Good morning, everybody. Happy to be hosting this panel with Amit Etkin, Founder and CEO of Alto Neuroscience. I'm going to kick it over to Amit to maybe just make some brief opening remarks, set the stage on the key programs and where they're at going into 2027. You also just had your earnings update yesterday, and maybe you can just give us a little tidbit there, and we'll do Q&A. Thank you.
Great. Good morning. Thank you for having us here. Alto Neuroscience, just to bring everybody up to speed, is a late-stage psychiatry drug developer, taking a variety of approaches across disorders like major depression, treatment-resistant depression, bipolar depression, as well we've done work in other areas like schizophrenia to develop novel drugs, often guided by ways to select patients for better response through biomarkers, things like EEG and cognitive measures. Our lead asset is ALTO-207, which is a fixed-dose combination of pramipexole, which is a D3-preferring direct full dopamine agonist, and ondansetron. That combination allows faster and higher dosing of pramipexole in that fixed-dose combination. Pramipexole itself has a boatload of efficacy history behind it, but the issue being nausea and vomiting that's induced by any dopamine agonist, and that's where the ondansetron comes in.
It's a program based on a lot of years of research, and we'll talk about that in terms of clinical efficacy. We also have ALTO-300, which is a drug, agomelatine, approved in Europe, being developed here as an adjunctive treatment in depression. The comparison set being antipsychotic medications, and the drug itself is a very well-tolerated, different mechanism of action, and we use an EEG biomarker to additionally enrich the population. Then ALTO-100, which is a pro-neuroplasticity first-in-class compound for bipolar depression patients characterized by impairments in neuroplasticity through a test of learning and memory. We put out our earnings, which essentially said last night that everything is on track.
The guidances are first half next year for ALTO-300 phase II-B readout, mid-next year for ALTO-100 phase II-B readout, and then second half next year for ALTO-207 phase II-B readout, with $320 million in cash covering us through 2030, and execution not only on those programs, but the full phase III program for ALTO-207 in treatment-resistant depression patients. The expectation that we've been discussing is that all of the data that we've been collecting, phase I data that we'll talk about, blinded phase II data, goes to the FDA for a Q4 feedback on the question of essentially, can we get started with a phase III program, which we're guiding to a start in early 2027, assuming alignment with the FDA. That's really the event this quarter.
But gearing up for that start, everything is tracking well in terms of guidance. Then likewise, we've been anticipating a start for a second phase III for indication expansion in ALTO-207 second half next year so that we have, through that, the broadest depression label, both adjunctive and monotherapy use in treatment-resistant depression patients, which is very, very exciting.
That's great. On the FDA dialogue around forging ahead with a phase III, pramipexole and ondansetron are well-known drugs, so what question are you trying to get the FDA comfortable with to give you the green light?
The main question is actually just the timing of start. So what we're aiming to do here is start the phase III before the phase IIb reads out. The primary motivation for that is, of course, to get the drug through NDA and approval as soon as possible. The information that we have on both efficacy and safety, not just from what we've collected but what we know as a whole, is quite massive. So really, it's a benefit-risk calculation here around the timing. So the benefit side, the argument is based on 25 years of research where different groups have shown consistent outsized efficacy signals, most notably the PAX-D study that came out last year in Lancet Psychiatry, which showed an almost Cohen's d of 0.9 effect size in very resistant TRD patients, and that's three times the average successful antidepressant effect size.
And that's been consistent across every study published over the past 25 years. On the tolerability side, there's data that we have collected, part of these phase I, for example, and that the originator of this combination collected, Chase Therapeutics, showing that adding ondansetron improves tolerability of pramipexole and allows you to dose faster and higher in a well-tolerated way, which is critical for harnessing that efficacy. Then we know certainly historically in terms of exposure levels and chronic safety of these two drugs, and doing this within a 505(b)(2) path, that we can harness a lot of that safety information.
The argument here is a risk-benefit argument to FDA to start that phase III before the phase IIb reads out, because if you then take those two pieces and you have enough visibility on safety and you have historical information about tolerability, then really there's not a lot of advantage to waiting for the phase IIb. This is for the FDA to decide. We've put a good argument in front of them. We've also proposed a specific design following the current phase IIb, very, very similar. It's an eight-week total treatment duration, the same two-week titration schedule.
While we are studying a dose of 3.2 milligrams of pramipexole together with 15 milligrams of ondansetron in the phase IIb, that would now be the highest of two doses that we would power for registration in the phase III, the lower dose being half of that at 1.6 milligrams. You have a nice step in terms of understanding dose response on the pramipexole side. All of those things we'll get feedback on, and we'll give them as well blinded data in terms of tolerability for the phase IIb that's ongoing.
Yeah. Okay, great. We'll get that update relatively soon?
That'll be either late this quarter or early next year, just depending on their feedback. We're obviously at the whim of their timing.
Right.
The feedback we will provide is this critical question of, can we proceed with a phase III, currently guided for early 2027, plus information on essentially what they saw. The phase I data deal with PK, the goal of the PK being to show that we have a modified release formulation where we have nicely PK matched these two components, pramipexole and ondansetron, which have two different half-lives, and that PK matching them certainly could contribute well to the tolerability profile, because having pramipexole stick around longer while ondansetron falls off obviously can create side effects that are not being countered. We have a dose ranging on the ondansetron side. The question FDA wanted us to answer was, can you get away with less ondansetron? We think 15 milligrams is still well supported by the data.
Blinded safety data from the phase IIb, where the critical question is really how many people get to that top dose, how well can they titrate, and then how many people drop out due to adverse events.
Yep. Okay, great. Well, best of luck with that. Maybe to take a step back, you and I have talked about the PAX-D data a bunch, but for those who are listening here or eventually read the transcript here and are kind of newer to the story, I guess one way I would sort of ask the question about PAX-D is, for investors that have looked at psych and depression for a while, we have seen a number of investigator studies that were ultimately not replicated in later-stage trials. What about this PAX-D study do you think kind of makes it uniquely de-risking for pramipexole?
When you think of a traditional investigator-initiated study, you are thinking small n, single site, an effect size, but the effect size is sort of on the maybe larger side than what proves out, but not massive. On all of those fronts, this study is differentiated. It is 150 TRD patients, very well-characterized patients through NHS records, very treatment-resistant, nine sites across the U.K., rigorously assessed, not only for an acute outcome, which was at 12 weeks, their primary outcome, but then kept randomized for 48 weeks and showing persistent benefit the whole way. The magnitude of that benefit, if you look at the primary outcome effect size, a Cohen's d of almost 0.9 is three times a successful antidepressant. On any of those fronts, this really is a standout result. You could ask, is that result an outlier to the rest of the pramipexole literature?
It is pretty clear it is not. Going back to 2000, there has been multiple different trials, both investigator-initiated and an initial industry-sponsored study by the originators of pramipexole, all of which have shown effect sizes larger than typical antidepressants or antipsychotics, really does not matter, large effect sizes. Chase Therapeutics in their phase IIa showed a Cohen's d of 1.1 at a higher dose yet. But the important thing is not only is there consistent efficacy, there is a consistent dose-response relationship, so you do need to give more to get more. The interesting thing is no single study underpins that relationship. You could actually fit that line without PAX-D, and you predict what PAX-D shows, it is almost on the dot for what it showed. It really fits in well with a very strong body of research.
That is great. Do you want to talk about the phase IIb design for ALTO-207 and how your powering assumptions and what is needed for a P-value compares to the historical effect size?
Yeah. Being mindful of the history of either smaller trials, fewer site trials, leading to potential smaller effect size, now granted, starting from 0.9 gives you a lot of room to go, we took a conservative approach in terms of the phase IIb. The phase IIb is a one-to-one randomization between what will be the high dose, 3.2 milligrams pramipexole, 15 of ondansetron, and placebo. Patients are titrated within two weeks with this custom modified-release formulation and then treated out to eight weeks at that dose that they achieve at the end of titration. It is 178 people. Think about the size of the PAX-D study, 150. It is powered at 80% power for a Cohen's d of 0.45. At 0.45, I think it is a blockbuster result.
It puts you really above standard of care with a drug that has a unique clinical profile in terms of not just antidepressant efficacy, but anti-anhedonic effect, given the direct dopamine agonism this drug impacts D3-preferring pramipexole, and other studies have been shown to have a really prominent anti-anhedonic effect, so I think that would be a great outcome. But that is actually below where we have shown in a meta-analysis. At a lower dose, you get efficacy. We showed at two milligrams in that meta-analysis, you get a 0.64 effect size. So, we are very comfortable with the level of powering to achieve a really attractive clinical profile, and then let the historical results be an upside to that.
Yep. Okay. As it relates to the dose range that you are pursuing and the combination and all that stuff, how comfortable at this point are you that you really have that right?
I think we understand the dose-response relationship, right? It's a question of how to map against that dose-response relationship. We picked 3.2 as a way to try to balance between efficacy and potential side effects. Of course, you learn as you go. We don't know what that max efficacy is. Could you go to five or 10 milligrams and continue to get more efficacy? That's just not known. But the other important aspect to understanding dose is not just the dose that we picked at 3.2, but what a lower dose would be that would still allow us to register a second dose and achieve efficacy based on information at least that we have, and therefore really define a cost-benefit relationship overall.
Where we settled was 1.6 milligrams, which is what we're proposing to the FDA, knowing that prior studies have started to show efficacy around 1 to 1.5 milligrams. So this feels a little bit above that as a good sort of anchor point for a low dose, and then we'll see how things play out. But it feels like we have a pretty good dose range represented.
You've learned a lot around the art of enrolling these studies, finding the right patients, managing compliance, all the kind of little things that go into running a successful psych study from ALTO-100 and the 300 interim analysis. As it relates to 207 at this point, can you talk about what you're observing so far as it relates to monitoring of trial conduct, types of patients that are getting in the study, and your comfort that this is a well-run trial at this point?
Yeah. On a number of different fronts, I think we're confident in that. You alluded to ALTO-100, which was a phase II-B in depression where the monotherapy arm, there was an adjunctive and monotherapy stratified populations within the study. We saw much lower compliance, much lower than expected in the monotherapy arm than in the adjunctive arm, and it really came down in that non-compliance clustered by site. It really says that the risk here is site-level execution risk, and we've seen that over and over in other studies. I was just with some big pharma folks a week or two ago, and they were complaining about that same thing. It is a systemic issue in our field.
What we have taken is a very hard-line approach, which is, first of all of our studies are adjunctive, so that means if you're coming in, I need evidence that you've been prescribed the medication, medical and pharmacy records, you're taking it, that you have refills, that I can test your urine or blood for that drug and see that it's there before you get enrolled. Doing that, we've reported for ALTO-100 and ALTO-300, multiple different occasions leads to at or close to 100% compliance. If you find people with the right patient-like behavior, they continue to take your drug and do what you expect to do. They're not in a trial for compensation. We use a variety of other things. We've brought some AI tools in that we've developed in-house.
We use sort of a verification of clinical trial enrollment in other trials, and we continue to do that in all of our trials for any indication. That's very much the case here. The implication of doing that, not only do you get higher quality patients, but enrollment is also slower because now you create a little bit more friction that you really have to prove that that's the right patient. We've already accounted for that enrollment rate in our projections, and therefore are confident in our guidance here with ALTO-207 second half next year. ALTO-100 and ALTO-300 we've already talked about. But it has taken a lot of retooling of the approach and understanding where the challenge is to running trials before you even test the science.
I think we as a field ought to take that more seriously, and we certainly within Alto have taken it very deadly seriously.
Yep. Okay. Maybe talk about the IP perspective and the underlying concept of the IP and why it is non-obvious to combine these two drugs when ondansetron is a well-known anti-nausea medication.
Yeah. First of all, our IP position is great. We have IP filed or granted out to the mid-2040s. Granted IP include, for example, the use of the combination for depression. The question is rightfully like, well, why is that? It's granted, but why is that non-obvious, right? Why was it granted? And the simplest answer there is, if you look at the apomorphine label, which is a different dopamine agonist, which preceded the work done on pramipexole, it contraindicates five-HT3 antagonists because the combination was shown to be not safe, led to profound loss of blood pressure, hypotension. Chase Therapeutics demonstrated that's not the case for pramipexole. And so when-
Why is it different with apomorphine?
Apomorphine's just a very dirty drug. Pramipexole is much, much more specific for dopamine and D3 more than D2. When you're told that something is unsafe and a contraindication and you demonstrate otherwise, there's definitely a non-obviousness to putting those together. It's also not the same mechanism, right? So there's nothing that says that you can necessarily block a dopamine, selective dopamine-driven nausea signal in the way you can block a serotonin five-HT3 signal in the gut, for example, with surgery or radiation, which is where ondansetron is indicated. So all of those things support a very strong IP package, and we continue to file IP to take us further and further out. We have modified release formulation IP, that PK matching. We have a custom titration schedule, which will be the starter pack that patients get commercially, and that also contributes to the tolerability profile, special populations.
We've actually done a landscaping against not just other psychiatry combination products like AUVELITY, which has already passed its generic entrant challenges, COBENFY, CAPLYTA, they're all method of treatment patents, but also the whole field around combinations. We've mapped the types of IP, the strength of our IP against other IP, and it comes out really quite favorably. It's broad, and it's really high-quality IP, very exciting to obviously protect the product for years to come.
What work can investors do, or what work have you done, or what analogs can you point to that makes you comfortable that a payer might not try to push someone to say, "Hey, look, these two products are cheap individually. Yeah, I see there's this marketed brand, but step through the generics," or something like that?
It's highly improbable in our view. You're talking about two drugs, pramipexole and ondansetron, neither of which are indicated for depression. You're asking people to get to a dose that causes most people to drop out due to nausea and vomiting before you then give them a tolerated version of the drug. That just doesn't really make any clinical sense. Ondansetron's actually also only indicated for three days of use, and you'd have to take it multiple times per day at different schedules than pramipexole because of the half-life issue. That is not just a nightmare waiting to happen, that is liability for doctors and for insurers, compared to something that would be well-characterized, safety known, on label, and well-tolerated.
But on top of that, if you just look at what has happened with AUVELITY, which is certainly a good analog as a combination product in depression, the insurance coverage is actually quite good. In fact, they guided to $4 billion in peak sales on depression for AUVELITY, as just that indication alone. And the reality is, it's not just payers who don't necessarily, as we're thinking about pramipexole and ondansetron here, won't necessarily have you step through, but clinicians, when they have the options of DIYing a combination, don't do it because it is not to anybody's benefit to do that. And that's especially the case here. If you try to DIY it and your patient ends up vomiting and in the hospital because they messed up the schedule, I mean, I don't know anybody wants to invite that kind of risk.
Yep. Okay. Phase II-B data next year, regulatory update on starting the phase III late this year, early next year. In the meantime, we are going to get data from 300 sooner. Maybe give a little bit of background around that program, and I think the one sub-question, excuse me, when investors ask me about it is just, the biomarker strategy there is super interesting. Is the thesis still that the biomarker is going to enrich this to generate a larger-than-average effect size, if this works?
Yeah. Let's take a step back, right? Agomelatine is a drug that is approved in Europe as a monotherapy antidepressant. We are developing it as an adjunctive treatment. Just stopping there, as an adjunctive treatment, it has very good clinical value. It is an extremely well-tolerated drug, different mechanism of action. As a clinician, knowing the side effects an antipsychotic has in terms of metabolic effects, weight gain, tardive dyskinesia that could last years after you stop taking the drug, even at the same effect size, I would rather prescribe agomelatine over VRAYLAR or REXULTI that have all done well, CAPLYTA, and so forth. Just as an adjunctive treatment, I think that there is value.
The question is on the biomarker, can you further enrich the response so it is not just the all-comer effect, which is already good, in terms of a reason to continue and bring the drug to market, but can you then further enrich? That is where the EEG biomarker comes in. We found that using machine learning, we replicated in an independent prospective sample of people taking the drug in our study, and then actually linked that signal back to drug mechanism of action. That is what we have seen both clinically and mechanistically as a way to potentially enrich that sample. We will see how the trial turns out. Scaling that EEG marker actually should be fairly straightforward. That marker is calculated from a signal from a single electrode, so we have done a lot of work in the scalability of that.
That is probably going to be something where a clinician, if they see all-comer efficacy and a biomarker, can then decide, "Do I want to do the biomarker in order to increase my confidence in prescribing that drug, but still be supported by the idea that all of these patients, that an all-comer population has responded, and then there is clinical value in this drug in an adjunctive context?" So we would consider that to be a complementary or enrichment biomarker. Biomarkers are new in psychiatry, but at least this provides another tool, and one that we think patients would find themselves compelling in a world where, again, speaking as a psychiatrist, we just kind of randomly select medications and tell stories about them. Having a test that identifies a person and ties to mechanism, I think, becomes a pretty compelling part of the clinical narrative.
What magnitude of difference would you want to see from the biomarker-enriched population versus the all-comers population to feel comfortable that that's providing this sort of extra benefit via patient selection?
It's a hard kind of number to hard threshold to really come to because different people have different perspectives. I think if you look at what you would consider to be kind of a weaker antidepressant versus a stronger antidepressant, even an effect size of a 0.1 Cohen's d effect size difference between an enriched and a non-enriched population would be pretty compelling. Anything more than that becomes more and more compelling. It's really down to what the cost is of running a biomarker, which we think will be quite cheap, doing this EEG at scale in the clinic, in the home, one-time test, kind of immediate result with data analyzed on the spot. That burden is pretty low. In fact, you could do it, you don't even need reimbursement for that.
You can just do it as a test to help guide the clinical decision and identify a patient as a potential candidate. As you get more and more enrichment, of course, it becomes more and more compelling. I think that the patients themselves want to know why they're taking what they're taking. How is it helpful to me? I think even a small difference could be quite meaningful clinically.
Okay, great. Last couple minutes here. Do you want to just briefly cover ALTO-100 and what we will learn from the next readout in bipolar?
ALTO-100, understand people might not give it the highest odds of success given prior ALTO-100 MDD results, but those were really colored by that non-compliance professional patient issue. If you actually look at the adjunctive pre-specified population there in that MDD study, it showed a Cohen's d of 0.47, nearly significant. It was not actually powered for significance. That gives us confidence in continuing this approach in bipolar depression, where all they have is antipsychotic medication, and a drug that, in this case, doesn't have those long-term effects, is well tolerated, a completely novel mechanism of action anywhere, that we've understood now what the molecular target is. It's a G protein-coupled receptor that regulates neuroplasticity, becomes really exciting and compelling, and identifying people through a test of learning and memory, which is an online self-administered test that people do.
In fact, you did at our investor day a while back.
I did okay.
Yeah, you did all right. You did all right.
I wasn't your worst analyst.
No, but I think you're right there in the middle of the bell curve.
Sure. That's fine. That's my life.
It's room to improve. But taking that test becomes a really exciting clinical opportunity. We'll see. People can value it however they want. We're funded to execute on it. Wellcome Trust has been supporting this study because of the importance of bipolar depression, which is so under-researched and undertreated. If that's positive, it'd be an incredibly exciting new vantage point on mechanism, on indication, and bringing a drug at that point to phase III.
Great. Awesome. Well, perfect timing. Thank you, Amit. Best of luck with what's to come. Appreciate it.
Thank you. My pleasure.