CEO Amit Etkin, welcome.
Thank you.
Maybe for those in the audience less familiar with the story, would you mind talking about the Alto thesis and then maybe talking about your pipeline and then milestones we can expect over the next 6-12 months?
We are a late-stage psychiatry drug developer. We take a precision lens to things. In other words, really try to understand using different ways of measuring biology in people who are the right patients for any particular mechanism, given the tremendous heterogeneity that we have in psychiatry as we define diseases now.
We also use those kinds of biomarkers to understand what drugs do in terms of how to dose them, and therefore also what kind of people benefit from that, and have applied it across a number of indications, including different aspects of depression, treatment-resistant depression, adjunctive treatments in depression, bipolar disorder, focusing on bipolar depression, and cognitive impairments in schizophrenia. It's really a very broad green field out there as far as I'm concerned in psychiatry.
While there are obviously treatments for some things, there's many areas with either no treatments, and even those areas with treatments, often they are lacking on a number of different fronts. Lots of opportunity. We are leveraging multiple different assets, multiple different drugs against those, and that includes ALTO-207, which we're developing in treatment-resistant depression. That's in a phase II-B, and we anticipate a phase III program soon. That's a drug that has a very unique efficacy profile we can talk about in a little bit more depth. ALTO-300 is in adjunctive treatment in depression.
That uses an EEG biomarker to select patients based on an approach we developed out of machine learning and validated as a way to identify those patients that we then linked to the mechanism of the drug itself. That's a phase II-B readout next year. ALTO-100 is in bipolar depression, and that's a potential first-in-class mechanism, enhancing neuroplasticity for patients with evidence of impaired neuroplasticity, as evident by impaired learning and memory, which requires neuroplasticity. A phase II-B readout next year. A lot of opportunities coming up, really, really meaningful clinical milestones. Now it's just kind of heads down execution on all these programs.
Great. Maybe we can start with ALTO-207 phase II-B data in second half 2027. Bigger picture, when I think TRD, it's really just SPRAVATO out there, approved and used. I guess [NVX] isn't really used.
Right.
Maybe to start, walk us through why you are highly differentiated compared to SPRAVATO, but not only that, to other TRD drugs in the pipeline as well?
Yeah. I think actually it's important to start by even understanding what TRD really means, because I actually think that the TRD definition that people have sort of moved to in their minds is actually perhaps more narrow than the real opportunity out there. TRD means two or more treatment failures. Two or more treatment failures includes a lot of the people in clinical practice who would get an antipsychotic, an adjunctive antipsychotic.
They're also the people, by the way, who will end up in clinical practice getting Auvelity. They're also the people who will get esketamine at the higher end of failures, or should they be approved at some point, psychedelics at the very extreme end of number of failures. That range, though, is the range that we're going after. What differentiates ALTO-207 here are a number of really critical factors.
Number one, this is a dopamine agonist, direct agonist strategy that has multiple trials of the pramipexole component of this fixed-dose combination of pramipexole and ondansetron that has demonstrated outsized efficacy across these trials. For example, PAX-D, which was a 150-person TRD study out of Oxford that published last summer, found a Cohen's d effect size of almost 0.9, three times larger than the typical successful antidepressant drug.
That kind of an efficacy signal is something that we've not seen and doesn't have all of the baggage, and certainly none of the REMS programs that an antipsychotic would have, or for that matter, that esketamine would have. It's a drug that's shown efficacy consistently. We've, in fact, meta-analyzed all of the data that have been published on pramipexole historically, and even with low-dose studies included, the average effect size is 0.64.
That's a very different profile than everything that's out there. Unlike partial agonists or antagonists of dopamine receptors, antipsychotics, there's no metabolic effects. There's no movement disorder effects. You can get tardives for years after a short course of an antipsychotic. Not the case here for a dopamine agonist, full agonist like pramipexole.
That positions a drug that is highly effective based on these investigator-initiated trials, and by combination with ondansetron, should be much better tolerated than pramipexole alone as a fixed-dose combination with a novel modified release that further better matches these two components, and for us, opens up a commercial landscape that's not just differentiating, as you said, from esketamine and potential future psychedelics, but really differentiating from all of the available options out there and displacing a very large market around antipsychotic medication, for example.
Okay, great. The phase II-B started a couple of months ago or so.
Yeah.
Maybe talk about the study design and then the powering of the study. You've mentioned Cohen's d of 0.9. I'm assuming you're not assuming.
No, we're not powering for perfection.
I would love to know-
Yeah.
...kind of what you're assuming drug and placebo behavior.
Yeah.
Great.
I mentioned PAX-D. PAX-D was an adjunctive study. You take people who failed, they stay on their drugs, and you add, in that case, pramipexole or placebo on top. Obviously, it's a very successful study in this population. That is also the theme that we are then following. It is designed as close to PAX-D as possible. Two to five treatment failures, you stay on that drug that you failed, you add pramipexole on top. We have an eight-week treatment period that includes a two-week titration. Part of the challenge and opportunity with pramipexole is how to get it to the right dose.
We know that we need to get to a higher dose. Because of a lot of data in the field, we've shown that there is a dose-response relationship amongst randomized trials, and that's been argued also amongst case report data. You want to get higher. Our target is 3.2 mg of pramipexole together with 15 mg of ondansetron, and that's dosed BID. Patients go through a custom titration schedule we've established that would actually look very much like the commercial starter pack that they would end up getting as a way, again, to bridge the trial to clinical practice.
It's 178 people in the trial. That means it's powered at 80% power for an effect size of 0.45. 50% power is around 0.3. In other words, anything from standard of care on up, we'd be in a position to detect statistically. To put a more direct point on it, any effect size reported in prior trials of pramipexole would be significant in that sample size.
Very well powered. If we end up getting the 0.9 effect size of PAX-D or the 1.1 effect size, which is over an eight-point MADRS delta that Chase Therapeutics found in their phase IIa study with this combination, we'll obviously be throwing a big party, and everybody's invited here. That's a great outcome, but it's not our base case. We needed to power this conservatively, make sure that we're in a position to detect a strong effect, and get the drug into the clinic where I think it'll meet a tremendous need.
Right. Simultaneously, you're planning for phase III, kind of "at risk," starting it up the first half of next year before the phase II-B reads out. Why is that first? Secondly, how many phase IIIs are we talking about? Just one? Because I believe the phase II-B is designed as a pivotal registrational quality study.
That's right.
Yeah.
That was part of our design thinking for the phase II-B. It's a unique situation, right? That we're in a position where we already know a lot about the drug, the active drug here, pramipexole. There's been a lot of studies. There's studies that are yet to come out that came out even in preprint this fall, further supporting, even since PAX-D, the efficacy of pramipexole. It's less a question of efficacy, that usually you wait on your phase II-B before starting the phase III. It's more a question of have we solved tolerability with this approach?
We'll know that already as part of phase I studies that we're doing, as part of blinded data analyses on the phase II-B, and that opens up the possibility of starting your phase III earlier and therefore putting yourself into an earlier position to seek potential regulatory approval. That's our strategy here. We'll expect to go to the FDA to seek alignment on a phase III design to start by early 2027.
That'll be an important outcome, a 2026 outcome here, as a regulatory catalyst event, if you will. That then allows us to compress the timeline to then get to NDA, and we're funded through 2029. We're funded through all of these activities, funded through the phase III. To your question, it would just be a single phase III then, given the design of the phase II-B being potentially registration supportive.
Understood. Yep. Then, yeah, one component is trying to mitigate or offset the side effects of pramipexole. You're using ondansetron very carefully in a titration schedule. Ultimately, what kind of nausea, vomiting rates you think we can expect? What does, for instance, pramipexole alone show in prior clinical studies?
Yeah. I think the rates of nausea and vomiting are perhaps less of the kind of operative rate as the rate of dropouts due to Adverse Events. You can get a little bit of nausea if you don't drop out. If you like the drug and you stay on it's fine, right? The dropout rates have been massive with pramipexole alone. Again, with the PAX-D as an anchor, 15% of patients dropped out due to AEs, incrementally more than placebo in the pramipexole arm. 20% versus 5%. Most of that was nausea and vomiting. If you can get that down, and an anchor point might be where an adjunctive antipsychotic is.
For example, CAPLYTA, a very successful adjunctive antipsychotic, has about a 9% dropout due to AEs on average across its phase III trials. That gives you a very commercially competitive product, and I think we'll be in a good place to get there with this combination, with the modified release, furthermore, PK matching two drugs with very different half-lives. You have a long half-life on pramipexole, a short half-life on ondansetron. On pramipexole is what's driving the nausea and vomiting, now if you can better match them, you can better counteract the nausea and vomiting. All of that, plus the titration schedule itself, should lead to strong tolerability.
Okay, something less than 15% placebo-adjusted delta-
For sure.
...dropouts due to AEs.
Yeah.
Yep. Let's just say hypothetically, it was like a Cohen's effect size of, I don't know, 0.25 or something lower than what you're 80% powered for. Do you still commit the phase III studies? The root of question is, at what point do you kind of discontinue the program entirely due to unfortunate results from the phase II-B, or do you commit phase III regardless?
We will have already started a phase III by the time we learn about the phase II-B. There's really nothing historically indicating that that's what you would see. In fact, the data historically indicate that maybe we're being too conservative in our estimates, right? If you take again that meta-analytic effect size of 0.64, including a lot of lower dose studies, we would be the second highest dose study, right under the Chase 4.1 mg that got a 1.1 effect size. Maybe the 0.45 effect size powering here is going to turn out to be quite conservative. Maybe that would've meant we could run a smaller study. Obviously you don't know that until you run it. I think it's better to be conservative than powering to perfection.
Understood. Phase III, again, starts first half 2027. What would the timeline be as we think about potential approval timeline? The phase II-B seems like it's only going to take a year or a year and a half. Will the same be for the phase III as well?
Yeah. We'll guide obviously more on timing once we have the final phase III design, once everything is signed off with FDA for that study. I don't want to get ahead of things that aren't yet finalized. However, one of the things we have messaged, again, with FDA feedback, is that we're expecting two drug arms, two levels of pramipexole. The FDA likes reasonably to look at dose-response relationships. We'll have a lower dose in there as well. That will certainly increase the overall sample.
Part of our goal is also to try to maintain as close to 50/50 drug and placebo. Even if you have multiple drug arms, they all would go into that 50% or so drug component. That helps maintain expectations for patients. That will lead to a larger study. Obviously, that would mean either more sites or it takes longer. We'll be more specific once we've finalized that. I do expect to be some dose response information in addition to the efficacy overall at the 3.2 mg dose.
Yep, very clear. Notorious in depressions, high placebo responses. How are you mitigating against a high placebo response or professional patients and so forth?
Yeah.
Maybe talk about that dynamic.
Yeah. We actually haven't had an issue with high placebo response per se. That comes from various other sources that you can control well, in terms of how you do your assessments and how you structure your trial and how you message expectations to patients. I think where the real risk is, that's a systemic risk across our field, is professional patients. We've seen it in multiple major pharma and biotech readouts over the past couple of years. It's always been an issue, but it's become more and more of an issue of late for a variety of structural reasons.
We've been doing is being very explicit, very directly messaging what our internal process has been to prevent as much as possible professional patients from coming in. Things like requiring medical and pharmacy records. All of our trials are adjunctive. You should have a record for having the diagnosis and been prescribed the drug, then we need to be able to measure it in your drug or in your blood or in your urine. We do that repeatedly throughout the course of treatment as well.
That tells us that we're getting the right patients who are showing the right behavior and not just in a trial because they want compensation. That's a huge risk, especially for, as it turns out, monotherapy trials, which are still very common in this field. We've seen in the data that we've reported even earlier last month, interim updates on the compliance pattern in both ALTO-100 and ALTO-300 ongoing studies, where we see 100% compliance in ALTO-300 with their underlying drug. You can't measure the drug itself because it's too short of a half-life, but you can measure their underlying drug, 100% compliance throughout double blind and open label.
With ALTO-100, where we've been measuring the PK of the drug in batches of patients rather than waiting all the way to the end of the trial, we've seen 97% compliance. These are really, really strong numbers and require you to have a very clear gate, which we run as a sponsor-led eligibility review that has to okay each and every patient as they come in. It's just a systemic issue with our field. I think we all have to act differently to prevent it and change the incentives. Until that changes, I think each company has to be very responsible for its own trials, and we feel needs to be clear in messaging how well that's worked.
Thanks. Maybe last question, then we'll shift to your other programs, lifecycle management opportunities for ALTO-207. Let's just say TRD in second half 2027 was successful. Do you feel compelled to move to MDD or not? It sounds like no, you don't need to, bipolar depression, other indications?
Yeah. Look, there's good data, even similar data, I would say, to major depression in bipolar depression, that certainly could be a next step. I think, again, starting with where I started this discussion of what do we really mean when we say TRD? There's ideas of much more resistant end that I think the field has in its narratives around things like psychedelics push to, TRD actually will cover the kinds of people who an insurance company would pay for getting a branded drug.
I think that really is a very wide spectrum, even under that label. By an adjunctive program as we are doing here, you don't have to come off of your existing drug. It fits much more the pattern that clinicians like myself use when you prescribe a new drug to a patient, especially one who's already failed a few drugs, which is, let's not rock the boat with their underlying drug. Let's add the new drug on top, see how it works, and then make a decision with the underlying drug.
Great. Thank you. Shifting gears to ALTO-300, agomelatine. There's a data readout now in first half 2027, phase II-B. Originally, it was guided to mid-2026. Maybe talk about what happened exactly, why the delay?
Yeah. The delay is really coming out of that quality filter that I mentioned, making sure that we have the right patients and are preventing professional patients from getting in. When we initially estimated what the readout timing would be, it was without a clear understanding of exactly what that ultimate recruitment rate is with that filter in place. We saw progressive acceleration as we restarted the ALTO-300 trial with this new approach, now having a much better understanding of where that recruitment rate puts us, and that trial got pushed out a little bit.
Obviously, I think we would all appreciate making sure the right patients are in a trial and giving us the best chance of success is the most important factor here. The other important thing to remember is that because we've now had that experience, the way we're projecting ALTO-207 has already accounted for the recruitment rate that imposing all these professional patient filters achieves. That gives us even more confidence in the ALTO-207 timeline for second half 2027 for that phase II-B.
Got it. Fundamentally, agomelatine is approved in the EU, was never approved in the U.S. Maybe walk us through, remind us what happened, and why you think you can get it approved this time around?
agomelatine ALTO-300 followed, frankly, much the same pattern as every all-comer antidepressant, which is that some trials work and some trials don't, and you need two positive ones at the same dose to be able to get it approved. Novartis developed it in the United States, Servier developed it in Europe, and Novartis saw, frankly, very much what Servier saw, which is they had excuse me, studies at different doses, 25 mg and 50 mg at separate arms. One phase III hit for 25 mg, another phase III hit for 50 mg, but they didn't get two at the same dose.
They would have to then do a whole new phase III to be able to get that second study. They were running out of composition of matter timing, they didn't proceed further with the program. That's exactly why the precision lens that we're taking is so important. Because if you can be much more consistent in the patients that you are identifying and that you think may have a better chance of success, then you can bend those odds, and you don't need to wait for two positives and have to do three or four or five studies to get there, that you can be a lot more deliberate in terms of the population that you're targeting.
That's part of the perspective. The other is, of course, we're studying this as an adjunctive treatment. Adjunctive treatment in people, therefore, who have failed a treatment also means a lower placebo response and people who are more clearly patient-like, that is, have a clear record of being diagnosed and being treated, as opposed to a lot of monotherapy studies where patients have absolutely no history of MDD or treatment.
Got it. Earlier last year, not this year, I think you did an interim analysis, and then I think you decided to continue the study and upsize it. What can we infer what the placebo-adjusted efficacy delta might have shown for you to upsize rather than fully stop for futility, for instance?
Yeah. The reason we did an interim analysis was really, again, to address that professional patient risk, which we saw in clear terms with the readout of the ALTO-100 MDD phase II-B, where the adjunctive portion of that study had a really nice effect Cohen's d of 0.47, and were 100% compliant with drug amongst those people we sampled for PK, whereas the monotherapy arm was 56% compliant amongst the people that we sampled for PK, and didn't show any effect. That led us to then say, "Well, can we identify those site-level execution risk factors amongst the baseline data for our ALTO-300 study to take out those sites that have the highest risk of bringing in professional patients?"
We did that in a blinded way, then ran the interim analysis to basically make sure that we have drug signal in there. As Andrew was saying, the resizing, which was a very slight resizing from 150- 200 patients with the biomarker. We set boundaries on either end for futility and for early success, which is a very high bar, as you can imagine, it didn't meet either of those boundaries. The effect size is somewhere in between, very consistent with a drug signal, but now amongst patients who are more confirmed in terms of removing that professional patient risk, revised our approach going forward. The upsizing was very modest, was just what was recommended by the statisticians to go from powering at 0.45 to powering at 0.4 Cohen's d.
Thanks. On the safety side, I think agomelatine at higher doses do show liver toxicity. Maybe talk about what you expect for your compound-
Yeah
...in the phase II.
At 50 mg, there is a slight rate of LFT elevations. They're actually not toxicity per se, in the sense that it doesn't lead to any long-term issues. These are actually adaptive liver enzyme changes at lower rates, by the way, than you see for other antidepressants and certainly for something like COBENFY, which doesn't even have LFTs in its label. At 25 mg, you get placebo-like levels of LFT elevations.
At 25 mg, you get the same level of efficacy as you get at 50 mg. In other words, we chose the 25 mg level specifically because meta-analyses have demonstrated there is no dose response on efficacy, but there is on LFT, so you might as well stick with 25 mg, get the efficacy at that dose, and avoid the LFT elevations as per, for example, the Novartis U.S. studies. That's where we are, and of course, we'll let the data dictate how things look, but that was the strategy for getting there.
Got it. What patents are you relying on this asset? Because I think agomelatine ran out in terms of exclusivity. Yeah.
We have granted a method of treatment in patient selection using the EEG biomarker, so it's very much directly matching what would be in the label. That's an approach we feel very strongly about. I also think it's important to take a moment to reflect on the importance of method of treatment in psychiatry. All of the successful programs, lumateperone CAPLYTA, Auvelity from Axsome, COBENFY from Karuna, esketamine from-
SPRAVATO.
...SPRAVATO from J&J, all are based on method of treatment patents. We've studied that landscape for all of our programs. We've reflected that in our IP position, which we believe is very strong, and, in this case, already granted IP around our patient definition.
Great. The last couple minutes is, you do have another program, ALTO-100, which we touched on. BDNF is the mechanism. Data mid-2027, so maybe speak to your confidence around why you think this could succeed in bipolar depression, and how do you power this study this time?
This is a novel mechanism of action. It's a potential first-in-class drug that enhances neuroplasticity directly. That's a drug that was developed really out of a functional assay looking for neuroplastic mechanisms. It's being given to patients here who have an identified deficit in neuroplasticity, as evident by deficits in learning and memory, which requires neuroplasticity as its basis. There's a good mechanistic match here in terms of what the drug does and how we define the patients.
Also in bipolar depression more broadly, which is where we're targeting, it's been repeatedly shown, just like in major depression, that people with cognitive problems like memory problems tend to have a much worse course of their depression, more chronic, more disabled. All they have in bipolar depression right now is adjunctive antipsychotics with all of their side effect burden.
The opportunity here is a really strong opportunity. The confidence that comes for us comes, for example, from the adjunctive arm of that MDD phase II-B study that I mentioned, which had an effect size of 0.47, and various other analyses we've done of the MDD data, which give us good reason to think that we're in a good position here to deliver a potentially first-in-class novel mechanism, which itself would be a super exciting outcome.
Great. Thank you. Three major readouts in 2027.
A lot going on.
big. Okay, well, thank you so much for the time, and congrats on all the progress, and thanks everyone for listening.
Thank you. My pleasure.