All right. Good morning, everybody. Thanks so much for joining the third panel. We've had a few really cool discussions so far. It's my pleasure to be hosting a chat with Marc Grasso, CFO of Alector, and Gary Romano, Chief Medical Officer. I'm gonna ask the team here just to kind of give a quick overview of where things stand with AL001, AL002, and any pipeline efforts, and then we'll do a deeper dive into the 002 program ahead of data of this year. So thank you, guys. And maybe, Marc, would you like to take it away? I appreciate you joining.
Great. Yeah, thanks, Paul. We appreciate the opportunity to participate again in your CNS Days. As a brief reminder, Alector's approach is to harness the power of the brain's immune system. We're seeking to offer broadly acting and potentially transformative therapeutics for neurodegenerative diseases. This is grounded in substantial genetic evidence. Just to recap, latozinemab, or AL001, is our most advanced program.
It's a drug that elevates the level of progranulin, and this is a secreted immunoregulatory protein. Its partial deficiency causes frontotemporal dementia and increases the risk for AD and PD. We have completed enrollment in a randomized double-blind placebo-controlled phase 3 pivotal study in frontotemporal dementia. We recently received breakthrough designation from the FDA for this program. Our second progranulin-elevating program is called AL101, and we've commenced a phase 2 study in early Alzheimer's disease.
Both of these programs are part of a worldwide partnership with GSK. This included $700 million in upfront payments, has $1.5 billion in development and commercial milestones, as well as significant profit sharing in the U.S. and royalties approximating the 50/50 profit share outside the U.S. Our third program is AL002. This is a drug that activates TREM2. This is a significant risk gene for neurodegeneration.
Its partial deficiency increases the risk for Alzheimer's. AL002 has completed enrollment in a multidose double-blind randomized placebo-controlled phase 2 in early Alzheimer's disease. We anticipate data later this year, and this will measure a range of clinical biomarker and imaging outcomes. This program is part of a partnership with AbbVie. They've paid over $250 million in upfront, and there's significant milestones they've paid.
They have an option at the end of phase 2, where if they opt in, there is a $250 million payment to Alector. This is a 50/50 global profit share from that point on. We're also progressing a novel portfolio of additional candidates supported by our Alector Brain-Carrier technology, also called ABC. If we have time, we can certainly talk about those as well. I think that's a brief recap.
Awesome. All right. So let's talk about the science behind TREM2, and I think the one sort of major debate with this target, right? So the genetic evidence, which I'm sure you'll speak to, is super, super interesting with this target, and I think informs a lot of probably the conversation here on why to pursue it.
I think the one challenge that some of the KOLs we've talked to have highlighted, including those who've been involved with companies in the space and even earlier stage work in these programs, is this question around inflammation in the brain and this subjective delineation between good inflammation and bad inflammation. And is there a time point with TREM2 where it's optimal to intervene versus suboptimal? I guess, how do you kind of go about trying to answer this question?
What got Alector and AbbVie excited about taking the genetic evidence behind TREM2 and pursuing this mechanism in a broader, I guess, for lack of a better word, idiopathic Alzheimer's population?
All right. Glad to welcome Marc. So yeah, so thanks, thanks, both. Good question. I guess Alector's approach has been really to follow the genetics, as you mentioned. And maybe somewhat simplistically to say, if the genetic mutations that increase risk for neurodegenerative disease are moving the biology in a particular direction that, let's say, as in TREM2, that loss of function mutations are causing increased risk of disease, then we should try to modulate that activity, that biological activity, in the opposite direction, in a direction that could be protective or preventative of disease.
I mean, we can cite the data that I'm sure you know, showing that high levels of soluble TREM2, reflecting TREM2 levels in humans, seem to have beneficial effects of slowing disease, whether measured by various imaging biomarkers or clinical outcomes.
In a variety of animal models, there are a lot of animal models out there, right? You're right. Some of them have shown clear beneficial effects of increasing TREM2 signaling. Some have suggested that microglia and activated microglia could eventually have a pro-inflammatory state, which could have a negative impact in biology of the disease and potentially in patients.
We just took the approach to go to the clinic and follow the directional genetics, as I mentioned, and carefully and with close safety monitoring, and see what - see whether there's benefit. I think there's been a lot of hand-wringing in this field for over a decade, worried about the animal models. Animal models only take you so far. In our study thus far, as you know, we've completed enrollment. We have patients that have been dosed now.
We just heard about a patient that's now been dosed for 3 years, 2 years in the double-blind study, and then in the long-term extension, and many who have been dosed for up to between 1 and 3 years. The safety signals have been, the safety has, knock on wood, been very good. We do have an ARIA signal, which we find intriguing, but so far, it doesn't seem we're doing any harm. And based on our unblinded data.
Okay. All right. That's great. We'll ask about the ARIA signal in a second. But one more question on this basic science debate here. And I'd like that you reference the sTREM2 data. And for those listening in, and Gary, you correct me if this is an oversimplification.
There's data from the Alzheimer's Disease Network Initiative that shows that CSF sTREM2 levels are inversely correlated with cognitive decline, right? And that's an all-comers Alzheimer's. Is that correct? I guess one thing when we've tried to interpret those data is really two questions. One is sTREM2, I think, right, is a byproduct of TREM2 shedding, right? So how do we know that higher sTREM2 in the CSF means more TREM2 in the brain? How do we not know it's the opposite, where it means that more is actually being degraded?
Then second, do we know if sTREM2 itself is just this inert byproduct, or does it actually have its own function?
Right. So a lot in there. I'll try to unpack it. So you're right. Soluble TREM2 is a cleavage product of TREM2. It's constitutively cleaved, irrespective of what's happening, whether TREM2 is signaling or not. So we, and I think many others, view it primarily as a surrogate, if you will, for how much TREM2 receptor there is in the microglial membrane.
There's some data out there, some suggestions that it may have, or some hypotheses that it may have other functions. But we see it primarily as a surrogate. And so, soluble TREM2 levels then reflecting increased TREM2 in the membrane, greater sensitivity for TREM2 to, or greater likelihood of TREM2 to be active in activating signaling pathways, which reduce the vulnerability of the brain to insults, including neurodegenerative diseases.
Okay. Makes sense. With the ARIA signal, so we agree that that's super intriguing. I guess a couple of questions there. At face value, does it make sense that agonizing TREM2 would actually lead to immune-mediated clearance of misfolded proteins in the brain, and that, as a result, would cause sort of the same vasodilation that we see with the A beta antibodies? I mean, really, what I'm trying to say is taking a step back. Is this an unsurprising finding in your mind?
No. I mean, we wouldn't, right. So what you're asking is, with activating, is it reasonable to think that increasing TREM2 signaling could cause increased phagocytosis and clearance of misfolded proteins? And we think so. So we're not, we probably shouldn't be surprised to see amyloid clearance in the presence of TREM2 signaling.
That said, and I want to emphasize this, we see that as just one piece, just the tip of the iceberg, if you will, one aspect of the benefits of the many increasing beneficial effects of microglia that we would see by increasing TREM2 signaling. So it increases microglial proliferation, survival, phagocytosis, and also all the other maintenance effects of microglia on brain health, including supporting synaptic function, astrocytes, oligod endrocytes, the maintenance and repair of the blood-brain barrier and the vasculature.
So we see all that the downstream effects of increased microglial functionality being beneficial in a number of ways. Mechanism. So amyloid clearance, yes, maybe. We'll see. We're definitely going to have the biomarkers to answer that question. But that's just part of the story.
Yeah. So when we've talked, it sounds like the DSMB of this study, right, closely vetted this ARIA signal and is comfortable, right, that you're not doing harm. Can you talk a little bit more about that? What data do they get to see? How do we know—what do we know about how the patients who had the more severe cases are actually doing?
M aybe you could just kind of expound upon why, at least in Alector's view, assuming I'm characterizing this correctly, that this ARIA is functionally, or at least likely, no different than the ARIA we're seeing with A beta antibodies and not some sort of immune reaction that is related to TREM2 and is not driven at all by plaques.
Yeah. So thanks for the question. So you're right. Our IDMC, of course, is looking at this data in an unblinded fashion. We're still blinded. What we can tell you just from the blinded data is that when we eliminated the APOE4 homozygotes from the study, early in the study, the incidence and the severity, both radiographic and clinical, of ARIA has come way down. And I think probably the—I mean, I could cite some numbers for you. So roughly 10%. So of all the ARIA that we identify with MRI surveillance, only about 10% is symptomatic.
A large majority of those have very self-limited and mild symptoms. And importantly, the number of the percentage of patients that have clinically serious ARIA, that is, that ARIA that's related to an SAE, that is, these are subjects that may have some lasting neurological deficits or even worse. That's less than 1%.
So we do feel that this—so hence, we're going forward with our study. We think that with the removal of the APOE4 homozygotes, we're very much in line with what's been seen with the anti-amyloid antibodies. And in fact, in many ways, in every way, really, that we can measure so far, this is indistinguishable.
The IDMC, they are aware of dose relationship, which we're not yet because we're still blinded. One thing I will mention, though, is that we've also asked them, going to this idea of pro-inflammatory effects of microglia, we've asked them to keep a close eye on clinical outcome measures, not for efficacy in a sense, but for safety.
If there was a pro-inflammatory effect, that what we were seeing on MRI was not ARIA, but something different with pro-inflammatory effects, we would expect it to maybe have clinical effects, which they have told us they've been watching this carefully and have told us to continue. That means that they're not seeing a concerning signal there.
Then lastly, we're seeing ARIA, just like with the amyloid antibodies, we're seeing it very early in treatment, after the first dose or first few doses, after which it drops off very precipitously. Again, if there was a pro-inflammatory effect, I would imagine that this is something that might increase over time and not have all the same characteristics as what we're seeing with the A beta antibodies.
Okay. All right. That's great. Thanks, Gary. So I guess taking a step back, right? I mean, this study has a unique design, right, where there's patients followed up to different durations, and you're going to have some data cuts, drug versus placebo, that are really well-powered, some that are maybe less powered but still informative. So you want to just talk about the trial design, how much data you're going to have at 12, 18, 24 months. And when we get this top line, how will you and AbbVie kind of be depicting the efficacy data and rank ordering this in importance?
Yeah. Thanks for that question. So yes, it's a, so for everyone's, just to remind everyone, this is a common close design in which patients will have, every patient, every participant will have a minimum of 48 weeks of exposure and up to 96 weeks of exposure. And we're using a proportional analysis method so that when we analyze the common close design, we're not just doing a time-to-event or time-to-fixed-duration study, rather, of just 48 weeks, but it'll include 48 weeks, 72 weeks, and 96 weeks.
T he distribution of those patients, it's somewhat heavier weighted at the end of the study at the 12 or I guess it's 48 weeks, 72, and 96 weeks. So we'll have somewhat more at the 48 weeks and then lesser amounts at probably going to be distributed somewhat like one-third at 48 weeks.
Then close to half of patients will have somewhere between 48-72 weeks. Then the rest will be out at 72-96 weeks. That's due to the fact that there was a big push. There was a lot of enthusiasm and a big push of enrollment towards the end of the study. So using that analysis method, we'll be using all of the data.
You asked a question about how are we looking at this? How are we looking? How are we going to make decisions? This study was designed to be a biomarker-rich study from the beginning when it was designed three years ago by AbbVie and Alector. It's even more so now owing to the really rapid advances in plasma biomarkers, particularly plasma biomarkers for phospho-tau species.
We'll be looking to show that AL002 slows the progression of Alzheimer's disease with a combination of, as measured by, a combination of clinical, functional, and biomarker readouts. That'll include, of course, our primary endpoint, but also other clinical outcome measures and also very much so the biomarkers, particularly the Alzheimer's biomarkers.
Yeah. Okay. Okay. Fair enough. Let's see. Anything else you'd like to add on AL002? I mean, I think the only other question that comes up on my end, on the investor side, is just we've seen these amyloid antibody, the launch of lecanemab, not launches. I guess donanemab hasn't launched yet.
But we've seen it go slow, right? It looks more like a rare disease launch than what you might expect from an Alzheimer's launch. And there's many impediments to that, but ARIA monitoring is certainly one of them. What's Alector's perspective on the A beta launch? And when we think about AL002, does this sort of, does it all change the clinical risk-benefit bar on what you need to have to have an exciting commercial product in Alzheimer's?
Well, we don't think so, not for this mechanism. So I think whether or not we lower amyloid, right, this was never designed or intended to be another way to lower it, just to lower amyloid, right? So we're not going to judge this based on, for example, a particular centiloid level.
If we don't get to 24 centiloids, which seems to be a necessary condition for clinical benefit with the anti-amyloid antibodies, given that there are many other aspects of this mechanism of action of restoring microglial function, it's not going to be thumbs up or thumbs down based on reaching that threshold. And I think with regard to how does the emergence of the anti-amyloid antibodies affect our development plans? Well, we have a monotherapy study that we're going to have results on by the end of the year.
If this is a positive study and we're going forward, we want to push this forward. We want to go forward as monotherapy. I think in the current landscape, we don't yet have a standard of care. We've been able to enroll this study and another Alzheimer's study that's ongoing. I think that we will definitely have a monotherapy approach to development of this molecule.
In addition, though, there is the potential, we think, for this to be additive or synergistic to anti-amyloid antibodies or even tau therapeutics at some point because the mechanism is really going to be complementary. For example, this could be sequential treatment, potentially, of an anti-amyloid antibody and then the TREM2 agonist. There's a potential for additive risk of ARIA, of course. We'll have to sort that out.
But one potential approach to this, and we've discussed this with a lot of thought leaders in the field, could be to treat first with an anti-amyloid antibody to lower amyloid levels back towards normal and then come behind with the TREM2 agonist as a maintenance therapy. And with amyloid levels being lower, we'd have less risk of additive effects on ARIA, I think. But we're also exploring that in animal models as we speak.
Okay. Okay. All right. Great. Well, we look forward to that data. In the meantime, the AL001 program and FTD continues to advance, right? You're at a pivotal study. We're about a year and a half away from the readout of that trial, right? Not to get too in the weeds on the science, right? But I think the one mechanistic question that investors who've looked at the progranulin space grapple with is just sortilin and is sortilin blockade the right way to upregulate progranulin?
Are you increasing the concentration of progranulin in all the key areas? What got Alector comfortable that this mechanism is more like an SSRI for progranulin versus just sort of almost like shifting the cellular biodistribution of the neurotrophic factor?
Yeah. It's a good question, Paul. So sortilin is one of a number of receptors, neurons, and glial cells that uptake progranulin. And our approach and the data supporting this is that if you can block sortilin, you can increase the levels of progranulin extracellularly. And this will not prevent progranulin from entering the lysosomes.
T he hypothesis is that through this blockage, you're going to increase progranulin levels, and you're basically going to maintain the progranulin that's necessary for lysosomal function. We've shown in preclinical studies that you can block sortilin, and you're not having a deleterious effect on progranulin levels intracellularly or lysosomal function.
Further, in the phase 2 study, which is the basis for the recent breakthrough designation for this program, we're showing that with blocking of sortilin, you can increase progranulin levels in CSF twofold, and you can normalize a number of different indicators of disease progression with this normalization of the lysosomal function, inflammatory mediators, and also showing a slowdown in cognitive decline as measured against matched historical controls.
Okay. Great, Marc. You guys did a blind interim analysis to this study. You found that the variability was a lot lower. Approached the FDA, and now you're going to readout this trial, right, at a lower end than originally planned. Maybe can you talk a little bit about that? And Gary, you and I talked about this a bit in January, right?
But I think my one question to you guys was lower variability is great and the simple math of powering, but is that at all bearish for seeing an efficacy signal if there's kind of less variance? So what's the right interpretation of that update? And I guess what gives you guys the confidence that this study is still a valid, well-designed test of this hypothesis?
Right. Right. So thanks. So we learned back when we—I think it's about a year ago—before we met FDA and EMA, we examined some of our original assumptions of variability of disease progression against data that was out there in observational studies. And based on that data, it looked like we were really being overly conservative about the variability that we expected to see in our study. And that allowed us to go with a lower but sufficient sample size. So as we mentioned last time, I think we did two things.
One was that we separated out. We originally planned to examine or to assess both preclinical subjects, that is, those who are at risk but are still presymptomatic, and also patients who were symptomatic. The presymptomatics brought a lot of variability to the table. The CVs were well over 100.
Whereas the CVs for symptomatic were lower and even lower than our original assumptions. So, for example, we originally thought that the assumption was that the CV for the symptomatics was somewhere around 77%. We now realize that it's somewhat lower, but not drastically lower. It's about 60%-65%. So what I'm trying to say is what we've done is eliminate the presymptomatics, which makes sense.
We're going to analyze them separately. We'll do sensitivity analyses to look at that subject, but they won't be part of the primary analysis. In the primary analysis, we now know that the CV in the symptomatics is lower. It's about 10%-15% lower than where we were. But there's still plenty of variability in room to see a difference between placebo and treatment effect. Of course, we're blinded to that, so we haven't looked at that.
But we're very confident that there's plenty of room to see differences between treatment and placebo.
Okay. Okay. Great. Well, maybe just to wrap things up, Marc, do you want to talk about the company's cash runway and your comfort that you're well-funded through both of these readouts?
Yeah, of course. Thanks, Paul. So from a runway perspective, we performed a raise in January, we have $620 million as of the end of December. And this is runway through 2026. This is a full two years beyond the anticipated TREM2 data readout, approximately a full year beyond our latozinemab phase three data readout. And this doesn't assume any additional milestone payments, including from AbbVie on the opt-in or from GSK.
I mportantly, this cash balance also allows us to advance our earlier programs and Alector Brain Carrier platform as well. So we think we're in a strong position as we progress, and we're looking at it from a conservative perspective. And obviously, the opt-in from AbbVie and potential future milestones from GSK would extend things further.
Okay. Great. Well, thank you guys very much. Congrats on all the progress. Appreciate it. And thanks for the great attendance on this panel and for the questions that were sent in. And we look forward to talking to you guys soon.
Thanks. Great event. Thanks, Paul.
All right.
Thanks, Paul.