist assess, and how would that checklist translate directly into real-world practice as Definium understands it right now?
That's an excellent question, and thanks for pointing out the timing. So we have a median and mean clearance rate that's about six hours, right? That is clearly a really good number for a clinical workday. We're really pleased with that. Across our studies, we're looking at 97% cleared by hour eight on the checklist. You asked about what's on the checklist. The checklist is something that is really a straightforward assessment of capacity, and so we talk about decision-making capacity in medicine a lot. Fundamentally what that's asking is, do you know who you are, where you are, when you are? Are you able to engage with the world in a way that is approximately at your baseline? So are you having ideas that would make it difficult for you to exist in the world?
Are you having perceptions that would make it difficult for you to exist in the world? It is fundamentally intended to be something that can translate easily into clinical practice. It takes less than a minute to administer, and it is something that along the way in our development program, we have been fortunate enough to be able to discuss with FDA at each of the study submissions so that we have got, really at the language level, at the word choice level, a process that has led us to this final checklist that we are using that was deeply engaged with FDA.
Holistically, how should investors think about sort of the real-world status of a discharge-ready patient? Should we be thinking about outpatient surgery? Should we be thinking about sort of post-ketamine discharge or post-ECT discharge?
All of those examples hold. In any case where we induce an alteration of consciousness in medicine, we have to set a threshold for saying, okay, this person no longer needs clinical supervision. Outpatient surgery is a fine example. You can imagine outpatient non-surgical procedures that people have done, where there's some degree of consciousness alteration to facilitate whatever is happening. Then there's a period where someone should be under clinical supervision, which we absolutely agree with DT120 is the case. Then people get better, and they get back to a place where they're able to function in the world. So that is a perfectly good analogy. Yeah.
I'm going to save the super unfair questions for the Chief Medical Officer for these last, or I have saved them for the last five minutes. Audience question, what does the Definium team see as the biggest risk ahead of the pre-NDA meeting? I think the investor question goes to what's going to be discussed at the NDA. What are the topics of most highest importance and criticality discussed at your pre-NDA meeting that's coming up?
Excellent question, and it's very much on our mind. One great thing about working with the division of psychiatry at FDA these days is how we've been able to have a rhythm of conversation, a frequency of conversation, both under the breakthrough status that we now have for GAD and MDD. I think just because, as we've heard from the division, they're incredibly enthusiastic about the category and want to be thoughtful about balancing safety and access, which is always the challenge. So the framing of the question, that there's a risk going into the pre-NDA, we.
I think it's just always clinical regulatory risk.
We really don't experience our conversations with FDA as risk points. It's all about charting a path forward, and it's about establishing the right setup for a submission that enables them to do their job as well as they can. There are always details of an NDA submission. Of course, it's a really complex and lengthy undertaking to write an NDA. But again, throughout this program, through our clin pharm studies, our safety studies, through our toxicology work, through our CMC work to make sure that we've really crossed all the T's and dotted the I's on manufacturing and controls. Then through our clinical efficacy research, everything was designed to be maximally interpretable and to sort of chart a clear path to a label that would be one that would enable real-world prescribing. We're excited. We're excited to go talk to FDA at our pre-NDA.
I wish I could tell you what a risk was, but we really don't experience these conversations as risk-bearing. They're opportunities.
Given the positive EMERGE data in MDD, and the strength of that data, right, Dan? Combination filing a possibility here? Can you file for GAD and MDD at the same time?
It's absolutely a possibility. It may not be the best course of action, and that's one of the things we'll discuss with FDA when we see them for the pre-NDA, is trying to figure out the right way to file for two indications. It may be sequential and overlapping, it may be in the same filing. That's something that we'll need to discuss with them, and it's certainly on the plan to discuss with them to be sure. This is as much about enabling the work they have to do as it is about our work. We have the data we have. We've got a second MDD study out there, in process that's recruiting incredibly strongly, and we're really excited about what that will be able to show.
This is one of the things that we have the opportunity to go get alignment on, and make sure that we and leadership at FDA have the same vantage on the thing.
Is the commercial strategy for DT120 at Definium Therapeutics going to be buy and bill? That was an investor question.
That's a great question. We know that there's a pretty even split in the world for SPRAVATO for buy and bill versus white bagging, which is for folks who aren't totally read in on the different mechanisms for getting drug to patient. The two major strategies are obviously clinics buy and hold drug, and they act as the dispensing entity to the patient. The other way to do that is that the practice writes a prescription, the HCP writes a prescription to a pharmacy and gets drug delivered in the name of the patient, so already labeled with the patient's name. Then the practice can hold that and then finish the delivery to the patient.
We think early on, as people get experience, we'll see more white bagging, and that over time, as clinics start to figure out their own rhythms, that the buy and bill will likely make more sense. But again, this is very similar to the question of sort of real-world practice patterns. Our job as the manufacturer is to make whatever path works for a clinical practice available to them. Through the course of establishing REMS and then ultimately, the sort of distribution models that we support, we want whatever works for a practice, both clinically and, of course, financially, to be available to them. So we'll strive to make the broadest range of possibilities available to practices.
Great. With that, we are at time. Dan, thank you so much for the time. Next up in the conference, we have Seaport Therapeutics, which will be led by my colleague, Joe Thome. Thanks, Dan.
Yeah. Thanks so much for having me.