Good afternoon, everybody. My name is Athena Chin. As part of the biotech team here at TD Cowen. Welcome to our 46th annual healthcare conference. I'm joined here today with CEO of Alto Neuroscience, Amit Etkin, who will be providing a corporate presentation followed by a moderated Q&A, and then I'll open up the floor for any questions. Please take it away.
Great. Thank you, and thank you for inviting us. This conference actually brings back as telling Athena, it brings back memories. This was the first conference two years ago that we did after going public. I remember it mostly 'cause the AV cut out in the middle of the presentation. My first presentation as a public company CEO, I had to do with no slides while somebody was fiddling with the projector. Let's hope this one goes better. It's a pleasure here to talk with you. This is our forward-looking statement and disclosures. You know, we have a lot going on.
I'm only really gonna cover two programs, but the mission here is to really transform mental health care, and do so with the lens of understanding what do our drugs do for the brain, who are the right patients, taking a precision psychiatry approach, to really make things work.
We have multiple late-stage programs here, as you can see, focusing on depression, adjunctive depression, as well as treatment-resistant depression, bipolar depression, and cognition in schizophrenia, all areas of tremendous unmet need. All of the programs aim to fill areas where there's just a lot of very demonstrable advantages to have. I'm not gonna go through all of them, but I'm just gonna highlight two, and those are the ones that I'll go through in more detail.
ALTO-207 is a fixed-dose combination of a dopamine agonist called pramipexole and ondansetron, which is an antiemetic. That's a program that is in a phase II-B and soon, next year, we hope in a phase III. On the other side of the coin is ALTO-101, which is our next readout, which is our earliest stage program targeting cognition in schizophrenia, an area for which there are no treatments at all.
In between are two other readouts, this year, both phase II-B studies, ALTO-300, which is a drug called agomelatine, drug approved in Europe and Australia. We're using it as an adjunctive treatment for a biomarker-defined population, in depression. ALTO-100, which is a novel mechanism of action that enhances neuroplasticity targeted for bipolar depression.
Let me dive, though, into these two programs to give you much more of a sense of what we're focused on. Heuristically, everybody understands that dopamine and motivation is important in depression, but connecting on the ball really is where the difference lies here. We've done a lot of biomarker work around dopamine biomarkers with EEG in people and have found that patients with depression, but even more so treatment-resistant depression as clinically defined, enrich for a hypodopaminergic phenotype.
The question then becomes: What is the best way to address that? Both our work and work across the literature has suggested a direct dopamine agonist approach may have really unique efficacy, and one, in particular, targeted for the D3 subtype of dopamine receptor. That comes in the form of pramipexole.
Pramipexole is a drug approved for RLS and for Parkinson's, is 20 to one selective for D3 over D2, and as I'll show you, has very compelling efficacy in depression, but like every dopamine agonist, comes with known adverse events that are dose-limiting in pretty severe ways around nausea and vomiting.
That's where Chase Therapeutics developed a co-formulation of pramipexole and ondansetron, showed it was safe and showed efficacy. We acquired that asset. That's now ALTO-207. The data that really tipped the scale for us, and I think also for the field, is the study from the U.K. that was published after our deal closed called PAXD. PAXD was a 150 TRD patient study across nine sites in the U.K., all very, very resistant, well-documented patients.
They were randomized to pramipexole or placebo, slowly titrating pramipexole to try to deal with the nausea and vomiting, targeting 2.5 milligrams as the dose, showed a couple of very striking things. First is in its primary outcome, which was at 12 weeks, they showed a Cohen's d of almost 0.9, which would translate to over 7-point delta on the MADRS for efficacy.
That is massive, full stop, particularly massive considering this is a TRD population. They took the next step of keeping these patients randomized all the way out to almost a year, 48 weeks. You can see that even though at 12 weeks you could start to look for additional treatments as long as you stayed on your randomized treatment, pramipexole never lost that efficacy, placebo never caught up. Very striking outcome.
Similarly striking outcomes on suicidality, functioning, anxiety, and as you'll see in a second, on anhedonia. The main downside of pramipexole that I alluded to already around nausea and vomiting leading to discontinuation is what they also saw in this trial. 20% of patients in drug discontinuing, but 5% in placebo makes it an effective drug that's very hard to use in clinical practice because of its very hands-on titration, and a lot of patients just simply can't tolerate it. The other side of a dopamine approach is anhedonia. Again, thinking of the reward system and enhancing motivation.
There were large effects, not only in PAXD, but in another more experimental medicine study that came out on pramipexole, this past fall, showing large effects on anhedonia, the kind that are not seen anywhere else in the field and there are some development kind of comparisons there. Certainly, SSRIs have not touched that target. What Chase did, as I mentioned, was do a proof of concept trial here. 32 patients, one-to-one randomized for the combination versus placebo, and they found on the MADRS an over eight-point delta corresponding to a Cohen's d of 1.1 at week eight. Already opening up at week two.
Perhaps even more strikingly on the CGI, which is a general kind of global symptom measure that tends not to differentiate drug and placebo all that well, they still found the same Cohen's d of one effect and opening up at two weeks. Very striking efficacy. When you zoom out and put that in context, now using the eight-week time point for PAXD so we can compare across trials, it looks there in the gray and blue bars like there's really evidence of a dose-response relationship here, which people have argued in the literature before. You do need to get higher, and you do need to be able to titrate there faster. Chase here was able to achieve a five-fold faster titration more tolerably than you could with titration along the label.
All of that stands in stark contrast to the very modest efficacy of standard of care, be it esketamine or antipsychotic medications or even psychedelics, which, you know, we anticipate will start reaching the market in the coming years. This becomes a quite differentiated therapeutic opportunity. That trial, the phase II-B trial, will launch the first half of this year to read out next year. Prior to the phase II-B reading out, we anticipate starting a phase III by early next year. Together, they would form a potential pivotal submission package.
The background then, you know, from a commercial perspective, is not only do you have differentiated efficacy for a mechanism here that's not targeted, but you also have a background of kind of commercial and regulatory success for a 505(b)(2) combination in the form of Axsome's AUVELITY, which has done really, really well.
Karuna, in KarXT taking that concept of an antagonistic combination, reducing side effects to reveal efficacy, also providing a really nice validation and backdrop. That's ALTO-101. Sorry, ALTO-207. That's, as I mentioned, a 2027 readout. ALTO-101, which is our next readout, is in schizophrenia targeting cognitive impairment. Cognitive impairment is the driver of disability in the disease. It is without any treatment at all right now, not even off-label.
There's nothing approved for this. Schizophrenia has a prevalence of 0.5%-1% of the world population. This is a massive unmet need. We're addressing it here with a PDE4 inhibitor. PDE4 inhibitors you're familiar with for their anti-inflammatory effects in the immune system. It's also been studied heavily for neuroplasticity and cognition in animals, and we've done a lot of work now in humans.
What we've shown is that we can enhance brain circuit biomarker responses with EEG, something we're calling the theta response here, which is a biomarker we've validated repeatedly. Picks up the deficit that drives cognition in schizophrenia over multiple prospective samples. This was data from a phase I study in healthy volunteers giving them single doses of the drug, showing a dose-related improvement in that function, and in the same study, also a dose-related improvement in cognition.
The study that is reading out around the end of this quarter is an 83-patient crossover design, placebo-controlled trial. Everybody's their own control. We're looking at a couple of things here. We're looking at EEG as the primary outcome. Again, that idea of circuit engagement with a validated measure of circuit functioning as a go/no-go, a de-risking data point for taking the drug into a phase II-B trial, which would then be longer treatment, parallel group design.
We will also be looking at cognition in two domains in particular, processing speed and verbal memory, and we're selecting patients here based on their processing speed abnormality, having linked that to the biomarker. Our base expectation here is that we make a decision based on EEG. It's not out of the realm of possibility that we'll also see cognition improve.
That would be a tremendous upside with only 10 days of treatment. That's a setup for the trial. The form of the drug that is being studied here is also as a transdermal formulation. The reason for the transdermal is that it slows down absorption of the drug. We have shown that doing so dramatically reduces the nausea and vomiting side effects that you see with frankly, any PDE4 inhibitor, but certainly demonstrated with this drug. You now improve tolerability and allow yourself to test whether you're getting the right brain circuit and cognitive outcomes.
Just as a little bit of kind of hot off the press, preliminary data here on our baseline sample for this trial, you know, I mentioned that there is a patient selection approach here, just like we do in all of our trials based on a processing speed impairment of at least one standard deviation. The resulting sample has a over two standard deviation impairment, so clearly impaired subpopulation.
The people, by the way, that we screened out based on not having sufficient impairment, they have near normal levels of processing speed. If you included them, as would typically happen in most CIAS trials , they would have made up 30% of the enrolled sample. Obviously potential for washing out a lot of signal. We also replicated now for the fourth time the relationship between processing speed and our EEG biomarker.
By doing that and enriching based on processing speed, we also have enriched based, enriched the population for an EEG deficit and a verbal memory deficit. A really nice way and a scalable way to define these patients. Then on the kind of, execution end of things, which is critically important in psychiatry, and we've been very, public and kind of loud about the importance and the risk around professional patients.
We do a lot to head off that risk. Just kinda highlighting how that was done here, we required medical records, pharmacy records, evidence that these patients are in fact real patients in stable treatment. That resulted in when you test these patients' blood for their antipsychotic, 95% compliance with their medication regimen. That other 5% gets screened out.
Really understanding that we're getting the right patients and enriching for the right phenotype to be able to get that go, no-go signal. Where we stand right now is a 101 readout , first quarter. We have two other readouts, ALTO-300 and ALTO-100, this year, and then ALTO-207, phase II-B reading out next year, but a phase II-B initiating this year, and cash to cover all of that. We have cash into 2028 allows us to really be disciplined in, understanding what our catalysts are, how they drive value, and then how do we push the programs forward from there. Let me stop there. I would love to see your questions, Athena, and then those of the rest of the folks here.
Great. Thank you. Starting with ALTO-207, a program of key investor focus. The phase II-B study that's supposed to start in the first half of this year, how are you thinking about the design, the primary and secondary endpoints?
We'll be a lot more specific about the details, but the design is gonna be pretty plain vanilla for what you'd expect a phase II-B to be. It's one-to-one randomization, eight weeks of treatment, MADRS as the primary outcome. No surprises there. We'll be looking at anhedonia as one of multiple outcomes, but it will look like any registrational trial.
The definition of the patients is clinically based on TRD, and that's because we are achieving biomarker enrichment already with a clinical definition. We will also record EEG and cognitive task activity so that in the final analyses, we can see whether there's an even more enriched population that could be pre-specified analytically for later trials.
Because of this effect size, there's no sense in prospectively selecting based on the biomarker 'cause we're already enriching and we're anticipating getting the effect based on just that clinical definition. All that will be baked into the analytic plan.
Since the study will be adjunctive, are there any medications that you will have to exclude based on the dopaminergic mechanism of 207?
Yeah. Principally an antipsychotic, right?
Mm-hmm.
You don't wanna take a dopamine antagonist while you're taking a dopamine agonist. I'd be interested just to see what happens when you do that, because they have kinda different profiles. That's the main, you know, drug you're avoiding, basically.
Got it. When we talk to doctors, they have said that they've used pramipexole in their TRD patients who have failed TMS or SPRAVATO. Is that how you're thinking about targeting in the real world, or are you thinking of potentially pursuing monotherapy in the future too?
Yeah. The way we're thinking about this is you have a drug that is easy to take as a regular kind of pharmacy-delivered drug. As we think of like the SPRAVATOs and ultimately the psychedelics, right? It's a easy to prescribe and take drug that would be well-tolerated by virtue of the combination with, if you look at all these prior data, large effect sizes.
Within the TRD world, we're aiming to supplant everything that's out there, whether it's esketamine, antipsychotics, and certainly ahead of psychedelics as they eventually come to market. The opportunity is even bigger than that. It's an adjunctive TRD study. The next step is an adjunctive label writ large, which just includes the people who failed one instead of two or more treatments, and then a monotherapy expansion from there.
An effective drug that's well-tolerated and easy to prescribe, there's no reason that can't be very widely used. The reason that current practitioners put it after some of these other treatments, even that varies, some of them will use it before, is all around the difficulty dosing the drug, all around the fact that you have nausea and vomiting.
Some people can't take any. Some people run into symptoms. You have to back off and wait a month or whatever until re-challenging. It's very hands-on. I think that speaks, number one, for the efficacy that clinicians see in their practice with pramipexole. Number two, for the one main problem that needs to be solved, which is tolerability. So we think that this should be a very widely used drug if what we're seeing in these prior data hold true.
In the real world, what have you seen as the highest average dose of pramipexole that physicians have been able to titrate up to?
There's a lot of caveats there. Highest average. Let me express it in a slightly different way, which is we've looked at real world data. We've asked what is the maximum dose, and how does that maximum dose relate to efficacy and kind of threshold? We looked at the NIH's All of Us dataset. 100,000 patients there had depression. 2,000 had gotten pramipexole at some point in their lives.
Over 75% failed to even reach as the highest dose, 1 milligram, which is the lowest at where the literature in the past has shown efficacy. Almost nobody reached the 2.5 milligrams that PAXD showed. Nobody reached the 4.1 milligrams that CHASE showed. There's a wide distribution, but the bottom line is it's a very hard drug to get to the right target.
The average time to dose escalation in that dataset was 273 days. It's just not a practical drug despite that efficacy signal, except for the few that have the time to put in. With a titration that's driven by a starter pack and the combination and also a modified release formulation that we've developed that better PK matches these two drugs. Ondansetron has a shorter half-life, it's gonna go away when you have pramipexole sticking around and therefore not antagonized. All of that should improve tolerability and allow you to get within a very short amount of time to a higher dose.
With that new formulation that you just mentioned, is the phase II-B going to have one formulation or two?
Yes. Well, it's a co-formulation.
Cool.
They're both intermixed, and that modified release slows down both-.
Mm-hmm.
slows down ondansetron disproportionately more.
Are you able to disclose what the dose of pramipexole was in that, in that formulation?
We'll lay that all out soon. What we've said is it's gonna be in between where PAXD was and where the.
Mm-hmm.
4.1 milligrams were. That's really trying to optimize on efficacy, where to some degree more is better, but also limit things like somnolent side effects, which aren't a big driver of discontinuation, but you'd like to limit that, thinking again about a really wide population of prescribers and the importance of tolerability. A lot of that is a little bit more art than science, drawing on the experience of a lot of clinicians who've used pramipexole over the years.
Do you have any concerns about the really rare behavioral side effects seen with higher doses of pramipexole? Some of the KOLs we spoke to said that they saw addiction to gambling, et cetera, when they dosed up, but it went away once they down titrated.
Yeah. Just to give everyone background, these are what have been called impulsivity symptoms. They're not impulsivity that you'd worry about in depression like suicidality. That actually dramatically decreases with pramipexole. Their impulsivity is in more reward-seeking, like gambling and things of that nature. It's been seen in Parkinson's populations at substantial percentages, but not in depression populations. The thought has been that it's really the state of the dopamine system.
In Parkinson's, you have few dopamine neurons and as a consequence, your postsynaptic dopamine system compensates by becoming hypersensitized. In depression, it's the opposite. Postsynaptic dopamine is hyposensitized. Here and there you get cases of people doing things like more online gaming, as happened for two people in the PAXD study.
As Athena was alluding to, it's usually something that responds really well to a slight dose downward titration. Those two people were at two and a half milligrams in the PAXD study. When they went to one and a half milligrams, those symptoms went away. I think it's going to be just something that's part of physician education, probably be part of the label because this is a 505(b)(2) strategy, and that's part of the Parkinson's label.
You know, it's not something we're particularly worried about, we're conscious of. It's also a marker of actually improving people's dopaminergic reward system, right? That they're seeking other rewards. Clinically, it's a very different experience being on pramipexole than an antidepressant as we know it through SSRIs and SNRIs.
Just lastly, on the PAXD study, are there any learnings from that U.K. study that you're taking to your phase II-B design?
Yeah. I mean, there's a lot of factors and, you know, now there's also managing patient expectations and PI expectations after, you know, the PAXD result. It's understanding how to deal with nausea, side effects that don't cause discontinuation, dose targeting, and things of that nature.
What it really did was provide a very strong base also on the safety side. How do you address the safety? How do you quantify sort of sub-threshold reward impulsivity type symptoms? All of that we've incorporated. Michael Browning, who's been the PI at Oxford of that study, has been tremendously helpful in guiding that.
On IP, what is the portfolio for 207 and when do those expire?
We have granted IP, expiring the late 2030s. That's the co-formulation and then the method of treatment for depression. That was IP from Chase. We've additionally filed IP around the modified release formulation, titration, special populations like biomarker-defined populations. Really broadComplementary, a set of IP that would take us into the mid-2040s. You know, for context, you could look at the IP for frankly KarXT or AUVELITY. Very similar flavor for that IP, and hopefully ours is even broader.
Got it. Now shifting to ALTO-101, with data expected this quarter. What do you hope to see in your top line results, on theta band? Kinda just reminding us how that relates to cognition?
There's a couple of important things embedded in there, right? One is theta response is our primary outcome here because we've validated that number of times. That is the go/no-go, and all we're looking for is statistical significance. There's not a, like, a minimum clinically important difference that's been established.
Maybe over time as a potential surrogate marker there might be, but right now statistical significance is what we're looking for. The other kind of embedded question is: how does that de-risk the program? That's really critical, right? We would love to see a cognitive improvement with only 10 days. That would be phenomenal, right? We can't count on that.
What we will be looking for, and what we've seen in our phase I before, is a relationship between the change in EEG and the change in cognition or any even directional effects in cognition that would help tell you that you're not just moving a circuit, you're moving the function of that circuit even in that short amount of time.
We also look at other EEG measures that all can tell us how well we've moved that circuit. If you're seeing not just one measure but multiple measures move, statistically significantly, that also will de-risk that phase II-B. You know, the last thing you want is, frankly, like what happened with Boehringer and their GlyT1 inhibitor, iclepertin, where they showed no EEG effects in a schizophrenia population and ended up failing phase III.
It's that incremental de-risking and informing of the phase II-B design that we're going to be achieving with this proof of concept trial.
Are there any exploratory measures as part of your study that you will be top lining as well?
Nothing at top line. There'll be plenty of other things that we'll analyze, obviously, you know, with medical conference presentations and papers as as is fit.
Got it. We have a few minutes left, and would love to open up the floor for any questions. Well, then I can hog the questions. I know that you also have a few datasets coming later this year from ALTO-300. Maybe can you remind us what the study is and what the primary endpoint results that you're expecting are?
Yeah. ALTO-300, as I mentioned earlier, is a drug called agomelatine. It's a melatonergic agonist and a 5-HT2C antagonist. Different mechanism of action versus every other antidepressant, and it's being developed here as an adjunctive treatment for depression. Again, think about our comparison set being antipsychotic medications. Even at similar efficacy, frankly, not having that side effect burden makes a big difference.
Here we're selecting patients based on an EEG biomarker. We developed and then validated that EEG biomarker using a machine learning approach, so agnostic for what comes out of that analysis. What we were able to see is not only do we predict and replicate the ability to enrich for better response on MADRS with that biomarker, but it turns out that biomarker also relates to the drug's mechanism of action.
If you do essentially the opposite of the drug, you either stimulate 5-HT2C or you decrease dopamine, you develop, either in animals or in humans, a more biomarker positive phenotype on that EEG biomarker. What we're looking for here is a fairly standard efficacy signal, so placebo-controlled MADRS change in the biomarker positive patients.
We'll also have a small contingent of the study that's biomarker negative. That can give you an initial idea of stratification of the enrichment, but it's really too small to be definitive in that sense. It's mostly there to maintain patient expectations, so people don't say, "Well, I'm in a trial with a biomarker, therefore I must have the biomarker," and that inflates placebo response.
That's, you know, the motivation for that in particular. We're expecting to see that as a simple yes or no stat sig primary outcome, and that will decide next phases for that drug.
Got it. With the last few seconds left, I know the biomarker is EEG-based. Currently, how are you thinking about potentially how practical it would be in the offices of psychiatrists?
You know, the background up to this question is that psychiatrists don't do EEG as a routine test. We actually think that this could be done extremely easily, and we've done a lot of work internally. The, the signal comes from a single electrode. You need a few more just to kinda capture and normalize the data.
We think that that can be done by a patient on themselves with a low cost, low channel count system in their home, and we've advanced a lot of that work internally with no experience needed to either apply the EEG or read out the result, which is algorithmic as a report, and that way, you know, not create a hurdle for a clinician.
That is very innovative. With that, we are at time, really looking forward to this year with the phase II-B study start for ALTO-207 as well as the datasets coming shortly.
Yeah.
Thank you very much.
Likewise. We're very excited as well. Thank you.