For those who do not know me, my name is Mike DiFiore. I'm one of the senior biotech analysts at Evercore ISI. Welcome to our conference. Right now, we have the pleasure of hosting Acumen Pharma, and with us in attendance is Dan O'Connell, CEO. Dan, welcome. Thanks so much for making time for us at the conference and for being here. Before we delve into Q&A, just maybe perhaps kick it off by giving us some business highlights and what to look forward to in 2025.
Thanks, Mike. Thanks to you and your colleagues for inviting us to the meeting this year and joining the conference so far. Sure. In terms of business highlights, this was a big year for Acumen. Acumen, as a company, is dedicated to advancing or bringing forward better treatment options for Alzheimer's patients, and we're doing that principally through the development of sabirnetug, which is a monoclonal antibody that has high selectivity for A beta oligomers, and over the course of this year, I mean, the big lift for us was to launch a global phase two study, which we initiated in the second quarter and have now guided the complete enrollment of a 540-patient study in the first half of next year. The other key metric for us was to conduct a phase one study in a subcutaneous option of sabirnetug, and that study was in the fourth quarter.
We'll read out in the first quarter of 2025.
Okay. And that's using Halozyme's?
That's using the Halozyme. Yeah, exactly.
Got it. Okay. Now, a lot to dig into here. I know you guys, sabirnetug, just to zoom out for a bit, you have a unique targeting profile. You target low molecular weight oligomers or both or?
Yeah, we actually, so sabirnetug was actually developed in partnership with Merck going back a number of years. And Merck had onboarded some technology out of Northwestern University, excuse me, that really highlighted the potential of an agent and antibody targeting these soluble diffusible aggregates of the A beta peptide that had a propensity to bind to neurons at synapses and cause deleterious effects, essentially being the trigger in the onset of Alzheimer's pathophysiology. And Merck advanced the program to sort of a preclinical candidate point and then elected to take a BACE inhibitor forward. And as a consequence of that decision, Acumen picked up the entirety of the program. We own it.
It's a wholly owned asset at this point, which has completed phase one testing in a patient study that read out last year and then is in the throes of this fairly sizable phase two study. So we think sabirnetug was raised against oligomers, these toxic species, and has high selectivity for the low-to-mid molecular weight is kind of what's been reported.
But not protofibrils, right?
We actually don't have protofibril data. We haven't tested that evaluation. The protofibrils are more of a linear construct than the soluble, more spherical aggregates that sabirnetug is known to target. And we think, you know, look, I think A beta exists in a variety of species. I think we can simplify that complex biology referencing kind of three major pools of soluble monomers, which is the largest form of A beta present. And then these soluble aggregates, namely oligomers, which include protofibrils, which are also soluble, and then fibrils and plaques. I mean, those are kind of the main forms. We just think that oligomers are the primary toxic agent in the onset of the disease. And by having an antibody that's directed at those species should yield a better safety and efficacy profile for patients.
Got it. You had a nice slide, Dan, in your R&D deck kind of just showing the range of molecular weight in terms of oligomers, in terms of how sabirnetug binds to it. It has nice binding, it seems, to ABO species weighing 10-11 kilodaltons and also from 100-200 kilodaltons. But there seems to be a gap between, say, 11 and 100 kilodalton molecular weight or kilodalton range. Is this concerning? I mean, are you kind of leaving any efficacy on the table because you're not capturing ABOs in this molecular weight range?
Right. Yeah. So I think, again, I want to reference the complex biology of A beta generally. So I think we had an R&D day on October 2nd. I'd encourage anybody that cares to dive into the science to take a look through that deck. We had a couple of external speakers participate in that day. We did kind of want to kind of orient people on the history and the rationale for targeting oligomers. I think in reference to that, it's some human AD brain tissue where there are these spikes of lower and mid-domain oligomers. I think it's to be determined whether we're missing anything in between. I think there's other data that we have to show a broad spectrum of binding across various different molecular weights.
Okay. Obviously, too, just to kind of round out the whole construct, you have an IgG2 construct with reduced effector function.
That's correct.
Why was it designed this way? Because people think about, you know, crenezumab had a very similarly reduced effector function, didn't really show an efficacy. But obviously, crenezumab bound everything where you guys are much more targeted.
Yes. Yeah. So yeah, the history of crenezumab is kind of a fascinating history to reflect upon. So that's an IgG4. So it's actually silent entirely and definitely has a high propensity to bind a monomer. So a monomer is a normal physiological protein and doesn't have a pathological role other than when it's in its aggregated form. So I think the IgG2 decision was actually one that was undertaken when the program was still at Merck. And Merck was intentionally looking for reduced effector function, looking for ways to avoid vasogenic edema was the terminology at the time. So we think, so the IgG2 has reduced effector function. It still has some effector function. And looking at our phase 1 study, we did have a couple of cases of ARIA, five out of 48 subjects exposed to drug.
We still think that has a, we're running a larger study to sort of determine what the rate of ARIA in a broader chronic-based study will be. But we do think that having some active Fc function is part of the mode of clearance and neutralization.
Got it. I mean, so looking ahead here, I mean, if we do see a good degree of plaque reduction, would that be explained by some sort of peripheral sink mechanism potentially?
Yeah.
Did you guys identify plaque?
Yeah, I think we would more characterize it as a shift in the equilibrium. So not peripheral, so a central sort of shift in the distribution of the A beta species. And as you can see in the R&D, the oligomers are proximal to plaques. And so by removing or neutralizing, essentially clearing those species, there may be other biological means that come into play. It may also be that we bind some degree of plaque in various instances, and there may be some direct clearance as a consequence of that interaction.
Got it. Got it. And the fact that it does not bind monomers, which are even more abundant in everywhere, if it did bind monomers, monomers would essentially kind of outcompete the actual oligomers, and you would get reduced efficacy. Is that the whole theory?
I mean, that's the explanation of crenezumab, that and the IgG4 potentially. But I think that the monomers are 7,000-fold higher in prevalence in an Alzheimer's patient. And so to the extent that your antibody is distracted from toxic species, yeah. And given these are, we're administering IV every four weeks. Given the relative amount of antibody that's getting into the central compartment, you just can't afford to waste it on a non-pathological target.
Okay. Just getting into the phase 1 INTERCEPT-AD results, patients were only followed for around 60 to 75 days, if I'm not mistaken. But you did see some interesting results in terms of plaque reduction. I mean, I think you compared to all the competitors, and at least over those initial 60 to 70 days, it looked pretty competitive in terms of plaque reduction. So my question is, could that have been influenced by the relatively later stage Alzheimer's patients enrolled in this study?
Yeah, a good question, Mike. I think that the population, you know, this is in the early AD population. So it's kind of folks that have established A beta pathology. I think it was important that we ran this study, both the SAD and the MAD in patients because we really did want to get safety target engagement evidence of the profile of the antibody. And we did that. And what we were a little bit astonished to see was the degree of the shift in the PET signal for the two high-dose cohorts and the MAD cohorts where the slope and the rate of clearance is consistent with Leqembi in particular. I don't know that these are patients that are more advanced and that I think it's just, in fact, I think the baseline centiloids in the phase one study were consistent with Clarity AD, the Leqembi phase three.
Okay. I don't know where I got that more advanced data point from, but I'll look into that.
I think what is interesting, I think it's important for us to establish that A beta treatment is sort of the cornerstone of what is likely to be the path to treatment for these patients. We've got a couple of approved products. They're pretty well established as some of the limitations in terms of risk-benefit profile. And we do think that agents like sabirnetug can usher in a better treatment modality for these patients.
Also, to talk about the novel assay that Acumen developed to demonstrate target reduction in the phase 1. I mean, since you guys are so targeted and bind such a particular species of oligomer, you kind of have almost a custom-made biomarker assay to kind of just track target engagement. Are you guys the first company to ever develop an assay that's tailored specifically to your antibody?
Yeah, well, you know, it's kind of, there's some interesting history. I mean, firstly, when the program was still at Merck, they ran target engagement assays in transgenic mice. And so sort of the concept of the target engagement assay was established in the non-clinical development of sabirnetug. When we kind of picked up the program, actually, part of the core group of the company is a group that had previously been at Eli Lilly, and they had a lot of experience with biomarker development and target engagement assays. And so what we elected to do was to develop an assay that measures the complex of the drug, sabirnetug, bound to oligomers. And this is an ELISA-based methodology.
And what we were able to show in the phase 1, even in the SAD cohorts, is beyond once you went over two mgs per kg, you actually developed or identified signal of target engagement. And then with the multiple doses, we saw significant effects and really almost at the highest doses, kind of a plateauing of target engagement. So that helped us think about, we don't need to go higher to get broader target coverage. And in fact, we've bracketed, I think, the efficacy range in the phase 2 study.
Got it. Got it. Before we get into phase 2, last question on phase 1, just the ARIA rates. I think the overall ARIA-E rate in the phase 1 study was around 13%. Again, on one hand, it's a sign that the drug's working. But on the other hand, how can you explain this? Is this just, could this be totally attributable by the fact that you have some effector function, as you said before?
Yeah. Well, I think, Mike, it's probably a bit of effector function. It's a bit of plaque clearance and/or plaque interaction. And I think the rate, it's a small phase 1 study. So I typically say the frequency was, we had five cases out of 48 subjects that were exposed to drug. And three came at the highest dose level of 60 mg per kg. One of those three cases was mildly symptomatic. The patient did fine. But that's also helped us establish sort of that dosing window where we can actually deliver higher levels of drug and don't seem to run into an ARIA safety limitation or ceiling, such as the more plaque-directed antibodies, the current agents that are approved and some prior agents, such as like aducanumab and so forth.
Got it. And what's surprising is that out of the very limited number of ARIA cases that occurred in the study, I mean, no ARIA-E occurred among the eight E4, E4 carriers or the homozygous patients in the study. Was that just pure luck or is there anything more that you want to tell?
I mean, it's a small n. And I think it's obviously something we observed in the phase one. We're going to be curious to see what the phase two yields. It could be that as a consequence of this oligomer epitope for sabirnetug, there's different species or conformers in the E4 homozygotes that sabirnetug does not see from an ARIA perspective. So I think we have more to learn in respect of what ARIA. And I think in particular, if sabirnetug does present a low rate of ARIA in those E4 homozygotes, I mean, that would be a real game changer for the study.
Sure. Sure. Okay. Now moving on to the phase two, three, I think, or just the phase two ALTITUDE-AD trial. Okay. Just standard question, what's the powering and effect size assumptions of the trial if you could disclose those?
Yeah, we haven't disclosed those. I mean, we've characterized the study as an adequately powered phase two study. It's a robust 540-subject study with two active arms versus placebo. We're conducting it in a registration quality way. So we think this potentially could be part of a filing package in conjunction with a confirmatory single registration phase three study, and we're looking from an outcome perspective at the iADRS, which is a composite score that includes both cognitive and functional measures. We'll also be looking at both imaging and fluid biomarkers and other clinical measures such as the CDR-SB.
And also too, the doses you're using, you're using 35 mg per kg and 50 mg per kg. How did you arrive at those doses?
Yeah, so like I said, we kind of bracketed the efficacy range, 50 being a little bit lower than the top dose in the phase 1 and 35 being kind of a mid-dose that we think can also be efficacious. I think the dosing decision was really predicated on the target engagement, that plateauing effect observed even at 25 mg per kg and higher, and then modeling that suggested that we would have adequate target engagement even at peak and trough levels at those two dose levels of 35 and 50.
Yeah. And you have a nice slide in your R&D deck kind of showing those CSF target engagement plots. And it seems like that had you gone higher than 50 mg per kg, it wouldn't have gotten you that much.
It wouldn't, and maybe a little bit more ARIA, right?
Yeah.
Exactly. It could be an off-target side effect.
Okay. Okay. Also too, prior and early in the conversation, you mentioned you're developing a subcutaneous version using Halozyme's technology. Have you disclosed what IV and subcu doses are being tested in the current PK study or no?
We have not. And I think we're still anxious to see that data and see what the path forward might be for subcu. We also are going to be predicating on a subcu strategy on the basis of the phase two IV data, right? I think that's going to be instructive. So that information should be available in the first quarter. And we're excited to see what the reveal is and what the possibilities would be for a subcu version of sabirnetug.
Okay. Wait, what's the half-life of sabirnetug?
The reported half-life is like two weeks.
Two weeks.
Yeah.
Okay. Okay.
Maybe 21 days, 14, 21 days.
Got it. And assuming bioequivalence and assuming success in phase two, will the subcu version be taken into phase three, presumably?
Again, it's TBD. I think it's, I mean, there's some precedent out there to look at including a subcu as part of an open label extension in a phase three. But we are, again, waiting on data and determining before prescribing what we want to do with subcu.
Got it. And just kind of one general question to kind of just tie this all together, Dan. I mean, just when you think about the generations of amyloid beta antibodies that have been developed that have failed, but that are not on the market and that will be developed, you guys obviously fall into that later camp of next-gen amyloid Aβs. And then you think about new technologies like Roche's trontinemab, it's not unreasonable to think that maybe in five, ten years, the Alzheimer's space will become kind of crowded in terms of Aβ mAbs. What are your thoughts on that? And how do you think sabirnetug could fit in there?
Yeah. No, I mean, I think it's an evolving landscape, right? And I think that some of these delivery modalities that Roche and Denali and others are employing are generating some interesting early data. I mean, at least the trontinemab data is interesting. I think whether or not that becomes sort of the mode of treatment is still TBD. It's still kind of early. Our view is that A beta strategies such as sabirnetug are going to be the cornerstone of treatment and that we will continue to iterate using biomarker information and potential other combination strategies to amplify the overall treatment benefit for Alzheimer's patients.
Got it. Got it. And just to level set too, since we're kind of getting up in the final minute, minute and a half, catalyst flow over the next 12 to 18 months?
Subcu phase 1 data, Q1. If I haven't mentioned it, we will complete enrollment in the phase 2 in the first half of next year. Then we'll be in line of sight to data within roughly 18 months from completion of enrollment.
I see. I see. Anything I didn't ask, Dan, that perhaps are more nuanced questions that you're getting from investors that you'd like to voice here?
Yeah. I mean, I think a point to emphasize is that a relatively small biopharma company such as Acumen has undertaken this fairly sizable phase two study and the enrollment rate has been spectacular. It's been really strong. So I think that's an attribute to the team and the ability to execute, but also the unmet need, the awareness that patients are looking for other and better treatment options and research opportunities. So I think it's been, I think the forward-looking analysis of this space and the long-term legs that A beta strategies will have is very strong. It's very healthy.
Got it. Got it. Final question for me before we have to cut out. Current cash position and what's your cash runway look like at this point?
Yeah, so at the end of the third quarter, we had about $260 million in cash and we've guided cash into the first half of 2027.
Got it. Got it.
So we haven't guided really on data readout, but I think one could infer from first half completion, cash is beyond a data readout.
Got it. Hey, there's a lot to look forward to, Dan. Thank you so much for making time for us. I think we're at time, but learned a lot and looking forward to keeping tabs on the story. Wish you the best of luck.
Thanks, Mike. Appreciate it.
All right.
Appreciate it.