Alector, Inc. (ALEC) Earnings Call Transcript & Summary

And thanks for joining us today, the Goldman Sachs Global Healthcare Conference. Really pleased to be joined by the team at Alector to kind of talk to the story where things are -- where things are headed. So I'd say, let me start at a very high level.How would you broadly describe Alector's approach, particularly in neurodegenerative disease, but also in oncology.

Sure. So as you know, that the dominant approach for treating all neurodegenerative disorders for the last few decades was to go after misfolded proteins. So there have been more than 50 clinical trials with drugs targeting the beta amyloid pathology, there have been 15 or 20 clinical trials going after the TAU pathology, there have been clinical trials going after huntingtin in Huntington's disease, [ senior ALS , SNP ALS ]. And so far, the majority of these trials were quite negative or marginally positive. We, 8 years ago, decided to take a completely different therapeutic strategy. And what we do is recruit the brain immune system to counteract multiple disease pathologies in brain orders. So conceptually, we are very similar to immuno-oncology. What revolutionized immuno-oncology was that instead of trying to kill tumors with irradiation or chemotherapy, we now recruit the immune system, primarily T cells to counteract cancer. We are doing the exact same thing, but in the brain, instead of going after individual pathologies like Abeta or tau, we are recruiting the brain-specific immune system to counteract all the disease pathologies. So we think that our approach which we designated, immuno-neurology could have as much impact on brain disorders as immuno-oncology has on cancer. And with this approach, we find close to 400 patent applications, we have multiple drugs in the clinics, dozens of immune checkpoint targets that we are pursuing, and we really hope to revolutionize brain disorders with this general approach.

Excellent. Thank you. So with that context in mind, let's start with the company's lead program, the sortilin monoclonal antibody AL001 for progranulin mutant frontotemporal dementia, we are going to call it progranulin FTD for short. Could you Walk us through some of the mechanistic rationale for increasing progranulin levels, particularly by sortilin blockade as a treatment for progranulin mutant FTD?

Sure. So progranulin is a very interesting risk gene for neurodegeneration. It is secreted protein that is immunoregulatory and survival factor for [ neo ] cells. So if you don't have progranulin, your immune system in the brain sort of go arrays, it becomes hyperactive and start damaging and destroying nerve cells instead of protecting them. and progranulin was shown to be a risk gene for practically every neurodegenerative disease that was studied. It's a regime for frontotemporal dementia. It's a regime for Parkinson's disease. It's a genetic regime for Alzheimer's disease, it's a genetic risk for ALS with C9 as well as sporadic ALS. So it seemed to sit very high at the regulatory cascade of neurodegeneration. It's one of the very few genes that are not disease specific. And in all cases, decreasing the level of progranulin leads to increased risk for neurodegeneration. So for example, frontotemporal dementia that -- in people that have the progranulin mutations, these people have one good and one bad copy of progranulin. They have 50% of the normal level of this immunoregulatory protein and they invariably develop frontotemporal dementia before the age of 60. And people that have regulatory mutations that decrease the level of programming just by 10% to 15% are at a higher risk of developing Alzheimer's disease, Parkinson's disease, ALS or LATE, which is a late form of dementia associated with TDP-43 pathology. So because loss of function of progranulin invariably lead to neurodegeneration, we undertook to develop a drug that elevates the level of progranulin in the human brain. And the way we did it is by targeting at degradation receptor for progranulin, this degradation receptor is a single [ trans memory ] receptor called sortilin and human genetics showed that mutations that decrease the level of sortilin elevates the level of progranulin in human without causing any other pathology. So we undertook to mimic the human genetic of sortilin ablation or sortilin decrease and develop a drug, an antibody drug that blocks sortilin, prevents its interaction with progranulin and prevents programming degradation. So we have shown that with this drug, we are able to safely and chronically elevate the level of progranulin in the brain and serum by two to threefold. And as long as the drug is applied, the level of progranulin is continuously elevated. And sortilin was shown by multiple animal model data and human genetics to be primarily at degradation receptor for progranulin. It doesn't seem to be involved in any of the programming functionalities, either the extracellular functionalities or the intracellular functionalities. So we think that blocking sortilin is a really good approach to elevate progranulin in the human brain. And conceptually, what we do is very similar to all the (sic)[ SNRIs ], basically serotonin and norepinephrine inhibitors that block the uptake of neurotransmitters and elevate neurotransmitters level by two to threefold. And this, as you know, sort of became very good therapeutics for anxiety and depression like Prozac and the entire cells of SSRIs. We are doing the exact same thing for progranulin, we elevate progranulin by blocking degradation cascade. And we are -- initially, we went to the clinic in [ STB ] that has progranulin mutations. We were able to restore progranulin back to normal level in these patients. And in an open label Phase II trial, and we were able to show that we normalize multiple biomarkers, including lysosomal and inflammatory biomarkers, we were able to normalize GFAP, which is a biomarker of neurodegeneration, and we were able to slow down brain volume decrease, like enlargement of ventricles, which sort of reflect neurodegeneration by over 50%. And comparing to age match control, we were able to slow down cognitive decline by close to 50%. With this data, we went to another subset of prostate dimension. These are patients with the C9orf72 mutation. And again, we were able to show that we elevated programming chronically two to threefold above the baseline level and we were able to slow down cognitive decline by 54%. We were able to reduce the level of neurodegenerative biomarkers like GFAP and the drug was completely safe over up to 2 years of treatment. And we are now in Phase III pivotal study in progranulin mutation carriers, the developed FTD, we are sort of expanding our trial in ALS, and we are planning clinical trials in Alzheimer's disease and Parkinson's disease because both the animal model data and human genetic indicate that elevating progranulin would be beneficial.

Okay. That was great. That was very comprehensive on the kind of data set. And I was kind of curious, from that Phase II trial, the INFRONT-2 trial, considered a 96-week study. As far as we understand, you presented 40 weeks of data in progranulin mutant FTD and C9orf72 mutant FTD. Could we expect additional data updates as those trials kind of roll -- continue on?

Yes. So in terms of the Phase II [indiscernible] this is working? Yes. In terms of additional data readouts from the INFRONT Phase II study, that study is ongoing. And we're looking forward to a complete data set for the FTD C9 cohort with the 12-month data set. We reported out on a subset of those data at AD/PD looking forward to that complete data set that will help INFRONT Phase III decision making.

Okay. Great. So -- can you remind us kind of the design of this Phase III INFRONT-3 trial that you mentioned earlier? Like how should we think about that study in terms of number of patients, key endpoints, that kind of thing?

Sure. So the INFRONT-3 is the Phase III study in FTD GRN patients. It's modeled after the INFRONT-2 study. It's a multicenter, placebo-controlled, double-blind randomized trial. The total target enrollment is 180 patients. Like you mentioned, it's a 2-year double-blind period, followed by a 2-year open-label extension. The -- It's a one to one -- I'm sorry, a 3 to 2 randomization from drug to placebo. And the primary endpoint there is the CDR® plus NACC, FTLD sum of boxes has been the endpoint in the INFRONT-2 trial.

Okay. So with this context in mind, when I come back to C9orf72 FTD. So you mentioned that with progranulin mutant FTD, you've had this mutation, you have lower levels of progranulin, sortilin blockade can kind of re-equilibrate that. But as you mentioned, in C9orf72, you don't have insufficient progranulin per se. So how do you -- can you walk through the rationale for why you want to have higher levels of progranulin in that context of normal basal levels of progranulin?

Yes. Yes. So there's rationale for FTD C9 cohort that comes from mouse models and also from human genetics. So we know that part of the rationale here is that with elevating progranulin levels, we can reduce the burden of TDP-43 pathology and mouse models of -- when you have a deficit of progranulin, you see aggregation of pathological dipeptides such as TDP-43 in mouse models of TDP-43 pathology by elevating progranulin levels, you can reduce that aggregation and prolong survival. We also see in clinical data sets from ALS patients, for example, that progranulin deficiency is associated with reduced survival and earlier onset of disease in patients with ALS or FTD associated with the C9orf72 repeat expansion.

Okay. Okay. So kind of from the C9orf72 FTD patient part, you mentioned that you tried to early look at AD/PD earlier this year. You're going to expand that recruitment to have a more complete data set later on. What are you to see from that kind of bigger data set to give you confidence to pursue a Phase III study?

Yes. So just as a reminder, what we saw in the data that we presented at AD/PD was essentially a very nice replication of what we saw in the FTD GRN data set. As Arnon mentioned, approximately a 50% slowing in cognitive decline, 54% relative to match controls from the all FTD database. We saw concomitant changes in biomarkers, such as GFAP, a marker of astrogliosis. And importantly, we also saw a very similar and favorable safety profile. So those data were very encouraging. I think what we'd like to see is when the full complement of patients reach that 12-month mark, just that nothing has changed from that interim data readout. We wouldn't expect major changes from that, but we'd like to see that full data set first.

Okay. Great. So at a high level, for AL001, a question we've gotten from a lot of investors over the last few months has been how to think about placebo effects in a disease like hereditary FTD. So how do you think about kind of placebo effects and their impact in the like C9orf72 FTD and progranulin mutant FTD?

Yes. I think whenever you have an endpoint, that's a subjective endpoint, you're always thinking about the potential for placebo effects, particularly early on in trial when you have contact with patients. And specifically when we think about behavioral variant FTD, you can sometimes see what looks like an improvement or perhaps small improvements that may be related to caregivers better able to manage patients or patients and caregivers better able to cope with the condition. In addition, as you see things like apathy and language impairment set in, you may see fewer of those disruptive behavioral symptoms manifesting that look like an improvement. So it really takes a well-trained radar to sort of tease some of this apart. And that's why we put so much time and effort into ensuring that the radars in our trials have the adequate training and experience to rate on the CDR sum of boxes. We provide centralized overt monitoring for the ratings to really try and help reduce that. I would say that, in addition, compared to some areas like psychiatry and major depression where we're really concerned about placebo effects. What we have here is a condition that unfortunately progresses quite rapidly. And so I think -- and it gets very severe at later stages. So I think particularly at those later stages of the trial, it's unlikely that a placebo effect is going to overcome that decline that you see in this condition.

If we look sort of, for example, in multiple Alzheimer's drugs that use practically the same CDR just with the [ report ], like if you compare all the anti-Abeta drugs to [indiscernible] like to historical Abeta -- in historical Alzheimer's patients, it doesn't -- nobody reported the long-lasting placebo effect with these CDR tests. And if you look at the FTD, there were 2 clinical trials, again, double-blinded, that failed. They were shorter. They're like 28 weeks. And again, if you compare the decline in this placebo-controlled trial to historical control, there was no reports that there was a meaningful placebo effect. So we don't think that there is a large placebo effect, especially over a year.

Okay. So on the CDR-NACC-FTLD sum of boxes that primary endpoint. So you mentioned there like behavioral features and the social features have this [ asset ]. So kind of on a more like numbers basis, what portion of that score do you think is exposed to those kind of early effects or caregiver effects that you mentioned earlier, Marc?

I mean I think it's hard to say. I mean, I would say that if you see some of that potential increase, it's likely to happen earlier in a trial than later. And so what we do see is, as Arnon mentioned across trials, whether it be an FTD or in Alzheimer's disease, you do see that decline. So I think the real risk of a placebo effect interfering with that decline in the placebo group is very low.

Okay. So stepping back, what is kind of the next cadence of news flow from AL001 or even AL101, the follow-on drug for sortilin? What additional data can we expect from those programs to kind of like guide the way forward as we wait for things like INFRONT-3 and other longer-term studies?

So yes, so as [ already ] mentioned, we are going to show another set of data from the C9orf72 early next year, let's say, again, larger cohort and longer treatment period. And then for the Phase III, we are going to report when we complete recruitment. And we have continued to update about our ALS trial that, again, will likely start early next year, but primarily as we are sort of accelerating to complete the Phase III, which is going to be a pivotal study and then to try to get approval.

On the ALS piece, obviously, ALS is a landscape that is shifting. And you may think a good job mentioning at the beginning that's both on a broad level on ALS or maybe new treatment modalities coming or not. And it is also on the targeted level, something like complicated data. So what do you make of the ALS landscape specifically in C9orf72 ALS and where one could fit in?

Well, so we've recently made a decision to sort of shift focus from a Phase IIa study that we've been doing in ALS C9 patients that was a smaller biomarker-oriented study to a more robust Phase IIb study that's in a more all-comers population. So I think that this gives us the opportunity to learn where AL001 might work better or worse. And again, the intent here is to more rapidly accelerate our development program.

Okay.

So in general, ALS -- 98% of all ALS is TDP-43 pathology with the exception of SOD mutation carriers and progranulin was shown to counteract TDP-43 pathology in animal models. And if you ablate progranulin, you actually induce TDP-43 pathology in adjacent neural, so there is significant mechanistic rationale for all TDP-43 pathology. So I think this is where AL001 can get in, but basically counteracting the pathology downstream of C9orf basically between C9orf and TDP-43 pathology. And the care treatment, again, there were several attempts to activate the C9orf, not particularly successful. We think that once TDP-43 is generated, it can become self-sustaining. So going to C9orf maybe similar to going after Abeta after tau, means I think you need to go a little bit downstream, and we think that, that's what our progranulin elevating drug will do. And you are right that the landscape is shifting, we are waiting to see what happened with the amylase drug, but our Phase II is going to be designed very similar to the amylase Phase II with all the lessons learned. So if the amylase drug gets approved, we hope that we can really have a similar design and try to see if we can get earlier approval in ALS.

Okay. Great. So I think we've really dug through on AL001 and will maybe shift gears over to AL002, the TREM2 program. So -- let's walk through the rationale for targeting TREM2 in Alzheimer's disease.

So TREM2 is really, again, a unique immune checkpoint to be targeted in Alzheimer's disease. And the main rationale is human genetics, in human, if you don't have TREM2 at all, like if you have homozygous loss of function for TREM2, you develop dementia by the age of 40 and it's deterministic every individual that does not have TREM2 develop dementia. If you have one of about half a dozen point mutations in TREM2, that partially decrease the functionality of TREM2, one of the 2 copies, you triple your risk of developing Alzheimer's disease. And there is at least one report of a mutation that leads to overexpression of TREM2. And in this case, the overexpression is protected, you reduce risk of developing from -- of developing Alzheimer's disease. So because of the strong human genetics that loss of function leads to Alzheimer's disease, whereas gain of function is protective. We developed a drug that mimic the gain of function feature that basically activate TREM2. And to activation of TREM2, you basically elicit increased number of the microglia brain immune cells, increased ability of the microglia to migrate to the site of injury, increased ability of the microglia to function in multiple aspects in phagocytosing and misfolded protein in our view, also increase ability to counteract other disease pathology. So we think that -- again, TREM2 is an immune checkpoint for the brain immune system. It's an activating immune checkpoint and our drug really recruit microglia to counteract the disease pathology.

Okay. So with that in mind, can you kind of briefly walk through the existing clinical data you have for AL002 and how we should think about kind of -- its kind of drug properties as what you have so far?

So first, we took the drug to animal models and in animal models of Alzheimer's disease, we were able to show that we are indeed increasing proliferation of microglia, increasing microglia migration to Abeta plaques, increasing ability of microglia to compact plaques and basically shield the plaques from the surrounding neurons and reducing the ability of plaques to injure the surrounding neurons. We also took the drug into animal models of multiple sclerosis where you need myelin clearance in order to regenerate myelin. And there also, we saw that in the presence of our drug that activates TREM2, we facilitate myelin regeneration and really see significant therapeutic benefit in this multiple sclerosis model. So with this animal model data, we took the drug to human. Initially, we did studies in healthy volunteers. And in healthy volunteers, we show a dose-dependent and suturable target engagement. We measure the level of soluble TREM2, soluble TREM2 is the extracellular domain of TREM2, which is clicked for microglia after activation. So we saw a dose-dependent effect on soluble TREM2 by the drug, and we show effect dose-dependent elevation of 3 biomarkers that represent microglia activation. This includes soluble CSF1R, which is an essential receptor for survival of proliferation of microglia, IL-1 [indiscernible], which is a soluble inhibitor of the inflammatory mediator IL-1 and SPP, which is, again, an inflammatory and synaptic binder that's secreted for microglia. So in healthy volunteers, we showed, again, target engagement and activation of microglia. We then went to Phase II in Alzheimer's patients. It is a double-blinded placebo-controlled study. And there, the study is still ongoing. It's blinded. What we reported was some cases of ARIA. ARIA is an imaging abnormality that can be seen with MRI. And what was reported so far is that all the anti-Abeta antibodies that remove Abeta plaques led to ARIA and conversely, all ARIA events were associated with drugs that remove the EBITDA. So although we -- at this point, we don't know for certain because we didn't unblind the PET imaging data, we suspect or we hope that our drug impact Abeta pathology among many other things. So, so far, what we've seen in humans is 4 different biomarkers showing harnessing of microglia as well as ARIA suggesting that the drug has activity and in disease context on hopefully on Abeta.

Okay. So kind of following from that, can you give us a little more detail on the design of the Phase II trial called INVOKE-2 for AL002 in Alzheimer's disease, like size, primary endpoint, that kind of thing, time?

All right. Yes. So the 002 Phase II studies enrolling 265 early Alzheimer's disease patients. It's placebo controlled as Arnon mentioned, and we've got 3 different dose arms. The study is a common close design. So basically, the last patient in will be treated for up to 48 weeks with an 8-week safety follow-up. Anyone that was enrolled before that will be trued up for 96 weeks. So we're really trying to gather a pretty robust set of data. And then at which -- and the primary endpoint there is the CDR sum of boxes, which is what's been used in many of the other Alzheimer's trials. And then as a secondary endpoint, we're looking at a lot of the other cognitive endpoints like the ADAS-Cog and [ R-bands ] and things like that. And then, of course, we're including a number of exploratory biomarkers, both on the imaging side with MRI as well as liquid biomarkers.

And it will be just to -- it's like 3 doses like we are testing low, intermediate and high dose as well as the placebo control. So we are really gearing up to do pivotal Phase III based on this Phase II data.

Okay. So with all of that, you obviously have to think about AL002 within the broader TREM2 landscape. As you're aware, there have been safety signals for various agents in the space. And what do you make of these adverse events that have been observed? And how they can be managed and how you think of them practically in terms of how you kind of run the clinical program for O2?

So there are at least 2 other additional reported programs on TREM2 and one of them is in a complete clinical hold because of safety and one is a partial clinical hold at the high dose. And again, we don't know what was the reason for the complete or partial hold, and it could be that there was some toxicity in the preclinical studies or in human. We sort of -- in non-human primates or in healthy volunteers, we did not see any safety signal that sort of led to any concern by the FDA. I mean the only signal that we saw was in alternate patients, which is the ARIA and sort of I think the consensus in the field now is that ARIA is part of the package of removing Abeta. So it's really an on target signal of efficacy or activity. So ARIA is the only thing that sort of we saw so far. And again, most of the ARIA is nonsymptomatic. It's just an MRI imaging with mild clinical symptoms. We had 3 cases of symptomatic ARIA. All of them were in APOE4 homozygous and we decided voluntarily to remove the APOE4 homozygous, which is 10% to 15% of the Alzheimer's population. And since we did that, we didn't really see any additional symptomatic ARIA, and we are continuing with the trial. So our assessment of the field is that the if there are adverse effect of clinical hold, they are drug related, they're not target related that sort of different drugs, designed differently, recognizing different epitopes and that sort of -- the specific design of the drug is really what may be associated with the clinical holds not the target. We think that the target is very valid and very exciting.

Okay. So...

Quickly add that one of the programs on full hold is preclinical and the other one is in early Phase I. So we're the only program in a Phase II. And we've tested in multiple healthy volunteers in Alzheimer's patients.

Okay. That's fair. So given all of this, again, kind of like the again, gut check question, we get from people is when can we expect data from INVOKE-2?

Sorry, from 002?

Yes.

Yes. So as I mentioned, it's a common close design. So once the last patient enrolls, then is 48 weeks follow up and then an 8-week safety follow-up. So call it about a year after the last patient. And the guidance we've given you, the study is enrolling well, it's ongoing. Given we've removed the E44, there's a small impact of that to operations. So we're guiding that we believe in the first half of next year, we'll finish enrollment.

Okay. Great. So stepping further back, can you walk through some of the remaining clinical programs? And I guess kind of I'm asking you to pick which child you love the most, like, particularly the ones you're most excited about.

So one of the problems that we are most excited about is a new Alzheimer's programs that we are developing. It's a program that targets a regime called MS4A. And this regime is unique because it appear to impact both disease initiation and disease progression. Either directly or indirectly, the protective variant of this target was shown to slow down cognitive decline after you develop Alzheimer's to slow down brain tissue loss, both in the Hippocampus and Cortex to slow down the conversion from mild colitis impairment to Alzheimer's disease, to increase survival with Alzheimer's disease and to increase the age of onset. So we think that, that's really a target that comprehensively impact many aspects of Alzheimer's disease -- and mechanistically, these targets seem to sit above TREM2, one of the things that this target regulates is TREM2 level. So we think that we have identified something which at least I consider a master regulator of the brain immune system that has multiple impacts on microglia survival proliferation, migration, and multiple aspects of functionality, including lysosomal function and mitochondrial function, again, ability to migrate. So it means I think that sort of this is really a very sort of definitive targets to recruit microglia to treat neurodegeneration. And this drug is going to be in the clinic later this year, and we will see how it works. Mean the other drugs that we will be in the clinic either late this year or early next year are 2 oncology drugs, sort of -- we have novel degeneration -- novel science company, but we are modulating the innate immune system in the brain, and we think that the innate immune system in the brain is very similar to the innate immune system in the periphery. So we think that many of our drugs are going to be like CD20 that started as targeted by Rituxan for non-Hodgkin lymphoma in cancer and now is the most successful drug for multiple sclerosis as ocrelizumab. So we think that many of our drugs and targets could traverse between oncology and neurology, and we are testing this hypothesis with 2 programs that will be in the clinic, sort of early next year. So we have, again, 2 or 3 additional programs going to the clinic as well as a broad portfolio of neurodegenerative immune checkpoint drugs that hopefully, will transform brain health.

Okay. Great. So finally, the question we're asking every company at the conference is what is the reason to own -- in this case Alector stock in the next 12 months?

Yes. We are developing a broad therapeutic strategy for dementia and other neurodegeneration, which we call immuno-neurology. We have a broad pipeline, we have multiple immune checkpoints that are genetically validated. And I think that our approach of immuno-neurology will transform brain health very similarly to what immuno-oncology bring with cancer. So I think that if you want to participate in this revolution, it's really -- we are inviting you to join us.

Excellent. Well, thank you so much, Alector team for joining us and thank you all for joining us this afternoon.

Thank you.

Thank you.