Great. It's my pleasure to be moderating this panel with Amit Etkin, Founder and CEO of Alto Neuroscience. We have a number of topics to go through, but I think as we were just discussing, Amit, we'll start with ALTO-207, which has become a greater and greater part of the story over the past six months. Maybe just rewind for people to give a quick background on, you know, the deal you did, last summer around this asset, clinical potential and TRD, and sort of what gets you excited, and then I'm sure there'll be many follow-ups. Thank you again for attending, and take it away.
Yeah. Thanks for having us on. Certainly don't need to give much background on the need around TRD and just that opportunity set. It's, you know, it's a massive unmet need. The question starts with what's the right biology to go after and what's the clinical evidence for that? We've done a lot of work over the years with dopamine biomarkers and interested in modulating dopamine in different ways for many obvious reasons around depression and dopamine. Our data has pointed to the more resistant end of the spectrum being even more enriched in the biology of the sort of hypodopaminergic phenotype. The question becomes how to best modulate dopamine.
Our data, as well as a lot of external data, point to direct agonism of the dopamine receptor, especially the D3 dopamine receptor, and that comes in the form of pramipexole. There's been a lot of great data I'll touch on in a second, the data most recently that came out after our deal from the PAX-D study. Direct agonism with a D3-preferring drug like pramipexole has led to really striking efficacy in a number of different trials. We've recently done a meta-analysis showing that on average across all doses, you're at 0.64 effect size, so really compelling clinical data.
The downside, though, of every dopamine agonist, and pramipexole included, is dose-related nausea and vomiting that limits both the dose you can reach and the speed that you can reach it, which makes it a very hard drug to use for the treatment of depression, where you're trying to get to a good dose, in fact, a higher dose, given the data in TRD as fast as possible. Chase Therapeutics developed a co-formulation of pramipexole and ondansetron that cuts down on the nausea and vomiting, allowing them to dose five times faster and to a higher target, leading to a really compelling clinical effect. That's the deal then we did in June is to bring that asset in.
They'd found in a phase IIa a Cohen's d of 1.1 over 8-point delta on the MADRS, in MDD, and that was just a super compelling package. We knew about the PAX-D results which were yet to come out and ultimately came out in July, which was an even more striking result for its size and rigor. It's a nine-site, 150-patient TRD study out of the U.K., randomized people for 48 weeks. Their primary outcome at 12 weeks showed an almost 0.9 Cohen's d effect size in depression, and that maintained the whole time. Placebo never caught up, even though at 12 weeks they could start adding new treatments, and pramipexole never lost its effect.
That's a really unique profile, really compelling, addresses anhedonia, which is something that's also been dramatically under-addressed in the field and does so in an equally compelling manner to depression symptoms as a whole.
Great. Do you wanna drill down a little bit more on just the context around this PAX-D study? Why was this study, like, conducted in the first place? Like, where did the thought behind it come from?
Yeah. The goal there was to try to encourage use of pramipexole by NHS. It was, in that way, a trial embedded within the NHS. These are patients whose depression is very well documented in NHS records. That's why they also did a long-term randomization and treatment course to really define the benefits and define, you know, adverse event profiles and so forth and why they used the primary outcome that they did, which was a self-report measure of depression called the QIDS, which translates into a MADRS. That delta translated into about seven points delta on the MADRS, but it's something that a clinician could do more easily because you just give the patient a self-report measure, and you understand how they're doing.
I see. The study went out to 48 weeks, I mean, which is kind of amazing, because you just never see that in psychiatry. Like, how in your mind when you look at this result and the extrapolability of it, like, how well run of a study was this? What was, like, the retention rate over time? Like, any caveats that we should interpret the data with?
I mean, there's always caveats with everything, right? Things in the U.K. are embedded in a national healthcare system. Patients are better characterized actually often than they are in the United States. As you know, there's a lot of those little differences, but I think overall the generalizability of the study is quite broad. It was very well run. You can actually see in their supplement how response broke down by site. This is another thing you never see in psychiatry trials where every single site separates. In part, the results are just so robust, right?
You almost hope and expect for that. I think that all speaks to the rigor. They retained people as well as could certainly be imagined in a 48-week course. You're gonna lose people towards the end anyway, but none of that really affects the statistics, which are mixed model, take all the time points into consideration. Extremely well done, run as a blinded study, just like any industry-sponsored study would be. You know, a lot to draw on there.
Yeah. Okay. Some doctors we've talked to will use pramipexole now off-label in depression or in TRD. You're developing the combination with ondansetron. Can you talk about the data we have for pramipexole alone versus the combo, and like how much ondansetron shifts the adverse event balance and widens the TI?
Yeah. Quite a bit is the short answer. We have data on the actual real-world use of pramipexole to start with, because it's out there, right? People use it for Parkinson's, RLS, so people can use it for depression. But what you see is that overwhelmingly most people don't even reach that 1 mg kind of minimal effective threshold, and very, very few people reach the 2.5 mg that PAX-D targeted, and nobody reaches the 4.1 mg that Chase was able to get to total. Then you ask how long it takes them to titrate there, and that titration is often very slow. You're trying to get past nausea and vomiting by going very, very slowly, then you run into symptoms, you have to back off and then re-challenge and so forth.
Even if you're able to eventually get up to two or maybe a little bit higher and not get all the efficacy that you could, your time to get there is very, very long. The data with Chase in their phase I where they did a head-to-head and tolerability between pramipexole and the combination is that 60% of their sample is able to get to 6 mg in only 12 days with no issue at all. Everybody saw a benefit in terms of having a higher effective dose, tolerable dose reached of pramipexole. In their phase IIa, they reached 4.1 mg in only eight days.
Those kinds of numbers and total dose and the speed to get there are just something you can't do with pramipexole as it is now, and that makes it a much more generalizable drug that could be used by any prescriber.
Yeah. Yeah. Okay. Makes sense. Do you wanna talk about the IP piece, right? Given that these are two drugs that are a little bit older and, you know, I guess, what would you say to someone who says, like, "Why is this not obvious to add something that mitigates side effects to something that has side effects?" Like, how have you been able to kinda put a thicket around this?
Yeah. First of all, we have granted IP on exactly that. The argument to it being, you know, obvious is, well, people actually haven't done it, and therefore, you know, that certainly demonstrates the non-obviousness. But there's also historical precedent around things like apomorphine, which is a different dopamine agonist, for which ondansetron is actually contraindicated. There was a need to even demonstrate the safety of this combination, which Chase did very well and opened up the ability to give both at the same time, and that supported very strong IP.
We have granted IP around the composition of the combination and around method of treatment in depression, and then filed IP around a modified release formulation that we've developed which better PK matches ondansetron with a short half-life and pramipexole with a longer half-life, and then a titration schedule that itself is optimized for tolerability, and then things like special populations like those identified with biomarkers. That's a pretty broad IP portfolio, but already the anchors are granted, and we've done a huge amount of IP diligence, needless to say, in bringing in that asset in the first place.
Yep. Okay. Makes sense. Do you wanna talk about the planned studies, design, powering, things like that?
Yeah. We're starting a phase IIb first half. That'll read out second half of 2027. The plan is to start phase III by early 2027, and we'll guide once we know what that looks like and so forth, in terms of its design and timing. For the phase IIb, it's a parallel group, 1:1 randomization, titrating up to 3.2 mg within about 13 days of pramipexole, 15 mg of ondansetron split, both are split BID, eight weeks of treatment duration, including the titration, and a sample size of 178.
The powering there is such that we have 80% power at a Cohen's d of 0.45, and 50% power, so our ability to detect statistical significance right around 0.3, meaning that all we have to be is really where standard of care is.
We're fine. With the effect sizes historically, we should be in really, really good shape. The plan is that phase IIb is part of the registration package. That'll obviously be up to FDA to determine in reviewing the data, but part of what we've been consulting with them on. The phase III, we would align with them for a start in early 2027 with some talks and phase I data that they wanted to see done in hand before that happens.
Okay. What was that specifically?
Two things, both fairly straightforward. One is just doing the 13-week tox on the combination, and that's to enable the open label part of the phase III, and frankly can add open label to the phase IIb at that point. The other is a dose ranging down on the ondansetron side. They wanted to see if there is a floor that we can define for the dose of ondansetron that's needed, and that's a phase I study just around tolerability, essentially the inverse of what Chase did in their phase I.
Right. Okay. Makes sense. You know, what have you learned about, you know, the nuances of executing these psych studies over the past couple of years, right? You did the interim on ALTO-300, you had the experience with ALTO-100. You guys don't use a CRO, you run it yourselves. You know, how should investors think about like the execution risk and the way you're tackling this problem?
We saw the risk, you know, we do a lot of like the stuff that you'd consider pretty standard, things like the SAFER interviews and whatnot. We saw a risk with the ALTO-100 MDD program pop up that I think has become more and more clear to everybody in the field, not just in depression, but in schizophrenia, and, you know, Alzheimer's just the same, which is the behavior of sites, is something that needs to sort of be taken in context. The potential for sites to, for various reasons, let in what we would call professional patients, has not been particularly well mitigated historically.
We in part, and, you know, I think very flexibly in part because we're able to do this through an in-house operations team, have instituted a much more rigorous approach, than I think has historically been done which is that we require medical records, pharmacy records. The patients have to look like patients historically. All of our trials are adjunctive, which means that you can measure the antidepressant or antipsychotic or mood stabilizer for whatever indication that the patient is on in their blood or urine. You should see evidence of refills in their pharmacy records before. None of that is what we could find in a monotherapy population, for example. We monitor their, for example, antidepressant use for their underlying antidepressant throughout the trial.
What we've seen is when we've required that now, everybody stays on their antidepressant throughout the trial. Like you have a much more real patient compliant population. And that's true not just of depression, that's true of bipolar disorder, that's true of schizophrenia. We've also implemented a sponsor eligibility review. We, with a number of psychiatrists in our leadership and medical team, review every single patient. We've brought in kinda homegrown AI tools to assess all of their structured interviews and to be able to process all of the information in their EDC and really kinda condense down what we understand about the patient, what the risks are around the patient, and what needs to be addressed by the site. That level of control, I think, is critical for neuropsych, and we've seen others follow suit.
Hopefully, that's a new standard. You know, we shouldn't be failing trials because sites bring in professional patients. If the science doesn't work, science doesn't work, but that, you know, we need to know where the source of risk is and address it.
You feel like the best way to tackle this is to run these independently and not rely on a CRO. Is that right?
Yeah. I mean, I just don't see at this point what a CRO adds. You know, the level of visibility and control we have by running it internally, I think, is stronger than you know, a CRO could bring anyway. They're not bringing magic, right? Ultimately, you care much more about the outcomes in your data than they do.
Right. Yep. Well said. Okay. Maybe just last questions that I get on pramipexole are just it's a dopamine receptor agonist. Maybe talk about any on-target side effects with that. Like how comfortable are we with impulse control? Are we totally comfortable on abuse liability? Do we expect this to be scheduled? Like maybe go through all that stuff.
Just starting at the end, it won't be scheduled because it isn't scheduled. There's no evidence for abuse liability. The drug's been out there for over 25 years used in Parkinson's and RLS. You know, we're dosing faster, but within the approved dose range, and so there's really no reason to expect anything, nor has FDA indicated anything on the abuse liability range, and certainly I would not at all expect it to be scheduled. In terms of the on-target side effects, there's two that we think about. Principally one is somnolence that I mentioned before. Usually, that can be dealt with a slight dose reduction.
You know, that's part of how we think of even our dose target for the phase IIb and beyond, to be able to address that side effect and still maintain a therapeutic dose. The impulse control side, just to be really clear what that means, it does not mean impulsivity like suicidality, which you would be concerned about in depression. In fact, in PAX-D, you saw a dramatic reduction in suicidality concomitant with the rest of the depression. It's basically when you stimulate the reward system, and that's part of why you're getting an anti-anhedonic effect, people do rewarding things more.
In Parkinson's, this is very different from depression, where you have a hypersensitized dopaminergic system because there's few dopamine neurons in Parkinson's, that rate might be around 15% of people doing things like gambling and shopping and stuff like that. PAX-D, for example, has a 48-week study, so it's a pretty, you know, clear bit of data, show that that rate was very low, much more mild, and very readily addressed with a dose reduction. There were two patients out of the 75 who were in the pramipexole arm that were playing online games more. They were both at 2.5 mg, and that went away when they reduced to 1.5 mg.
It's something that we'll be watching out for w ill certainly be part of the label in the way it already is for the pramipexole label because as a 505(b)(2), you're taking all the safety data that's out there anyway. It's an opportunity just to educate physicians. It's not something that concerns us, though we're watchful for it. You know, it just shows that you really are stimulating the reward system in a unique and powerful way, and that's how you treat depression
Makes sense. Okay, great. Well, let's maybe spend some time on ALTO-101 since that data's coming next, and we have seven minutes left here. Do you wanna just introduce the concept, Amit, as it relates to, like, the mechanistic rationale, why you chose CIAS and the design of this study?
Yeah. CIAS is the cognitive impairment in schizophrenia is the driver of the disease. It's the thing that starts earliest. It's persistent throughout life. It drives disability, inability to hold a job, relationships, support oneself, live independently. And it's substantial. You're talking about 1.5 standard deviation on average in this population, and you see that in many patients with schizophrenia. Importantly, there are no treatments at all. Everything that has been tried in the past has failed, and if you start to kind of track back to why, that starts to position how we've developed our program.
In some part, prior programs have not even assessed effects on the brain and kind of durable effects on the brain of the drug and just gone right for late-stage studies and often not even enrich the population in actually demonstrating cognitive impairment, which a lot of patients who might come to trials may not have a sufficient cognitive impairment 'cause they're able to negotiate trials, right? We tried to solve all of those things. Let me just start with a mechanism, and then I'll talk about the design and how it reflects that. This is a PDE4 inhibitor.
Part of the concept here is going after cyclic AMP is a very important second messenger system and not going after glutamate receptors, nicotinic receptors that historical programs have had where there's a limited ability to modulate them and a lot of compensation. The PDE4 system has been implicated in cognition for a long time. We know that one thing that cyclic AMP does is it regulates the activity of particular kind of interneurons, the inhibitory neurons in the cortex, that act like gates that drive a certain rhythm in the brain that helps brain regions communicate, and you can pick it up with EEG. As the drug improves that and improves plasticity to kind of embed that in a longer-term, stable neural structure, you should be able to see those EEG changes.
This theta response, this rhythmic response, is one of the areas of greatest deficit in these patients and is correlated with cognitive impairment. The study design is primarily on that as the primary outcome to see are you engaging the right target? Can you improve with 10 days of dosing? It is a crossover design here. Everybody gets drug and placebo. Can you improve that circuit marker that directly reflects the driver of cognitive impairment and the drug's mechanism and how we think it works at the molecular and cellular level? Over time, that translates into cognitive improvement is the thinking. We'll also be assessing cognition in that trial, specifically two forms of cognition, processing speed and memory, both of which we've linked to the drug or others likewise to similar mechanisms. They're well-chosen.
We just don't know whether 10 days of treatment is enough to move that. If it does, phenomenal outcome. That's something that's never been seen before, but the EEG should be enough to move the drug forward. That all said, there is precedent. We've moved cognition and EEG in a single-dose healthy volunteer well-powered study, and roflumilast has shown effects again on EEG and cognition with an eight-day crossover in CIAS. I think it's a good setup. 83 patients were randomized. We did patient selection based on the magnitude of the processing speed impairment. We have a two standard deviation impairment at baseline in processing speed. That enriched also for EEG and for memory. A good well-characterized population, well-powered study, and we'll get the data in a couple weeks.
Do we, on this EEG biomarker, do we understand what is actually, like, an effect on that that would correlate or is beyond the signal-to-noise ratio?
I think just relying on standard statistics, if it's statistically significant, indicates a clear effect of the drug. We have seen in the prior phase I that the degree of change within subject in EEG correlates with the degree of change in cognition. We'll look at that as well here. That gives you kind of additional color on what a change in EEG means, but ultimately, seeing a statistically significant change in EEG is the signal for us.
When we think about cognitive impairment in schizophrenia, like, how much of the actual dopamine blockade mechanism of the antipsychotics is a contributor to that, do you think?
If anything, existing drugs work against you rather than-
That's what I was gonna say. Yeah.
Yeah. Mostly for the anticholinergic effects. The newer drugs have very little anticholinergic effect and tend to be neutral for cognition. The older drugs can really make cognition worse.
Interesting. Okay. All right. We'll get that data in a couple weeks. Is that right?
Yep. Around end of the quarter.
Okay. We have the ALTO-300 data this year. Maybe just help set that data up quickly on just sort of, you know, what would be the hurdle for moving ALTO-300 forward, both as it relates to, like, the effect in kind of your primary population, the biomarker population, but also showing a difference between the biomarker positive and negative patients too.
Yeah. This is using EEG as the patient selection marker. It's powered really just for that primary population. We have a much smaller biomarker negative population, mostly to maintain expectations for patients so they don't just think I'm in a study because of a biomarker response and that inflates placebo. It gives us an initial view of enrichment but not a very definitive one. I would think of it really as moderate change, you know, at six weeks, just in the way we understand it for any depression trial. Statistical significance in that primary population is the signal to move forward. For a drug that's really, really well-tolerated, historically has always been well-tolerated, as an adjunctive treatment looking to displace antipsychotics, which even with equal effect size on the tolerability wins alone, and obviously, we're looking to enrich further beyond that.
Okay. And that date is when exactly?
That's mid this year.
Okay. Great. We're right up against time. Anything else you wanna highlight, Amit?
I think that's great. We haven't even talked about ALTO-100, which is yet another catalyst second half of this year. A lot going on to be very excited about.
How long is your updated cash runway now?
Through 2029.
Okay. Great. All right. Well, thank you very much for joining. Appreciate it.
Yep. My pleasure.