Scientist Stories: Ajay Verma, Vaccines to Treat Neurodegenerative Disease

Turning now to another very significant disease, a neurodegenerative disease, our next speaker is Dr. Ajay Verma, Chief Medical Officer at United Neuroscience. And he is a neurologist who's devoted his career to the advancement of neurotherapeutics and neurodiagnostics. He received his MD and his PhD in neurotoxicology at Johns Hopkins and did his residency training at Walter Reed and was on the staff at the Uniformed Services University of the Health Sciences, the U.S. military medical school, for many years before moving on. He served in leadership roles at Merck, Novartis, and Biogen, where he directed drug development, biomarkers, and experimental medicine efforts in Alzheimer's disease and Parkinson's disease, ALS, neuropathic pain, multiple sclerosis, and a significant number of other brain and neuromuscular disorders. So thank you very much.

Thank you. It's a pleasure to be here. It's a very important topic I'm going to be talking about. I think our generation really faces a major crisis, actually. This is— we may in our lifetimes witness one of the biggest calamities that our species has ever faced, which is the rise of dementia. The rise of dementia is really exploding at an exponential rate. If you're over 65 right now, you have about a 1 in 8 to 1 in 10 chance of getting Alzheimer's. If you're over 85, it's about 1 in 3. And there's no disease modification therapy on the horizon. So this is, the better we do in cancer and cardiovascular, we actually promote this problem in a way. And I wanted to highlight that a little bit. For most of our species, we've only been a few hundred million people on the planet. And it's just this last 200 years that we've spiked up to 7 billion. And of course, this is due to a lot of stuff that we've done with sanitation, food availability, antibiotics. But I think a lot of this is also vaccines. And I'll come back to this in a minute. It's not just that there's more people, we're living longer. We're seeing diseases of the elderly that we never saw before. Probably in caveman times, they never saw Alzheimer's, or in the Roman times because people didn't live that long. Because these diseases take a long time to develop and they really develop late in life. If we think about it from a vaccination perspective, I actually think of this as immunoengineering. I think we're actually actively maintaining this peak. By immunoengineering people in their first decade of life. I would put it to you that we're actually doubling people's life spans early on by protecting them from the environment. If that went away, I bet this number would drop pretty quickly. So the thing that's always puzzled me is why stop there? We've actually immunoengineered humans to adapt to modern times. Now that we're living longer, what can we do for the later years? Can we also protect, you know, seniors with vaccinations? Perhaps not against just exogenous pathogens, but against some of the endogenous pathogens. And what I'm specifically talking about is some neurodegenerative neoantigens, epitopes that our body hasn't seen that creep up later in life, particularly at a time when our immune system is weakening. And what are these? These are actually proteins that we've had our whole life, that actually misfold and tend to aggregate and are meant to be cleared all the time. But with aging, particularly in the nervous system, we have old cells that don't replicate, that live for 90 years, you know. Their upkeep mechanisms get gummed up, for lack of a better, more sophisticated term. And we end up having various beta-pleated, you know, sheets of a variety of proteins that do normal functions in life, but later in life when they aggregate, they present epitopes that are really neoantigens. And we know these as beta amyloid plaques, tau tangles, alpha-synuclein Lewy bodies, you know, other protein aggregates. But what's really happening when they aggregate is they're expressing epitopes that weren't there previously. And once they form these little new surfaces, they actually create a new molecular interactome. And they start to actually bind to each other and gum up and so on. And there's different strains of these misfolded proteins that are now being appreciated, much like prions, that depending on the way, you know, what's expressed on their surface, they have different behaviors in what they interact with. Many of these can be spit out from cells, especially when they're injured, and be taken up by other cells and moved along circuits. And where they end up, you know, hiding out in the nervous system basically represents the clinical phenotype. You know, if they gum up various motor systems, you present with a motor phenotype. If they're more of the thinking areas, you get a cognitive decline. And there's often a mixture of these in neurodegeneration, so it's not usually one protein. So this is one way of sort of looking at it. And you might then question, well, if you're forming these neoantigens, you know, shouldn't your immune system see them as being neoantigens because they haven't been there your whole life and do something about them and maybe get rid of them? And it may be that it does, actually, because even though I said, you know, how bad I painted the picture about Alzheimer's disease and dementia, most people actually don't get dementia. It's still a relative minority. And when we actually look carefully, there's actually many normal people that are elderly that are walking around with endogenous antibodies against misfolded A-beta, tau, alpha-synuclein. We've coined a little term internally in our company to call these endobodies, meaning just that endogenously— presumably these are being generated by endogenously being vaccinated. You know, someone just Their immune system is seeing these things and they're potentially protecting themselves. We haven't really proved that yet. This is sort of an association. But when I was working at Biogen, we were actually working on aducanumab, which is a monoclonal antibody that targets aggregated A-beta. And we actually licensed this from a Swiss company called Neuroimmune. And they had actually fished this out of just old people. They were studying really old people in their late 90s, early 100s that were 2 standard deviations better physically and mentally than their peers. Lo and behold, a lot of them just walking around with lots of these antibodies. Aducanumab was actually cloned out of such a person. Through a long process, basically, we had to murinize it to test it in mice, but it's already fully human. Once we did that, it went right into clinical trials. It's really a monoclonal antibody targeting A-beta that's invented by God, but selectively amplified by these companies. And now it's in clinical trials and it's the leading candidate actually for disease modification therapy in Alzheimer's. It's the first one that's shown that you can actually not only halt amyloid plaque progression but actually reduce it and maybe even stabilize cognitive benefit. We took it through the Phase 1B and then I left Biogen and now it's in Phase 3. And a lot of promise for this. So the company that I joined, United Neuroscience, Our thinking is, well, if that can happen endogenously, surely we can direct the immune system to actually recognize these neoantigens and fight them off through vaccination. And I don't need to review all the benefits of a potential vaccine versus a monoclonal antibody to this crowd, except to point out that this is not easy to do, right? People have tried this before. It's not easy to control the immune system Particularly when you're targeting it at an endogenous protein. There's a risk of unleashing it and getting T cytotoxic responses and autoimmune reactions. The challenge is really, when you vaccinate, how do you do this so that you can selectively get antibodies produced without activating cytotoxic T cells? We know that in order to do that, you do need some T helper cell help here. How do you do that? That's what our company has been working on. And just to walk you through, you know, how we got to that stage, this was tried before. This is not a new idea. So Elan Therapeutics tried this about 20 years ago. They used the full A-beta peptide. Here it is, 1 to 42. And what this peptide is, it's actually a little middle piece of this larger protein called amyloid precursor protein. So it's clipped out, and it's about 42 amino acids. And 15 through 42 are the sticky portions of the— I mean, of this peptide. So they tend to glom together. And the 1 to 14 tends to stick out in the dimers and oligomers and fibrils and so on. And that's been shown recently that that portion, you know, sticks out. However, they had used the entire 1 to 42 peptide. And when they went into mice with it, it worked in both a preventative and a treatment mode. So if you started immunizing animals early or even after they had plaque, you could actually reduce the number of amyloid plaques in their brains. And when they took it into man, however, you know, they ran several studies and a lot of people got dosed with it, but the study had— the studies had to be stopped early on because they had about 6 to 8% meningoencephalitis, this autoimmune reaction, largely T-cell mediated against, you know, brain and meningeal tissue. And so unfortunately, you know, this People had a lot of promise for this, but it had to be stopped. Even though they stopped dosing, many of the subjects that had been dosed that didn't have encephalitis continued to be monitored. A lot of them did well. There were several papers published about this. In fact, when some of them went to autopsy and their brains were looked at, they actually had a— compared to control patients where you might expect a lot of amyloid, which is shown in the black here, some had prominent clearing, others had big patches of clearing in their brain, which is not typically what's seen in pathology. So the idea was that this could work if we got rid of that safety issue and then really tried to turn it into a drug, you know, in some way. And so it was discovered later, of course, that this peptide actually has, you know, specific B and T cell sites as the more the immunoinformatics world came along. And so really, from starting around 15 is where all these T cell epitopes are, and they had used the entire, you know, peptide. So what then others and ourselves did was really focus on this N-terminal end that just had the B-cell epitopes. And, you know, we tried to see, okay, can we make an antibody— can we make a vaccine that just has that epitope? But obviously that's just a B-cell epitope, so you got to link it to some T helper, you know, thing. And then a lot of other companies used ugly things like KLH or tetanus toxoid or something. Well, our parent company had developed this library of T helper peptides. And I'll talk a little bit more about these. So what we basically did was swap out the rest of that molecule and put in our T helper peptide. These peptides are derived from various pathogens like measles and hepatitis and so on. We have a whole collection of these. And our vaccine technology basically uses these and couples them. So here are T helper peptides. And through a link— these are about 20 mers, so it's pretty small. These are all synthetic peptides. They do have, you know, some analogy in nature, but they're really actually modified artificial sequences from various viruses and so on. We have a proprietary linker peptide, and then we put our B-cell epitopes here. So we're very good at sort of, you know, finding the right B-cell epitopes that are going to be exposed on the surface of some protein, empirically making lots of these vaccine constructs with lots of B-cell epitopes and these, testing them all empirically, seeing which one gives us the right antibodies. And our vaccine also has CpG, not only as a TLR9 agonist as an adjuvant, but we actually engineer in various charges into the peptide so that it complexes with the negatively charged CpG in a certain way to hold the epitope just right to present it a certain way. Now we try to predict this, but obviously we have to do a lot of iterations and just empirically test various combinations of this to see what works the best. And basically then we have Adjuvax. So it's a very clean system. But what it does allow us to do is put in multiple antigens. You could target more than one protein, two different epitopes on the same protein, a phospho and a non-phospho epitope. So you can put them all in the same vaccine. We also have a number of these T helper peptides. We have about 3 dozen of these. And some of them may have different MHC, you know, restrictions. So we can actually potentially customize them for different ethnic groups. Something we're working on. But overall, it's a pretty simple, you know, as far as a drug goes in terms of vaccine, pretty simple construct. This just illustrates what these peptides do. So here's just the B-cell epitope itself. If you give that to guinea pigs, you don't get that much of an immune response. This is log 10 of just antibody titers. Whereas if you couple it to something like KLH, of course you get— well, most people don't show you is that most of your antibody response actually goes to the carrier here, right? They'll just show you this. So in our case, we're lucky in that we get a really nice response to the B-cell antigen, but our T-cell epitopes are silent. And we've seen this consistently with various targets. Moreover, we have a whole bunch of these different ones. And we can actually— even though— so here's A-beta 1-42, which has T-cell epitopes, so it's going to give you a response. If you take those out, just do the 1 to 14, you don't. But we have these different epitopes that can actually shape the magnitude and, you know, the amplitude and the speed at which you get to a certain titer. So we have the ability to shape the immune response as well with these T helper peptides. Okay. When we actually applied this to A-beta, then we took it into various animals and we get very predictable responses across species in guinea pigs, monkeys, baboons. And in monkeys, we can actually dose— we can lay off, come back and boost, and boom, you go right back up again. These were tested in animal models of Alzheimer's disease and shown to lower amyloid plaque, just like all the other monoclonal antibodies did, which is what's shown down in these diagrams. And then we took these into man to see if this platform would work in man. The platform was actually worked out in animal health and has been very successful in veterinary applications. But this was the first application in man. And so we went right into patients. These were mild to moderate Alzheimer's subjects. That's what these red numbers are supposed to indicate. And we, you know, wanted to just measure the safety and immunogenicity of it. We looked at a variety of things as far as the humoral response. Some of these aren't complete yet, but we're finishing these, as well as the cellular responses. And we We're delighted to find that actually this was with 19 subjects, all mild to moderate Alzheimer's subjects, and we found that everybody responded actually. And we, you know, with these 3 doses, 0, 1 month, and 3 months, we start to see a titrated antibody response that peaks around 4 months or so. If we don't dose anymore, this comes down. We think this is largely because we're primarily expanding B-cell clones and not activating a lot of T-cells. We have to really confirm all that.. But we do see the response come back down, which can then be boosted right back up again. In this Phase 1, we also saw, you know, nice response across the age span. We didn't really see any drop-off of immunosenescence. We didn't have people under 50, but at least, you know, we were delighted with this response. And when we actually characterized the antibodies, what do they actually recognize? Just as we predicted and programmed in by our V-cell epitopes, This is from a human subject over different weeks after inoculation. And these are dot blots. They're— we're very good at making these proteins. So this is the alpha helix monomer versus the beta sheet, you know, various size oligomers and fibrils. And you can see that the subjects actually develop over time a response just to the aggregated forms. And moreover, their sera can be used to stain Alzheimer's pathology and it recognizes the amyloid that's in the plaques and the perivascular spaces very nicely. We've done a very— we do very thorough peptide maps to see what are these antibodies actually reacting against. So this is the APP molecule that I was showing you. These are various 10-mers that are walking down it, and right where our lead peptide is, that key aspartate is necessary, that's where most of the patients, you know, show their response. You don't see a smear. So it's for a vaccine, we're very happy with this. It's a very precise response. And we tend to see this in animals and humans as well. We've done a bit to assure ourselves we're not activating T cells. We need to do more here. But we have tried to take peripheral blood cells and stimulate them with the antigen. And we don't see evidence of T cell inflammation. And it's been very safe so far. Phase 1 was very well tolerated, just some local site reactions. We did a lot of imaging, and there was, you know, a hint— this was a very small study, it was open label— but there was a hint that if the higher the titers, maybe you'd progressed a little slow. But the numbers were so small. It did look better in the earlier the disease was, so the subjects with an MMSE greater than 20, which means earlier disease, tended to do better during the study. But all this really informed us is to go earlier in our next study. And so we did that. And this is our Phase 2A study, which is currently running and will be finishing up at the end of August. This is 43 subjects, has a placebo arm in it, 2 different dosing schedules where we're doing the same priming that I showed you, but we're boosting every 3 months or every 6 months. And this one actually, having come from Biogen, you know, we did a lot of work on really quantifying the lesion to see how we do Alzheimer's drug development. So these were subjects that were screened for making sure they had amyloidopathy with an amyloid PET scan. And I'll show you the layout of the study. So here's the study. These are those first 3 boosting doses. And in this arm, we're boosting every 3 months. Here it's every 6 months, and then the placebo. Placebo arm. And because of the prior experience, and also what people are seeing with monoclonals in the field, there is early on some brain edema and the, you know, the meningencephalitis that happened in the past. So we were very careful to do lots of serial MRI exams in people to make sure they didn't have any inflammation or edema. And so we did a lot of that. We had a PET scan at the beginning, at a year, year and a half. We also have a number of cognitive scales and functional outcomes, as well as some exploratory MRI markers in there. We're blinded to the study currently. We're just starting to get the data together in a blinded way for QC purposes. But what we're delighted to see is that we're not seeing any evidence of brain inflammation or edema. This is a problem that other immunotherapies have seen in this field in particular. Antibodies against amyloid, for example, here's aducanumab, for example. They see this phenomenon called ARIA-E, which is a brain edema that's treatment-induced. It happens very early. It's dose-dependent. They see quite a lot of it, particularly in people that are ApoE4 carriers. A prior antibody also saw something like this. We're lucky that we're not seeing any of this. And we don't know exactly why this is. We think it's due to the titrated response. When you start a monoclonal therapy, obviously you go up right away and then over and over again with each dose.. And that's where they tend to see ARIA-Es in this early period. And it's dose-dependent. If they titrate, they reduce the rate of it quite a bit. Well, vaccines are so titrated in their response anyway that we're just not seeing that. So we're— now you could say we're not seeing it because the antibodies aren't getting in the brain and they're not reducing plaque and so on. But so we've been looking carefully at that. And again, we're blinded. We don't know which arm, you know, various subjects are in. We're going to unblind pretty soon. But we are seeing a lot of subjects that show no evidence of any brain problems, but they are showing a decline in brain amyloid. This could be placebo, I doubt it, that's not what I saw at Biogen. But we were still blinded, so just a caveat on that. However, in some subjects, we're not seeing this in every subject, we're still blinded. But what Biogen reported over a year, they showed a dose-dependent decline in amyloid in the brain at 6 months and 1 year. And this was really what launched them into phase 3. And in some subjects we're seeing something similar with the vaccine, with IM dosing, without the— so we'll see. I think we have a lot to learn about this program. We have a lot of room to go on the dosing, on the adjuvants. But we've been creeping because we wanted to be safe first and foremost. And this is where we are with that program. Oops, I'm sorry. That's our lead program up here. We're really learning a lot from that, how it works in humans, what are the right biomarkers. What we've done is quickly proliferated that into a big pipeline now. We should have our IND for our Alpha-syn program, which is the aggregated Alpha-syn, by December. We should start dosing in April of next year. We have aggregated tau vaccine, it's about 6 months behind it. Then as soon as we get that lead, we'll start our triple combo vaccine with all 3. There's a few other targets. But we've also proliferated a number of humoral targets. So here, we hope we have a little bit lower bar because the antibodies don't have to get into the nervous system across the blood-brain barrier. And, you know, they might be able to neutralize— we've got animal evidence that this works. They can neutralize some of these humoral mediators. And so these are in development for neuroinflammation, neuropathic pain, migraine vaccine, anti-CGRP.. And we have a few other things that are cooking along. Just wanted to show you, you know, with the A beta, the world knew a lot about that N-terminal sequence that I was showing you, the 1 to 14. With the larger proteins, you know, we don't know what the right epitope is yet. So, you know, instead of just following the competition, we're actually empirically just chopping up the whole protein and making lots of vaccines across the whole thing. So we'll make like 100, you know, different ones. Put them all into guinea pigs. We have this elaborate screening procedure where we'll take them through, you know, various in vitro and in vivo assays to make sure we have the right type of epitope and then do an efficacy study in an animal model and then have a clinical lead that goes forward. So that's kind of our vaccinology meets, you know, biotech drug development. And what we're— this is back to that diagram I was showing you that we're really finding that these strain-based approaches that these different prion-like, you know, species that are formed by these misfolded proteins are very relevant. Some of them are very relevant to disease, others aren't. And so we've actually hooked up with some good collaborators to use this as a screening approach, actually find those titers that are, you know, going after the disease-relevant strains of these. And it's working really well. And I just want to sort of finish up by showing you that, you know, this is very specific. Vaccinate guinea pigs, or hopefully people soon, against A-beta, we just get A-beta, you know, aggregated antibodies. These are all spotted on the membranes here, soluble versus the aggregated forms. If we vaccinate them against alpha-syn, we just get the alpha-syn and not the others. Tau, blah, blah, blah. If we do all three, you get all three. So really excited about our combo vaccine approaches. This is very hard to do for other companies, for other drug types. We have to match, you know, different doses and the this and that. So we're counting on the body actually being able to do combination therapy if we give it combination vaccines. And so this is where we are. And our hope is really to do for neurodegenerative diseases what vaccines have done for other disorders. And I think the power of vaccines to prevent disease and also to do combination therapy— we give our kids MMR and all these these combination vaccines. This is potentially the best way we're actually going to attack these diseases because we're realizing now it's not just one problem and we probably need to hit many different ones. So this is in a nutshell what we're doing as a company. And I'm glad I could share some of this with you.

Thanks. Thank you. Very exciting work. And we could certainly use some good news in this area. But it's also, for me, very positive to see how you have incorporated learnings from past failures as you have shaped your agenda and your research going forward. Clearly you've generated some interest. Lori? Thanks for that. In light of Joel Dudley's big, big paper in Neuron last week, humongous breakthrough, are you looking at some way to together with HHV-6, HHV-7 as your antigens and perhaps piggybacking on that as immunotherapies?

That's a great, great comment. And yes, we're looking at that. I think that that whole field has actually come and gone before. And it's a big impact paper. But people— I remember back in the '70s and '80s even, they were thinking of hepatitis virus. And a lot of it had to do with the anatomy. They thought because where it comes into the nose and the circuits that it travels in, where you see the pathology of Alzheimer's, that it may be a viral type of disease. There is a longstanding story that's the real biology of some of these amyloidogenic proteins may be to be like a really primitive immune system and wall off various infections and form plaques around them. That's another emerging area. So it could be that they go hand in hand, like if you have certain viruses and infections in the brain, promote amyloidogenesis to wall them off. This is another story that's brewing with Rudy Tanzi at MGH. But how it all works together is not clear, but I think it's going to be multifactorial. So, and I do think the vaccine approach offers the best way to go after a multifactorial problem.

A couple of other questions.

Yes. Yes, so for the combo vaccine, I'm wondering what the rationale for that is. Does Parkinson's predispose to Alzheimer's, or is there any connection between these? Great question. As a clinician, I can tell you, and most neurologists will tell you, you often don't know what somebody has when you first see them. And it takes about 2, 3 years to really declare whether it's, what kind of dementia, or it's, and often it's a mix. People have both Parkinsonism and Alzheimer's. It's not as clean as in the textbook. And so the vision we have is, you know, if we can show these are safe, if we can make a dent in early disease, and new biomarkers and so on, then why not have an anti-dementia vaccine that protects you against multiple pathologies and everybody when they're 45 gets this? So that's the idea. It's really, and unfortunately by the time you actually diagnose someone with some of these subsets, it's too late to actually start therapy. So that's been the driver. But there is data that in Parkinson's not so much, but in Alzheimer's absolutely, you have alpha-syn and tau and A-beta. And so on. You know, as a neurologist, I can definitely say I don't know. So that's why we have great advisors like Wade. But absolutely, we are looking at that. I don't know which way it'll go, but we're just starting to look at the subtypes and setting up all those assays. We have a lot of samples left over from our trial. We've got brand new ones coming. In fact, one of my goals in coming to this meeting was to meet others to collaborate with.

Yeah.

Great.

Well, thank you so much.