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

Great. Well, good afternoon, and welcome to the Barclays Global Healthcare Conference. My name is Carter Gould, senior biopharma analyst here at Barclays. And as we enter the afternoon, we're very pleased to welcome Alector to the stage. Joining us from the company is Co-Founder, CEO, Arnon Rosenthal. Arnon, thank you very much for joining us. I know it's a busy time for the company. You guys had data at AD/PD host -- had a call yesterday on that. I think the presentation was today or -- I think that's right. So I know a busy time for you. Thanks for taking the time out to speak with us today.

Yes. Thank you for the invitation, and thank you, everyone, for joining this [indiscernible].

Perfect. So maybe before we jump into Q&A, I think it would be helpful to give a little background of the company and whatever you think would be helpful to people before we kind of jump into the questions.

Yes. Absolutely. So the premise behind Alector, the reason for its existence is to develop an alternative therapeutic strategy for dementia and other generation. As you know, the dominant approach to treat Alzheimer's and Parkinson's and ALS and Huntington disease was to go after their respective misfolded proteins. So the general idea was that each neurodegenerative disease has a typifying misfolded protein. Alzheimer's is typified by A-beta and TAU. Parkinson's disease is typified by alpha-synuclein. ALS is typified by SOD and TDP-43. Huntingtin is typified by Huntington's disease. And the idea was that if you remove, prevent the synthesis or disaggregate any of these proteins, you develop a cure. And unfortunately, this approach is not panning out as we hoped. And our idea was instead of going after misfolded proteins directly, to recruit the brain immune system to do their work for us, to cure the disease for us. So conceptually, Alector is very similar to immuno-oncology. What revolutionized immuno-oncology was that instead of trying to kill cancer directly, we now recruit T cells to do the work for us. And we are doing the exact same thing. Instead of going after the pathology directly, we recruit the brain-specific immune system to do the work. And the scientific rationale for this approach is really human genetics, the emerging human genetics primarily for Alzheimer's disease in the last few years, showed us that most of the risk for Alzheimer's disease actually resides in the immune system. The majority of the risk for Alzheimer's disease are practically immune checkpoints for the microglia immune system in the brain. So all our brands really are genetically validated. They target risk -- genetic risk genes, and they really modulate these risk genes to act opposite of the direction of the risk. And with this approach, we developed a broad portfolio. We have more than 400 patents covering this immuno-neurology idea with dozens of targets and technology. We have multiple drugs in the clinic, including drugs for frontotemporal dementia, which is the second-largest form of dementia for early age, multiple drugs for Alzheimer's disease; and sort of we are going drug for ALS; and we are planning to go also into Parkinson's disease and other neurodegenerative we're solving.

Perfect. Plenty to jump into, particularly with the data yesterday, but maybe just start with more of a fundamental question initially. Why was FTD such a good place to start when you think about sort of your initial effort?

FTD was sort of -- we started with a genetically sort of defined subset of frontotemporal dementia. These are people that develop frontotemporal dementia because 1 immune regulatory protein called progranulin is deficient. So basically, people that have 1 good and 1 bad copy of the secreted immune regulatory protein, progranulin, invariably develop dementia before the age of 60. And despite of dementia, it's called frontotemporal dementia. So it was conceptually a simple situation where a monogenic disease, 1 gene, secreted gene is mutated, and if you found a way to restore this immune regulatory gene, we thought that we will have a therapy. And that's what we were able to do. Basically, we were able to develop a drug that restore progranulin back to normal level by blocking at the gradation cascade. So the mechanism of our drug is very similar to the SSRI that led to Prozac and other many sort of antidepressants, basically, block the reuptake of serotonin, and it's a very safe and effective way. And we are doing the same thing. We block the reuptake and the gradation of this secreted protein. And by this, we were able to double to triple its level in the brain. And in FTD patients that have a genetic mutation, this -- and able to restore progranulin fully to normal level.

So I want to dig into that because the Phase II that you've been running has given you sort of a step-wise approach to show, first, that you can restore it to normal; and then also move to, I guess, super physiologic levels. And maybe kind of walk through that and the importance -- we're going to -- I'm sure we're just going to straight -- jump straight into the data yesterday, but maybe just walk through the importance of that...

Yes. Absolutely. So yes, so initially, there is progranulin mutation carriers. They're 50% of the normal level, and with our therapeutics, we were able to restore it back to normal. It's, in a way, enzyme-replacement approach. It's a simple concept. But -- and we -- in our Phase II, open-label Phase II, we were able to show that sort of reverse multiple disease biomarkers with this approach that we slow down brain tissue loss as measured by volumetric MRI, and that we slow annual decline of cognitive deficits by 48%. So we are in the midst of registrational Phase III with this indication. But the human genetic of progranulin shows that it's a risk gene for Alzheimer's disease, for Parkinson's disease, for ALS for late, which is another type of dementia. So practically, every neurodegenerative diseases, this even small deficiency in the level of progranulin, the amount of progranulin is a risk. So we wanted to venture beyond this enzyme replacement indication to other indications where we think that even physiological level of progranulin is rate-limiting. And the C9 and other genetic subset of frontotemporal dementia, people that have C9 open reading frame 72 was an opportunity to do that. So these people, again, they develop frontotemporal dementia before the age of 60. It's a very rapidly progressing disease. Human genetic shows that decreasing progranulin exacerbate this disease and that over-expression in animal model is protective. But the people have -- sort of these patients have normal level of progranulin. So we hypothesize that elevating progranulin two- to threefold above the normal level will still be therapeutically beneficial. And the data that presented yesterday sort of are consistent with this hypothesis. Again, we showed significant slowdown in cognitive decline of 54%, which is, we think, very profound. We show a normalization of disease biomarkers. And we think that in combination, the totality of the data from the progranulin mutations that leads to FTD -- from the C9 mutation that leads to FTD suggest that we have really an effective drug and that the drug could be effective in other indications where progranulin is at normal level.

So I want to come back to those aspects. I think one of the things that maybe was lost yesterday is that you were able to do this all safely.

Yes. Absolutely. So the drug is very well-tolerated. I mean these sort of individuals are either healthy volunteers or different types of FTD mutation carriers have been treated with this drug for 2 years now, and there are really no drug-related adverse effect. It's a safe drug, and it's -- it's a safe drug when you store progranulin back to normal level, but more importantly, in healthy volunteers but in C9 new patient care turned out to be a safe drug also when we double to triple the level of progranulin, again, for a long period of time over 2 years now.

So the question I would have asked yesterday, if I didn't have to run off your call and start the conversation here, which would have been, what are the implications of that when we start to think about the ALS study, when we think about the other kind of studies above and beyond sort of the PGRN population?

Yes. So this really supports our hypothesis that in disease states, in neurodegenerative diseases, progranulin physiological level is still rate-limiting. And that if you elevate progranulin beyond physiological level, you will have therapeutic benefit. So this is what we think we achieved with the C9 of mutations that are associated with FTD. And we think that this would read through to the other large diseases, Parkinson's disease and Alzheimer's disease, and we are planning on testing this hypothesis sort of in the next few months.

Okay. So I want to come back to that in a second, but let's talk about some of the biomarker data and go into a little bit more detail there. So when we thought about the original population you evaluated the drug in, we kind of went through this -- serum neurofilament went from being super important to then you guys talked it down a little bit and being sort of less relevant in that population. And then yesterday, you showed very good serum neurofilament data. I guess now, when we think about the C9orf population, has it taken on more relevance? Is it less significant? How much weight should we really put on that data?

I think that neurofilament is still an exploratory biomarker. There was initially a lot of excitement about neurofilament because it tracks disease progression really well with multiple neurodegenerative diseases, AD/PD, multiple sclerosis, Huntington. But when you look at tracking therapeutics, it doesn't track therapeutics very well. Like, for example, the drug that -- the ALS drug that Amylyx is advancing now that probably will get approved chose pretty good clinical efficacy. And actually, the level of neurofilament increased over the 6 months clinical trial period. And there are multiple examples now that the levels of neurofilament really are disconnected from clinical efficacy. So I think that neurofilament is an intracellular protein, the neurons that's secreted when nerves are injured. So if there is more injury, the level is higher. And I think that eventually neurofilament, if you have a good drug, will go down. But I think that now that it will be a trailing biomarkers, not a predictive biomarker of efficacy, I think that it's going to be a very slow biomarker that we will see clinical efficacy and other biomarkers responding significantly faster than neurofilament. And there are mechanistic explanations for this. It could be that it takes -- that neurons start to function before they are completely healing and stopped secreting neurofilament. But more importantly, many of them is folded proteins that we work with, like TDP-43, which is the typifying misfolded protein in FTD and ALS; or huntingtin protein, which is the typifying protein in Huntington's disease, actually suppress transcription of neurofilament. So if you have a misfolded protein that suppress transcription of neurofilament and you have a drug that counteract this misfolded protein, neurofilament will actually increase initially. So you have competing activities. You have elevation of transcription of neurofilament and reduction in secretion of neurofilament from injured neurons, then I think it's going to -- it might take several years for neurofilament to reach an equilibrium starting going down. There is actually 1 example in lysosomal storage diseases with no generative pathology for which there is enzyme replacement therapy. The clinical benefits are immediate, but it took 4 years for neurofilament to go back down to normal. So I think the time frame is going to be much longer than we expected. And as a result, we view it now as a much less important biomarker that would predict efficacy. But still, as you mentioned, like we see stabilization and possibly even a trend or decline. So I think it may take a little bit more time, but it will happen hopefully eventfully.

So one of the key headlines yesterday, as you mentioned, was the impact on the clinical [ trial ] that you measured and pretty profound, north of 50%. At the same time, it was done against the sort of matched historical control, which not everybody is always familiar with and that we had to go through this with the original data set as well. I guess one critique of that was it was drawn from a different database. Can you just maybe just walk through why you had to do that, if that's worrisome or not or just your view?

Yes. So the 2 clinic -- Phase II clinical trials that we did with progranulin mutation carriers and C9 mutation carriers that suffer from FTD were open-label trials. And in order to monitor efficacy, we had to do what we call a synthetic control or what we call digital twins. And the digital twins in both cases were taken from FTD consortia. So there are 2 academic consortia that basically follow FTD patients over time and look at the rate of disease progression, survival, effect of biomarkers, and in some cases, they look at brain tissue loss with MRI. So there are -- there is a European consortium, and there is a U.S. consortium. And they have very similar groups of patients, and they're actually going to integrate soon. So the first group of control patients that we took for the progranulin mutation carriers from the -- from one of the consortia called GENFI and the C9 -- for the C9, the control that we're taking was -- or the other consortia ALLFTD. The type of patients are very similar. And basically, 1 consortia has more patients with more data for 1 disease and less patients in this data for the other disease, and they are just more accessible from 1 consortia for 1 disease and the other consortium for the other disease. There is really no meaning to this outside of that. It's just what was available for us as a control. And I want to emphasize that the control were selected blindly. We basically selected -- there's actually a small number of patients to start with, but we selected patients that have the same starting parameters as our treated populations, like the same cognitive deficits at baseline, the same level of neurofilament, similar age, similar gender, similar diagnosis. And only after we selected the control patients, we look what happened to them a year later. So the selection was blinded before we knew what the disease progression is going to be. So we think that we did have a pretty rigorous control selection, and that the control is really reflecting what will happen in real life to untreated patients. And again, if you compare to our control with the progranulin FTD, we all see an activity as we see more than like around 50% or more slowdown in cognitive decline annually. And we think that that's a pretty profound effect, if it sort of repeat in Phase III.

Right. One of the -- you also were able to show some patient-level data on those clinical measures. And I recognize the ends are very small in some of these. But for a number of those -- a couple of those patients, you were able to get back and improve relative to baseline, which, generally, we don't see in these sort of neurological diseases. Can you just talk about the importance of that? And how commonly when you look through some of these databases, you would see -- is that something that happens commonly with noise? Or just kind of how outstanding and important is it?

Yes. We measure cognitive decline every 3 months. Unfortunately, the consortia measure cognitive decline every year only, so there is less detailed data on cognitive decline, on variations in cognitive decline. But you are right that generally, you don't see improvements sort of cognition in these small variations. But basically, nerves are damaged and destroyed and if they -- they don't repair spontaneously. So although we don't know, it's possible that the improvement may represent damaged neurons that stopped transmitting signal, stopped functioning, but once you remove the insult that they could reconnect, this could heal and basically may represent physiological improvement. So that is a bit intriguing, and we would see what happens in a larger trial. But it's not inconceivable. Basically, a lot of the dementia, I think, is caused by injured neurons that are not -- that once the levels are there, it's reversible. But in general, you can still restore if you stop the injury. In stroke, for example, you have very large paralysis, and over time, the neurons recover, and you see improvements -- in many acute injuries, you see improvements. So with the drug, it's possible that you see improvement in dementia features, if the drug removed the injury and the neurons recover. And again, this is an intriguing part of the data that we will keep monitoring.

And when you think about sort of like your working theory around the kinetics of being able to show that clinical response, there was a number of -- like we were talking before, there was a number of patients that were showing pretty robust declines that we're sitting through 2, 3, 4 months and then had pretty profound improvements. How do you -- at this point, how do you think about the pace of clinical improvement and the kinetics at play?

Yes. I view sort of what we see is like braking distance in a car. So when a drug -- car is accelerated, and that's what I see like the disease progression, when you press the brakes or remove the disease injury, it takes -- in a car, it takes some time for the car to stop. And the same happens with an injury or with disease. Once you removed the insult, it takes some time for the disease to respond to the lost insult. So we think that, basically, the disease will slow down and eventually stop and maybe we'll get better with time. And C9 is a slower progressing that is compared to the progranulin. So it's possible that, basically, it will take shorter time, just maybe 2 to 3 months for the disease to -- basically for the injured neurons to heal and start functioning again, and then you see either slowdown in cognitive decline or even maybe improvement. It's like no different than any other injury. Like if you stick a knife in your wound -- in your hand and then you remove the knife, you don't heal immediately. You just take -- it takes some time to heal and basically get back to normal, and that's what happens with the injured brain. The drug removed the insult, but it takes the brain some time to heal and start functioning again.

Okay. Move away from the nice first...

Yes. Sorry. People are sensitive.

All right. So you do have a Phase III ongoing. Can you talk a little bit about your confidence on the pace of enrollment and that we'll see data in line with the time lines, how you kind of communicated to the Street?

Yes, we are in the midst of a registrational Phase III for FTD with progranulin mutations, and the trial includes both symptomatic mutation carriers and presymptomatic mutation carriers. Basically, we are trying prevention and reversal models in a single trial. And the FDA agreed with us to combine the 2 trials, and the FDA agreed with us on the clinical endpoint that will be a registrational endpoint. And we have close to 50 sites worldwide, and the recruitment goes according to expectations. It's a rare disease. It's the first drug that's ever really tested in Phase III. So we are really offering genetic screening. We are increasing awareness of the disease, and recruitment is what we expected it to be.

And maybe taking a step back, when you think about sort of the agency's interaction with CNS-focused companies, clearly, it's been a colorful 12 months or so. On one hand, you could absolutely point to evidence that maybe they're course-correcting now. At the same time, you see things like ALS where they seem to be as hyper-accommodating as they ever were. From your perspective, what's sort of the state of the state in FDA and how they're approaching neuro?

I don't think -- I think that they took sort of a strategic decision first to rely more on biomarkers, not just from the clinical endpoints, which are very difficult in neurodegeneration. And I think that they decided to really make approval as easy as possible and maybe allow sort of drugs to be approved and then monitor following approval, which I think is a great thing for neurodegenerative drugs. I don't think that they are course-correcting because of recruitment. I think that after recruitment, they gave breakthrough therapy to 3 additional Alzheimer's drugs, so they're not course-correcting. You mentioned the Amylyx had drug that they are considering approving after a Phase II trial. I think that sort of this is the trajectory. I think that neurodegeneration regulatory component is going to be like cancer: it's going to be hopefully more risk-taking and more innovative. And I think that it will help the field tremendously.

Okay. So you have a number of other programs that we spent the first 20 minutes talking about progranulin. Let's talk about TREM2. For example, it's obviously a space that's been in the news of late. You had some competitors get put on clinical hold. You've had to make some changes as well. Can you maybe just walk through kind of where that stands and some of those changes?

Sure. So yes, so we have a drug in Phase II. This is targeting TREM2. TREM2 is immune checkpoint for the brain immune system. It's a very profound risk gene for Alzheimer's disease. Mutations in TREM2 provide as much risk as APOE4 allele, which is the most famous risk gene for Alzheimer's disease. And because of the human genetic, we developed an activating drug for TREM2. We have shown in human that the drug engage target, that the drug recruit microglia with 4 different disease sort of downstream biomarkers. And we are in the midst of Phase II, and again, recruitment is going extremely well. There's a lot of excitement about non-A-beta therapy for Alzheimer's disease. And again, as you mentioned, we are the only company that have a TREM2 drug that's not either on parcel or complete clinical hold. And I just want to mention that our progranulin elevating franchise is in partnership with GSK, with -- we just did the partnership last year. We received $700 million upfront, and it's a 50-50 cost profit sharing with commercial rights. And that we have 2 programs in partnership with AbbVie. One of them is TREM2. Again, it was -- they were partnered at preclinical stage. We got $220 million upfront, and it's again worldwide profit cost sharing. So we have a very ambitious program beyond TREM2 and progranulin. We are taking 3 additional drugs to the clinic this year, and we have sort of $930 million in cash that allow us to really execute our progress.

Those were timed extremely well, given sort of the market today. So -- but we're looking forward to all the updates we will get on the progranulin front over the coming couple of years. So Arnon, thanks very much for joining us today, and we had a great conference.

Thank you for hosting me.