We're going to go ahead and get started with our next session. I'm Andrew Tsai, Senior Biotech Analyst at Jefferies, and it's my pleasure to have Amit Etkin with me today, CEO of Alto Neuroscience. Welcome, Amit.
Great. Thank you for having us.
To begin, could you give us an introduction about Alto, your thesis, and then talk about your pipeline and then the milestones? That could be very helpful for our discussion.
Yes. We are a late-stage psychiatry drug developer in phase II-B, and hopefully soon phase III, for three different assets in high-need populations, including treatment-resistant depression, adjunctive treatment in depression, and bipolar depression. We've also done work on schizophrenia and other areas. The organizing principle of the company is really around trying to understand the biology of humans with these conditions, because we worked so abstractly in psychiatry in the past, so that we can understand with biomarkers, how do you dose a drug? What does it do to the brain? We can understand who are the right people to do that to, what are the ways you select patients that may respond to your drug, and use precision, in this case, in different ways for different programs as needed, but in all cases, to really maximize the clinical impact of these drugs.
As I mentioned, we're in phase II-B for three assets, ALTO-207 in treatment-resistant depression as an adjunctive treatment, ALTO-300 in adjunctive treatment of major depression, and ALTO-100 in adjunctive treatment of bipolar depression. These programs all reflect different amounts of knowledge about each of the drugs and different amounts of evidence of clinical efficacy, with ALTO-207 really taking the lead there. ALTO-207 is a fixed-dose combination of pramipexole, which is a dopamine agonist, and ondansetron, which mitigates the nausea and vomiting that you have with every dopamine agonist that precludes you from dosing high enough and fast enough for it to be a commercial large-scale treatment in depression. Pramipexole has a wealth of data behind it showing outsized efficacy signals, most recently the PAX-D study, which was a 150-patient TRD study showing an almost 0.9 Cohen's d effect in these highly resistant patients.
There we have a phase II-B ongoing that's on track to read out second half next year, and aiming to start the first phase III early 2027, pending alignment with FDA, which is a meeting that's upcoming, or feedback that's upcoming. Then a second phase III for an indication expansion to add monotherapy TRD second half of next year. ALTO-300, its phase II-B is reading out on track for first half next year, and ALTO-100 on track for mid-next year. So a lot going on in this next year. Then the cash position right now that covers us through 2030 and execution of the entire phase III funding and NDA submission for ALTO-207.
Cool. Perfect. Thank you. I wanted to spend a lot of time, if possible, on 207 today, which, like you said, has potentially pivotal phase II data in second half 2027. To start, this is for TRD. When we think TRD, we think SPRAVATO. Can you talk about just the underlying value proposition differentiation over SPRAVATO? Then there are also possibly psychedelics coming to the market, and maybe talk about that differentiation too.
Yeah. This is actually a really important point, because TRD is actually much more than just the population that you'd think of with SPRAVATO or psychedelics. TRD is defined as two or more treatment failures, right? Two or more treatment failures, another way to say that is anybody who could get a branded medication, right? Your primary care patient who's getting AUVELITY, your general psychiatry patient who's getting CAPLYTA or VRAYLAR or REXULTI. Those are all functionally TRD patients. That is the audience we're going after here, and that's how we've designed our development program. At the end of that chain, four, five, six, whatever failures in real life, that's the esketamine folks, and even afterwards is the psychedelics. We actually see the landscape as much, much broader, still under the TRD lens.
Keep in mind that a drug like AUVELITY, for some insurers, you need to fail three drugs before you get there, and yet they failed their TRD study, and that's a $4 billion peak sales guidance for depression. The opportunity is huge here with a drug that has outsized efficacy at a lower dose than what we're using. In a meta-analysis, we found that there's a 0.64 Cohen's d, so twice the average effect of a successful antidepressant when you look across the whole field with, as I mentioned, PAX-D at 0.9 and the prior phase IIa program from Chase Therapeutics, from whom we acquired this asset, at 1.1, an eight-point delta on the MADRS. With that kind of profile, a unique anti-anhedonic effect by virtue of this dopamine agonism, which other programs don't have, we're looking to displace everything.
Displace AUVELITY, displace CAPLYTA, displace things like esketamine and psychedelics, which require interventional context, REMS programs, and so forth. This will be just a standard mail order pharmacy, no REMS program type pill that we all understand how to use and prescribe and contract with insurers for and so forth.
Taken at home too instead of a psychiatrist.
Yeah, totally at home.
Okay. And so you mentioned throughout the body of literature, I think the average Cohen's d for pramipexole has been about 0.64. How did you power this phase IIb then for your own second half study?
Yeah, so that's a great question. So that average effect size in the meta-analysis at a lower dose, at two milligrams, we're using 3.2 milligrams, is a fantastic effect size. That would be just a phenomenal outcome for a drug, right? But we've taken a perspective here that we actually want to power this conservatively to that, for a range of obvious reasons, right? You want to make sure that you're not powering it to perfection, but also really just powering it to what you think a good drug, a potential blockbuster, exciting effect drug would look like. And for us, that was a Cohen's d of 0.45, so that translates to maybe four, maybe even 3.5 point delta on the MADRS.
At that level, you're above standard of care, and with a profile that doesn't include tardive dyskinesia, for example, and weight gain metabolic effects like an antipsychotic, and an anti-anhedonic effect, I think that's a really exciting profile for a drug. Then let the upside that has been seen repeatedly with pramipexole in the past be that upside, but we've powered the study at the point where we think it makes a very compelling commercial asset.
Ultimately, you've got an efficacious drug. I mean, when I talk to doctors, they're saying, "I'm using it off-label.
They love it, yeah.
You're really solving for the safety part.
That's right.
Can you talk about why you're confident the dose you've chosen for ondansetron, that antiemetic, to offset the safety side effects of pramipexole, why you think you have the right dose to mitigate the AEs?
Nausea, vomiting?
When you think about the various things that we're doing to mitigate primarily nausea and vomiting, which is the dose-limiting AEs, they come in essentially a couple categories. One is just simply the use of ondansetron, but on top of that, also a modified release formulation that better PK matches pramipexole, which has a longer half-life, and ondansetron, and then a titration schedule that also considers, in a non-linear way, how to get as many people as possible to the minimum effective dose, which historically has been around the 1 to 1.5 milligram threshold, and then ramp them up to that top dose. We're trying to achieve tolerability across that whole package. The dose component for ondansetron is essentially addressed through phase I work, and we've done that.
Chase Therapeutics before did some of that, but we've done that more explicitly by looking at the phase IIb dose of ondansetron and the lower dose, so 7.5 milligrams BID, 15 milligrams total, versus 2.5 milligrams BID, so five milligrams total, versus no ondansetron in our phase I program. And of course, there is just the observation around tolerability for the phase IIb study as it's going on, which is actually where you care most about it, because that's the patient data. All of that information has then gone to FDA as part of a request that we filed to start the phase III before the phase IIb reads out.
Our goal has been to start that phase III in early 2027, whereas the phase IIb reads out in the second half of 2027, and basing that argument on the known efficacy over multiple independent trials of pramipexole and the observed safety and tolerability across phase I and blinded phase IIb data thus far.
Can you give us a frame of reference of how high of a discontinuation rate due to these AEs pramipexole alone shows? I think you just mentioned, Chase had the combination that they did in phase IIa. Did you see lower rates of AEs and discontinuation rates with the combination?
Yeah. When you look at, I think the best analog in terms of what happens with pramipexole alone is a publication in 2000 from Corrigan et al. They titrated over two weeks, just like us, dosing BID, just like us, with just pramipexole, and their high-dose arm was five milligrams, so similar to where we're targeting, and they saw 58% of people drop out due to AEs. That's obviously not a tractable profile, right? I mean, it's so many, in fact, that they couldn't even run statistics in terms of outcome on their sample because they had such high rates of dropout. Through the development of the program by Chase and then by us, there's been an iterative understanding of how to titrate, certainly the importance of a modified release formulation, and the selection of the ondansetron dose.
Where we have consistently discussed would be a great discontinuation rate since the beginning is really where CAPLYTA is. CAPLYTA is widely considered to be a well-tolerated drug, obviously very successful outcome, and that was at 9% discontinuation due to AEs in the drug arm.
Being near there would be great, and that would be a fantastic clinical profile that we think would support a lot of use in the clinic together with the effect size we expect to see with the program. Obviously everything is ongoing, but all of that information from an FDA blinded safety perspective is there in what they're reviewing and will give us a sense. We have mentioned in our earnings release last week that we'll give the market an update on what that blinded tolerability looks like and really help people understand however it is that the FDA comes out and provides us feedback, on what basis did they make that?
I see. Was this the FDA’s request, to be clear, not some proactive move by you in order to start the phase III, just to be clear?
No, that is a proactive move by us.
It was a proactive move.
Because the default is you wait until the phase IIb reads out, but we are asking here to start the phase III before the phase IIb. If you do not talk to them about it, you are just taking on more risk, because they could give you feedback on a design element that could be important, or they could ask you to wait. That is not our base case, certainly, but it is much better to ask than to assume.
Right. To be crystal clear, the modified release that you are working on for this drug combination is for the ondansetron component, or?
It is actually for both. They are together modified, and the modification achieves a couple of really important goals. One is that PK matching.
The shorter half-life ondansetron now matches the longer half-life pramipexole. But we also delay the Tmax, the rate of absorption of both drugs. Certainly, for things that induce side effects, the speed with which they get to Cmax is a factor. Also, ondansetron itself, in order to be modified, you bring down the Cmax, right? You squash the Cmax to essentially lengthen it out, and that achieves another side goal, which is further reduce any risk of QT prolongation, which has been observed at higher doses, primarily IV. Speed to Cmax and higher Cmax. With ondansetron, we think that risk is low anyway with an oral 15 milligram dose, but even lower when you reduce the Cmax by half and delay the Tmax with this modified release. That all benefits our safety profile.
I see. You have completed that phase I study with the modified release, and you plan to share the whole safety data to the street come year-end or early 2027 in conjunction with-
Yeah.
Phase II blinded safety AEs. Is that the right?
Yeah. So there's really several key points here that I think underpin the profile, the safety profile. One is the phase I data, and part of that is just the PK component, right? Is the modified release a modified release? Then there's really two things that I think are most salient with respect to the phase IIb blinded data, and that is the discontinuation rate due to AEs and the proportion of people hitting the top dose, 3.2 milligrams. Those are really the things that I think FDA would be looking at to understand, based on the whole body of tolerability data, do we have the titration schedule right? Do we have the dose selection right? Therefore, we're proposing no change to either the dose or the titration schedule in that phase III, what would become the top dose.
We're adding a second lower dose, because the base expectation always is that you have a dose response analysis in some study, and this would be that study. But those are really the two numbers that underpin an understanding of tolerability and modifiable factors in the design that support tolerability.
If you've finished the phase I, are you able to confirm with us whether you have seen dropout due to AEs below, like, let's just say 15%, supporting your thesis here? Or do we need to wait? I mean, it feels impossible that at this juncture, that you didn't see something. It feels impossible to me, at least, that something went awry.
We're not proposing any change in the design from the phase IIb to the phase III on either titration or dose. We'll give a lot more detail, obviously, after the FDA. I don't want to get it in front of FDA review.
Okay.
But certainly, we're not proposing anything based on a lot of data that we're seeing.
Great. So that's around year-end or early 2027, where you're-
Yeah, we're just waiting for feedback.
Okay. Ultimately, we'll wait for the phase IIb data set on efficacy, but there's a great body of evidence to suggest you might exceed what you've powered for. Let's just say, it somehow didn't work out. Your phase IIIs would be going on. Is there a way to tweak, adjust, amend midway?
Yeah. I mean, the likelihood, given where we're powering and where historical studies have been of it not working out on efficacy for effect size reasons, you would think is low. The way we powered it, every single pramipexole study ever, effect size at one milligram or more would be statistically significant. This is very deliberate in terms of our powering. The phase IIb is at 80% power. Phase III is at 90% power. So you could always increase the sample size to improve the powering further. I think that's probably the main lever. Maybe you think about the role of anhedonia, and is that an enrichment factor? But ultimately, we have nothing right now that would indicate that that would be a plan.
I see. The phase IIIs, I think one starting first half, the other one starting second half. Can you walk us through, or just give us a hint how long it would take to finish those for the top-line release? I'm just trying to gauge when this ALTO-207 could be approved ultimately.
Yeah. So we'll give more guidance when we start the study, but you can look at other TRD programs that usually around two and a half, sometimes longer years in similar populations and similar types of patients. The other important thing to mention with that second phase III is that is a monotherapy TRD study. Importantly, to be able to avoid the prevalent risk of professional patients, especially in monotherapy trials, that will be essentially the same population of people who would otherwise get adjunctive treatments, verify that they're on antidepressants, that we measure in their blood or urine, then wash them off in order to then put them in a monotherapy trial.
It becomes very similar as a clinical population to our adjunctive population, but then supports a label that's really broad, that is monotherapy or adjunctive use, that speaks to the primary care doc thinking about AUVELITY, the psychiatrist thinking about CAPLYTA, REXULTI, and so forth. Because biologically, there's really no reason why this drug, and frankly, every drug that we've studied in the field, only works as an adjunctive or monotherapy treatment. That's the way we thought about our development program. That's the way we thought about these iterative phase IIIs.
I see. Yes, you feel confident on success. Placebo is always Achilles' heel.
Can you talk about study execution, what you have done to ensure placebo response is not going to be very high?
Yeah, 100%. This is a huge area of concern, and it traces back to a couple of factors that you control, right? There are some things that you cannot, but one of those is patient quality. We have been very explicit that we have put in a sponsor eligibility review where we internally, a team of psychiatrists, review every single patient. We require medical or pharmacy records to document their depression and their prescription. We require measurement of the drugs that they are taking in their blood or urine, and then we do that throughout the course of their treatment. That defines real patients you can confirm. A lot of studies in this space still do not do that, and I think that presents a tremendously high risk.
We have, on two different occasions, reported for the ALTO-100 and ALTO-300 programs that are ongoing, that doing that results in at or close to 100% compliance during the study. Clearly, that translates into patients showing patient-like behavior and taking the drug, and that is going to also serve to reduce placebo. In the ALTO-207 program, we also have central raters, and that allows us to have fewer raters in the number of sites, reducing variability, allowing us to keep tighter control over those raters. We have also done things like excluding patients who, over two MADRSs during screening, show a greater than 25% decline. We do not measure symptoms all that frequently, every two weeks during the trial, so you are not basically implying to patients that they should be responding sooner by virtue of how quickly you are measuring symptoms.
We have also developed and validated an AI rescorer, essentially, of transcripts of MADRS interviews, and we record every single interview in every, not just MADRS, but things like SCID as well, with our patients, and in fact, put that in front of FDA as well, and sought their feedback, and have been doing prospective validation of that tool. What that tool gives us is not just the ability to potentially even use that to score primary outcome, and that has been some of the discussion with FDA, but as a QA tool internally, in near real time, after an interview is conducted, the transcript is analyzed, and you can see, is that rater off from the other raters? Then why are they off? Are they not going deep enough in certain areas?
And provide that feedback to their supervisors to be able to really keep the performance of each rater tight. All of that reduces placebo rates, improves your standard deviation, which is really the driver of effect size and therefore statistical power, and just keeps a tight leash on the whole study execution.
Thank you. Very clear. end goal here, maybe a dropout rate, on safety, a dropout rate closer to what CAPLYTA showed of 9%. To be clear, do you expect any SAEs in this study? Could there be SAEs?
I mean, with this kind of population, there's always the potential of SAEs. But the things that the AEs that we're mainly keeping track of are the common things, nausea, vomiting, that leads to dropout AEs. That's really what is primarily on our mind. There are very infrequent things like impulse control disorders that can develop with chronic use of pramipexole that appear to be much more of an issue of Parkinson's, and historically, that's where that has been described. They seem to occur at somewhere between 0% and 0.7% in depression, and unlike Parkinson's, seem to respond well to just a dose reduction, people staying on the drug as opposed to needing to discontinue it.
We've seen that from both randomized trial and case report data. That is really the main thing that we have on our radar, but as I said, it's something that you see at a much, much lower rate with depression. It's something that's also on the label for aripiprazole, and you see with aripiprazole at a similar rate in depression as with pramipexole. So, it's a contextually important thing.
It's also an on-target effect when you activate somebody's reward system. That's how you treat their depression, and you may just need to adjust the dose in some people so that they can get the right level of reward system engagement.
Okay. Bottom line, the benefit-risk profile seems quite favorable regardless.
Certainly feel that way.
Okay. I think that's all the time we have, and you have two other data readouts next year. We'll get to that next time.
Stay tuned. It's coming.