We're going to get started with our next session. I'm Andrew Tsai, Senior Biotech Analyst at Jefferies, and it's my pleasure to have Amit Etkin, CEO of Alto Neuroscience, beside me. Welcome.
Thank you.
Maybe spend a couple of minutes talking about the Alto story, what you're trying to achieve, and then you do have a lot of programs. Maybe walk us through the milestones we can expect over the next 12 to 24 months. That'd be great.
Awesome. We are a psychiatric drug developer focusing on disorders like depression, treatment-resistant depression, bipolar depression, schizophrenia, but really primarily taking a biological or biomarker lens to things. Trying to understand who are the right people for any one of our drugs, what are the right subpopulations, how to kind of go beyond the typical clinical definition and get biological insight, as well as understanding what drugs do in order to really target how they should be developed, things that have not really been done from a precision medicine lens in our field before, but have had a lot of promise and validation in other areas of medicine. We have four different drugs in phase two programs at various stages right now, all of which are going to read out in the next two years.
A lot of exciting promise in areas like cognition and schizophrenia, which is where our next readout is, Alto 101, first quarter of next year, where there are no treatments. Areas where when you have cognitive impairment, a frequent portion of a frequent issue in schizophrenia that is the driver of disease for which there are no established treatments. Other assets like Alto 207, which is a recent program that is now in a phase 2B, or about to start a phase 2B, targeted at treatment-resistant depression patients for which there are not a lot of alternatives. There is just a handful of really high-burden, low-efficacy drugs, and this is one that holds a lot of promise. That is in 2027. In between, we have Alto 300 mid-next year and Alto 100 in the second half of next year, both of them phase 2B programs.
In the case of Alto 300, targeting an adjunctive treatment in depression using a biomarker defined out of an EEG signal that can be gotten readily and at scale and actually by patients in their own homes. Alto 100 is targeting in a phase 2B patients with bipolar depression defined by a cognitive biomarker, a poor memory test biomarker that identifies a certain subset. That's also a phase 2B in an area where all those patients have is antipsychotic medication. Really, in all cases, we're looking for impact. We're looking for impact in really high-need areas and looking for differentiated profiles in part by a biomarker approach that I think is different from what you see historically in the field.
Very clear. I wanted to spend some time on Alto 207, which you acquired back in June. You got this from Chase Pharmaceuticals for a very small upfront. Congratulations on the deal since then. How did you get it for so cheap? Maybe talk us through the history, the recent history of that.
Maybe it's not so cheap when you look at the Coruna deal with Lilly. That was $100,000 upfront. I'm not quite sure that we're competing. The motivation for that asset was several-fold. First of all, we have been doing work on biomarkers of dopamine, both increased and decreased dopamine function as measured by EEG for years and interested in this kind of mechanism for years. That had pointed us to treatment-resistant sectors of the depression population as particularly enriched for a hypodopaminergic phenotype. The strategy being then you want to increase dopamine signaling, in particular with the dopamine agonist. That is very different from the other approaches that have historically been tried, which have really been around increasing the release of dopamine.
That dopamine agonist approach scientifically came together with findings that we were aware of from the PAXD study, which is a study of 150 TRD patients out of Oxford here in a NINDSITE NHS-based study that showed an outsized clinical efficacy signal, a Cohen's d of almost 0.9 in TRD patients, so three times the average antidepressant effect size. Those two things, then the biology and the compelling clinical background to the dopamine agonist portion of this fixed-dose combination, really motivated that deal.
What Chase had done was show that the main downside to a dopamine agonist, which is well known across all dopamine agonists, is the nausea and vomiting you get as you have more of that mechanism on board that limits the ability to dose it at the individual level and also limits generally the speed with which you can dose, both of which are critical factors for depression. They had shown in a fixed-dose combination with ondansetron that you can really cut down that nausea and vomiting and dose five times faster than the label to a higher target and reach an efficacy signal that was quite impressive. They found an eight-point difference on the moderates in a phase two proof of concept study in depression, a randomized trial in depression.
All of those things came together with a long-term relationship we have had with the founders of Chase and really made this the right deal for us and for them now. Now, as people have seen, I think the ability on our end to accelerate the program into a phase 2B and soon into a phase 3 as well by early 2027.
Understood. Thank you. You met with the FDA back in September, October to discuss the program. What exactly did you need to discuss and what exactly did you get FDA buy-in around?
Yeah. We were looking to address a couple of things. One is for the phase 2B, is there anything else that we should be thinking about in the design to position it best to address the efficacy picture? There was really nothing that needed to be done. That's great. The phase 2B is now on track to launch first half of 2026. That's been going very well in preparing for that. We actually pulled in that timeline because of that. It's mainly at this point literally making the pills and putting them on stability for long enough. The other thing that we asked them is, as we prepare for the phase 3, what are the gates? What are the rate-limiting elements for a phase 3 program? They really came down to two things.
One is a TOX study to bridge to chronic TOX, which you need to do for the open label extension and the ultimate ICH exposure guidelines. The other is a dose-ranging phase one study on ondansetron. In other words, how little can you get away with in the fixed-dose combo and still be covered in terms of the nausea effects? Both of those things we could do and are doing right away. That opened up the possibility for a staggered start acceleration of the phase 3 program. Rather than some time in 2028, where it would happen after a phase 2B readout and an end of phase 2 meeting, now we can pull it in and guiding that to be by early 2027. We'll have an FDA interaction on that phase one data and TOX data later in 2026 to set that up.
That acceleration, I mean, that's obviously very compelling for a program that could reach the clinic a few years later, pulls in our NDA timeline by at least a year. We took that opportunity and really pushed forward with 207 now in this accelerated.
Fashion.
Yeah.
Before you start the phase two in first half 2026, what else do you need to complete or finish? I think you mentioned stability testing.
Yeah, just CMC at this point.
Okay. Everything's good to go from there. Can you talk a little bit about the phase 2B design? How will you power the study? How big of a study? How many arms?
Yeah. The phase 2B is going to be relatively straightforward, one-to-one randomization, one dose of the one target dose of the drug. The drug itself is titrated. The way we're titrating it is with a couple of things in mind. One is, of course, the target dose here. We'll say more about where that is, but it's higher than what has been used in prior studies. Also, the nature of the titration, aiming to get people to at least where we think the minimum effective dose is, get everybody there, and then push people to the top dose. That titration is really in a one-to-one randomization drug versus placebo with one target dose. It'll be an eight-week duration study. We'll say more about that, elements of the design around that eight-week treatment as we finalize and get ready to launch the trial.
It is going to look more or less like a plain vanilla design. It's really meant to be potentially registration supportive. So fairly straightforward in that regard. That's probably going to be a little bit different from the phase 3, where we anticipate having to probably do more than one dose. This is the kind of efficacy dose. There'll be another lower dose that we will likely look at in the phase 3 to try to define, again, where that bottom end is, because typically FDA does like to have a dose-ranging component. That wasn't necessary in the phase 2B.
Okay. Interesting. As phase 2 is going on, data maybe 2027. Phase 3, you'll start it early 2027, maybe "at risk" a little bit.
A little bit at risk. That's right.
Will it be two phase 3s, one phase 3? How are you thinking?
Just one phase 3. The aim here is to get the phase 2B to be registration supportive.
What would be the go/no-go threshold for this phase 2B readout in 2027? Is it two points? Is that enough? Placebo adjusted? Or do you want to see something higher than that?
Yeah. So we obviously want to see something higher always, right? Historically, where pramipexole has been has been very large effect sizes. Even if you look at the oldest randomized trials around the one to one and a half milligram amount, which is kind of where efficacy was being seen, that's still a Cohen's d of 0.5, whereas standard of care for these patients with either esketamine or antipsychotics is around 0.25-0.3. Where studies have been since at higher doses, including Chase's own study, was a Cohen's d of 1.1 at a 4.1 milligram average dose. That's quite a lot of efficacy there. That was an eight-point difference in the moderates on the Chase study. We would love to have that kind of an effect, needless to say.
Given the beneficial tolerability profile of the drug relative to an antipsychotic where you're getting chronic metabolic effects, movement disorders, tardives, and so forth, or esketamine, which is not well tolerated and quite burdensome to deliver, even a two-point delta in a TRD population would be great. Obviously, this is aiming for more than a two-point delta here. It's really a different strategy with an asset that has a lot of validation to the mechanism and is unique in a sense that all the other things like antipsychotics that increase dopamine levels, none of them directly drive dopamine signaling, which just from our general knowledge, right, it's a very important target for depression.
Very clear. Thank you. And phase 3 starting in Q1 2027, is it fair to assume data for that study could be 2028, or is it closer to 2029?
We haven't started the study, so it's hard to guide. I think 2028 is a reasonable assumption at this point.
Ultimately, this could come to the market 2029.
Yeah. We won't get into specific dates, but whether we're talking 2029 or 2030 or whatever, it's all pretty soon.
Accelerated and.
Pretty soon.
Okay. You mentioned earlier in your prepared remarks, you're pursuing this biomarker approach at Alto. What biomarker are you approaching here for this study?
Yeah.
Exactly.
The strategy here is a little bit different in that because of the large effect size, you do not have to a priori select people based on a biomarker. We know that the TRD population is already enriched in a hypodopaminergic state, as reflected by a biomarker. We also know that from FDA discussions we have had before for our Alto 100 program, that essentially the rubric, which is one that logically makes sense that the FDA holds, is if the drug works in the general population, it will be labeled in the general population. Even if you have a biomarker that is a perfect biomarker for response, if you do not have to use the biomarker, it will be labeled generally with an enrichment marker that identifies who the better responders are.
That's the strategy we anticipate taking here by pre-specifying biomarkers as part of the phase 2B and phase 3, essentially complementary analysis. Where that changes for other assets, like for example, where we think the case might be for Alto 100, is where you do have to do that biomarker in order to find efficacy. It's not going to work in the general population. There, it's more of a companion diagnostic that's required for the label. That's ultimately a letting the data speak case in the labeling. You still get a biomarker that identifies some enrichment, even as part of the clinical trial section, even if the drug shows a 1.0 type effect size response.
Do you think, I mean, before we shift gears to your other pipeline assets, do you think we can expect you to do more BD deals, hunt for new assets or more assets in 2026? Or do you think you're right-sized?
I mean, right now, I think we're pretty right-sized. We're focused on what's ahead of us. We are always thinking, what are the opportunities? We have an engine for insight that's differentiated, right, with the biology that we're doing. We have an in-house data science team that's constantly churning on new ideas and new approaches. We wouldn't want to miss an opportunity like what we took with 207.
Very good. And so speaking of, in the meantime, you do have the CIAS schizophrenia readout coming in early 2026. Maybe walk us through the study design of this compound for this compound and what is the go/no-go threshold for you to move forward?
Yeah. CIAS, cognitive impairment associated with schizophrenia, is the driver of disease. As I mentioned before, there are no treatments, full stop. The opportunity here for improving cognition and with it functioning of these patients is massive. The trick is, though, that as we know, everything has failed in the past for various reasons, one of those simply being that the development has not really been de-risked through a careful biological biomarker-based approach. Understanding what do you need to actually move in the brain to make a difference? What do you need to show in earlier stage studies to know that you're going in the right direction and then dosing the drug and finding the individuals who better respond to that drug based on that biology? That's the first piece that we did here.
We found we looked at all the biomarkers associated with schizophrenia in the past, put them head-to-head against each other, found that one called a theta response is the most discriminative between patients and controls and the best correlated with cognition, and then took that finding and prospectively replicated it twice to show that this is consistently the best biomarker. Kind of the concept being towards a surrogate marker here for cognition and schizophrenia showed that it's relevant to animals as well. We then found that in a large phase one study, 40-person crossover design study, so everybody's their own control getting drug and placebo, that Alto 101, which is a PDE4 inhibitor, thought for a long time to improve cognition, in fact, improves both that EEG biomarker and cognition itself in single doses in healthy people, and that the two relate to each other.
In other words, by moving EEG, you get this correlate improvement in cognition. That tells you you're engaging the right target that's meaningful for patients. You're getting already evidence of cognitive improvement. That sets up the phase two proof of concept trial. That proof of concept trial has as its target, as its primary outcome, that EEG measure, the theta response. If that moves significantly, the drug advances to phase 2B. If we also see improvements in cognition, those are secondary outcomes. Ten days of dosing, as we are doing in that trial in patients with schizophrenia, may not be enough. In the event that it is, that's an even better outcome. That's certainly an upside scenario that would have us even more excited to advance the drug into phase 2B.
The other thing that we did is we reformulated this drug from an oral to a transdermal formulation, slowing down its absorption, which essentially took care of the class-wide side effects that are associated with PDE4s, which is nausea, vomiting, and diarrhea, which makes a really, really big difference. You can get higher doses and obviously long-term tolerability. You put all those pieces together, and there is a lot of excitement about this readout, a lot of potential for this patient population, but a very careful set of considerations on what is good and what's great, of course, and how you advance the drug by an incremental biological de-risking and then put it into long-term phase 2B type trials, which will look at both cognitive improvement and functional improvement.
Very clear. Recently, you've pushed back the data readout very slightly from, let's call it second half 2025 to early 2026 because you validated, I believe, in your press release this biomarker. What exactly happened to cause that delay?
That was more of just a mechanics of trials and the fact that the holidays are here and the trial involves changing the patch daily in the clinic. People do not like to come in on Thanksgiving and Christmas and the like. That is sort of a minor change. The thing that you alluded to, which was our second prospective replication of that EEG biomarker in patients with schizophrenia, totally independent sample in that case, people who were admitted to the unit over years at the University of Texas and then stabilized and then assessed. Very real-world population. Again, theta showed the best discrimination between patients and controls and the best correlation with cognition.
It's hard to kind of understate except for people who've been in the field for a long time how rare it is to see replications and how rare it is to see it like every single time. That gives me a lot of confidence and excitement about this biomarker outcome.
Very good. We'll wait for that data readout coming up in early 2026. Following that, very shortly afterwards, you have the agomelatine asset with an MDD data reading out. Maybe talk about your excitement for this compound, what it is, what gives you the confidence that will succeed in MDD.
Yeah. So this is agomelatine Alto 300. It's an antidepressant with a unique mechanism. It stimulates melatonin receptors and blocks 5-HT2C receptors, the consequence of which is to increase dopamine levels. It's a drug that's approved in Europe and Australia as an antidepressant as a monotherapy, not approved in the U.S. So this is a U.S.-focused NCE strategy. Two things actually motivate us in particular with this drug. One is its really superior tolerability compared to everything else as an antidepressant, as compared to an antipsychotic, which, as we discussed before, carries all these chronic metabolic and movement disorder and other kind of tolerability side effects. We're developing that drug as an adjunctive treatment, that comparison set being antipsychotic treatment.
We also found in a large phase 2A study and then prospectively replicated that an EEG biomarker, which we established through machine learning, could predict outcome with that drug and that specific versus placebo. That allows us to select those patients. About half of those patients with depression have that biomarker. We then were able, even though the biomarker was identified empirically based on machine learning, to go back and understand the why of it and link that EEG biomarker in studies in animals and in humans to the drug's mechanism of action. In essence, that biomarker reflects a high 5-HT2C tone or a low dopaminergic tone, the opposite of which is what the drug achieves.
That comes together quite nicely, a drug that would have a real clinical need addressed and the ability to find the right patients with a mechanistically meaningful and yet machine learning-derived biomarker. That is a mid-2026 readout for a phase 2B study. If that is positive, that goes into a phase 3 program. Obviously, a lot of reasons to be excited about that. Frankly, excited about 2026 as a year of a lot of different readouts coming.
One thing you did in this study specifically was you took an interim analysis earlier. This was it earlier this year?
Earlier this year. It was a long year.
Right. Where you decided to, or the independent committee suggested to upsize the number of patients. What can we infer from that recommendation in terms of the efficacy delta that this drug could be seeing or showing?
Yeah. The motivation for that interim analysis, which we do not normally do, is really to remove the site-level execution risk that we had uncovered in the Alto 100 program about a year ago, which we did in a blinded case review, took out data from four sites, found in the interim analysis that there is a drug signal there. That supported continuing the study with a very slight upsizing, essentially powered at a Cohen's d of 0.4 instead of 0.45. Fairly similar overall final sample, 200 biomarker-positive patients. Perhaps more importantly, it was really driven by the insights from that Alto 100 trial, which pointed to certain types of sites commonly used in everybody's development programs, bringing in a lot of professional patients. We have completely overhauled our requirements for trials.
For all trials, we require medical records, pharmacy records, blood or urine measures of the underlying antidepressant, a sponsor eligibility review. We brought in AI tools to analyze interviews, all homegrown, to really bring a new level of execution rigor over what is going on at sites. Since then, I have heard that big pharma and biotechs have done the same following us being kind of public about the source of the outcome for 100 and our solution, which is actually quite heartening to me to see that people are taking that approach. Certainly, drugs should not be failing for those reasons. We have seen that improve the quality of patients coming through on all programs. I think at this point, honestly, we should be asking everybody, why are we not doing this across the field? It is critical to get that execution right.
Understood. Very clear. Final question is the data readout will be on 200 biomarker-positive.
Correct.
Okay. Very good. Thank you so much for giving us an update today.
My pleasure.
There's lots of things going on over the next one to two years, and hopefully, big success to come.
I appreciate that. Thank you.
All right. Thank you, everyone, for listening.