From Our Neurons to Yours

Why new Alzheimer's drugs don't work | Mike Greicius, Stanford University School of Medicine

• Wu Tsai Neurosciences Institute at Stanford University, Nicholas Weiler, Mike Greicius • Season 6 • Episode 1

Share episode

In the past few years, Big Pharma has released not one, but three new treatments for Alzheimer’s disease.

Aducanemab (2021), Lecanemab (2023), and Donanemab (2024), are the first treatments to effectively clear the brain of amyloid plaques — the sticky protein clumps whose build-up in the brain has defined the disease for decades. The problem? They may not help patients at all.

Today’s guest, Stanford neurologist Mike Greicius, considers the new amyloid-clearing drugs a major disappointment — and worse, says they likely do more harm than good for patients.

Despite this critique, Greicius, thinks that the next few years will be an exciting time for novel Alzheimer’s therapies, as growing biological understanding of Alzheimer’s risk and resilience bear fruit with promising new approaches to treatment.

Learn More:

Greicius is the Iqbal Farrukh and Asad Jamal Professor of Neurology and Neurological Sciences at Stanford Medicine, and a member of the Knight Initiative for Brain Resilience and Alzheimer's Disease Research Center at Stanford University.

Amyloid Drug Skepticism:

Alzheimer's Genetics Research:

Episode Credits

This episode was produced by Michael Osborne, with production assistance by Morgan Honaker, and hosted by Nicholas Weiler. Art by Aimee Garza.

Send us a text!

Thanks for listening! If you're enjoying our show, please take a moment to give us a review on your podcast app of choice and share this episode with your friends. That's how we grow as a show and bring the stories of the frontiers of neuroscience to a wider audience.

Learn more about the Wu Tsai Neurosciences Institute at Stanford and follow us on Twitter, Facebook, and LinkedIn.

Nicholas Weiler:

Hey, everyone. Nicholas Weiler here. Before we get into this week's episode, we have a favor to ask. We're working to make this show even better and we want to hear from you. We're in the process of gathering listener input and feedback. If you'd be willing to help out, send us a short note and we'll be in touch. As always, we're at NeuronsPodcast@Stanford.edu, and you can find that address in the show notes. Thanks so much. Now, let's get to today's episode. 

This is From Our Neurons to Yours, a podcast from the Wu Tsai Neurosciences Institute and Knight Initiative for Brain Resilience at Stanford University, bringing you to the frontiers of brain science.

In the past few years, big pharma has released not one but three new treatments for Alzheimer's disease. These drugs are not just aimed at softening dementia symptoms, like some that have been on the market before. The new drugs, including Aducanemab, Lecanemab and Donanemab, which are a dagnabbed mouthful, are the first treatments that are designed to actually clear the brain of amyloid plaques.

The sticky protein clumps whose buildup in the brain has defined the disease for decades. And these new drugs do a great job of clearing out those plaques. The problem, they may not help patients at all. Today's guest, Stanford neurologist Mike Greicius, has been vocal in plainly stating that he considers the new amyloid-clearing drugs a major disappointment.

And worse, that they likely do more harm than good for patients. We asked Dr. Greicius to come on the show to explain his concerns about these new drugs. Now, despite his criticisms, Greicius actually thinks that the next few years will be a tremendously promising time for novel Alzheimer's therapies. New clinical trials are targeting other forms of Alzheimer's pathology, such as the protein tau.

And growing genetic understanding of Alzheimer's risk and resilience centered on the well-known APOE gene family, are also suggesting promising, new approaches to treatment. But before we got into any of that, I started by asking Dr. Greicius to explain why he wouldn't prescribe any of the new amyloid-clearing drugs to his own patients. Mike Greicius, welcome to the show.

Mike Greicius:

My pleasure. Thanks for having me, Nick.

Nicholas Weiler:

So you see Alzheimer's patients. Do you prescribe these new amyloid-clearing drugs?

Mike Greicius:

Yeah. I do see Alzheimer's patients, and no, I'm not using these drugs. I'm not convinced that they're effective.

Nicholas Weiler:

Can you tell us why not? What is your concern about the effectiveness of these drugs?

Mike Greicius:

Yeah. So it's a bit of an involved tale and I'd like to break it down into two pieces. The first piece is if you believe the results at face value, the size of the effect over 18 months is really quite minimal.

Even pharmaceutical companies have looked at how much change you need to see or difference between placebo and active treatment to say, "Okay, this is really doing something."

Nicholas Weiler:

And when you say doing something, do you mean the amyloid clearance or improvements in memory and cognition?

Mike Greicius:

Well, no, improvements clinically. There's no question in my mind, but that these anti-amyloid antibodies are very good at removing amyloid plaque from the brain. I don't think anybody would contend that. The question, the one we care more about, is how are patients doing as the amyloid plaque is getting clear?

Are they doing better than the people on placebo? It's the clinical outcome I think that we really care about. So these studies have suggested that on the CDR, some of the boxes, this is a clinical outcome measure that we use for studies. A physician needs to see a change of about one to two points over the course of one year in a patient in order to say, "Hey, this patient looks different today than they did a year ago."

So one to two points at one year allows a physician to say, "Hey, the patient looks worse." What we've seen in these trials is something between about a quarter of a point and two-thirds of a point over 18 months between active treatment and placebo. So this is a minuscule change that even over one year's time, would be hard to detect.

And over 18 months, I think, is essentially going to be undetectable to patients, to their family members and to physicians. So that's the first piece. If you believe the results at face value, they're pretty underwhelming. And then the second piece is I don't really believe the results at face value, which is a whole nother problem.

Nicholas Weiler:

So you don't think that even the tiny improvement that is being reported reflects something real?

Mike Greicius:

Exactly. I do think it reflects some changes, but I don't think it's due to the biology of the treatment. The big problem with these results as they're reported, is that there's an issue called functional unblinding, which I can unpack a little bit. It's fairly straightforward, I think.

But the bedrock principle of clinical trial science is that trials should be double-blinded, placebo-controlled. And double-blinded means neither the patient or their family members in this case, nor the treating physician knows whether the patient's on active treatment or placebo. That's the notion of a double-blind.

And with these trials, there's a very common side effect called ARIA that I'm sure we'll get into, which I think is very likely to be unblinding patients and their study partners. Study partners are particularly important in Alzheimer's trials because a lot of the questionnaires have some subjective questions.

How do you think the patient's memory is over the last month or two, for example? And they'll ask that of the study partner, so three parties need to be blinded, the patient, the study partner, and the treating physician.

Nicholas Weiler:

So ARIA is this inflammation and brain bleeding. And that's something that if you get that, you're probably on the active drug.

Mike Greicius:

Right, exactly. That's the concern we've known about, so ARIA is this incredible public relations acronym. It sounds really lovely like an opera. It actually stands for amyloid-related imaging abnormalities. The abnormalities that we see, you mentioned one of them is brain swelling, and the other is brain bleeding. These can be quite dangerous.

And in fact, there have been a handful of deaths both in the Lecanemab trial and in the Donanemab trial related to ARIA. We're not 100% sure why this happens, but the general notion is that the amyloid plaques that we see in the brain's substance itself of the brain, is what we want to get rid of. But there's also amyloid plaque in the blood vessels that lead into the brain.

So we think that when these antibodies see the amyloid plaque in the blood vessels, that causes an inflammatory response that can lead to leakage of these vessels, so either swelling or bleeding. And these, when patients are monitored very closely like in the setting of a clinical trial, these are manageable. We hold the dosing, we reimage, we wait and see, and typically it resolves on its own.

But even in the setting of the trial, there were some deaths related to this, so it's a very serious side effect, which is its own problem. But here it's also quite likely, I think, to be unblinding the patient and their study partner because we don't really see much of this in the placebo group.

Nicholas Weiler:

So the issue here is if people are getting unblinded, either the patients or their family members, their study partners.

There may be a placebo component coming in here where people know they're getting the drug and so expect to get better. And then that's going to maybe account for some of these very small, reported improvements.

Mike Greicius:

That's exactly right, yeah. You might think that the placebo effect doesn't have much of a role in something like Alzheimer's disease where the patient might already have some memory trouble. But these are early Alzheimer's patients, many with mild cognitive impairment. They're not even demented, some of these patients.

And remember here the placebo effect can be both on the patient but also on the study partner, who's certainly got the cognitive wherewithal to have a placebo effect, if they think, "My spouse is on the real deal, I'm hopeful. I think they're doing a little better." There's a tendency to overreport how the patient might be doing.

Nicholas Weiler:

Right. You're looking for good signs.

Mike Greicius:

Of course, that's why you're in the trial. That's true of the patient, the study partner, it's true of the physician. Everybody wants to see these drugs work, so I think that there is potential for a pretty substantial placebo effect.

As you pointed out, you wouldn't need a lot, given how small the difference is between active treatment and placebo to begin with. If even half of that were due to functional unblinding, I don't think these would be significant results anymore.

So I'm quite concerned that this is playing at least some role, if not the main role, in these reported clinical outcomes.

Nicholas Weiler:

So I could imagine people saying, "Well, maybe the results are small, but there are no other available treatments for Alzheimer's disease so what's the harm here?"

And I think we've previewed that a little bit with our conversation about ARIA. So yeah, is that the main thing that you're concerned about, that these side effects are bigger than the potential benefits?

Mike Greicius:

Yeah, yeah. And just to back up a little bit, we have other approved treatments for Alzheimer's that are by no means disease modifying, which has been the pitch for these anti-amyloid antibodies. We have treatments that help a little bit with symptoms and that are quite safe and that we use a lot, but they definitely don't slow the disease down.

So if I believed these clinical outcomes, I might be willing to roll the dice. There aren't lots of deaths, but two at least in Lecanemab, three at least in Donanemab in the trials, in the setting of the trials. I think we're going to see more as this gets picked up in the community use, we'll see more deaths. It's harder to link them once the trials are over to the medicine.

The reporting is voluntary. And even short of death, patients can get symptomatic from ARIA, they can have seizures. It's a concern. I think when you couple that risk with a few other issues, one is there's a fair commitment of time. So Lecanemab is dosed twice a month. You're at the infusion center for several hours. There's a significant out-of-pocket cost. So Medicare does cover Lecanemab.

It will probably cover Donanemab, but that leaves people that don't have supplementary insurance on top of their Medicare with a 20% out-of-pocket copay. So currently, Lecanemab is priced at $26,000 per year. That'd be 5,200 bucks out-of-pocket for patients. And similar pricing, a little higher for Donanemab, would be about $6,400 out-of-pocket. So there's just not much, in my view, that points me to want to use it.

Nicholas Weiler:

Yeah. Well, I can imagine that some listeners may have family members with Alzheimer's and they might have been hopeful about these new drugs. So it's maybe a little disappointing to hear they don't seem to do much.

They've got some side effects. They're going to cost a lot of money and time. What's the point? Why have these missed the boat and what do you think is a better way forward?

Mike Greicius:

Let me address the deflation factor because I feel it strongly. I don't like to be in this role. I don't want to be like a therapeutic nihilist. I'm not, I don't think. I think I'm a realist. I think it might be helpful for some listeners to hear that I'm not alone in this.

There are other neurologists who really don't think that these medicines work. And I think for some family members, it might be helpful to hear somebody, who has looked at this pretty closely say, "You don't have to bother with this." People feel compelled, they feel guilty, they want to try anything.

Nicholas Weiler:

Right.

Mike Greicius:

Really, you're not missing anything if you don't take this. You're probably better off not taking it. It's really kind of agonizing at some level. We, the field, develop these mouse models of Alzheimer's disease, they're very good at generating amyloid plaques. They're actually not good for much else, frankly.

There's not a whole lot of neurologic deterioration or behavioral issues with these mice, despite the fact that we throw one or two nasty Alzheimer's mutations at them. But they're very good at generating plaques, and that allowed the pharmaceutical industry to generate drugs that could remove the plaques.

I feel a little complicit in all this because we, as a field, gave the pharmaceutical industry a target. We said, "It's the plaques, get rid of the plaques." And lo and behold, they delivered with medicines that actually are quite good at removing amyloid plaque, and yet, our patients aren't doing really any better.

So by no means, I feel like we can't get through this without the pharmaceutical industry, so I'm not an anti-pharma person. I think we gave them the wrong target. But there is no question that somewhere early in the disease course, some misprocessing of the amyloid protein gets the ball rolling. I think that's undeniable from the genetic studies.

Nicholas Weiler:

So just to jump in for a second, you were saying that you think that the amyloid plaques are probably the wrong target.

Does it seem like amyloid is the wrong target or are we not targeting it early enough?

Mike Greicius:

Right. Yeah, so that's tough. I would argue that it's more that the plaques are not the right target. So this goes back before we had treatments, before we had the ability to image plaques in life, which we can do now with a PET scan. There are lots of really beautiful, elegant, painstaking studies where they would do autopsies on patients, see how much amyloid plaque and tau tangles that they had.

And then go back and ask, "Did these people have memory trouble? Did they have a dementia retrospectively?" It's been pretty clear from the get-go, once we started looking at these studies, that the places in the brain where you see amyloid plaques are not necessarily the parts of the brain that are sick.

So for example, there's a big part of the brain called the medial prefrontal cortex kind of right behind your forehead. That's one of the first places to lay down amyloid plaques, even in healthy older people, but also always in Alzheimer's disease. And yet, that's one of the last places we think to get sick. It has normal usage of sugar, normal glucose metabolism until late in the game.

And then the flip side is that one of the regions of the brain that we think gets hit earliest, the memory centers of the brain or the hippocampus, is one of the later regions to get amyloid plaques. So right off the bat, it should have been clear, yes, amyloid plaques are a part of the picture, but they aren't really in the places of the brain that are sick necessarily.

But that's also, I think, combined with this problem of, "Well, maybe it is amyloid plaques, but we're just not starting early enough." Now that we have this amyloid PET scan, which was this incredible contribution to our diagnostic toolkit in the early 2000s. It's very easy to give somebody a little injection in a vein.

Put them in a scanner like an MRI, and you can see whether this tracer is binding to amyloid plaques in the brains of humans. So we can detect this very easily now. We see it in healthy older people. Once people are over 55 or 60, you'll start to see some percentage of the population that's healthy develop these amyloid plaques. Logistically, it's tough to go very early.

It makes for a longer trial, a more expensive trial, probably more individuals enrolled in the trial, so we've already gone fairly early. These people are symptomatic but they're not yet demented, so they have memory trouble but they can still run the show on their own. And we've seen these middling results that I described earlier.

And we've even tried in some early genetic mutation carriers to intervene here at an earlier stage without much sign of any efficacy. So I don't think it's the timing necessarily, but it's still a little bit of an open question, for sure.

Nicholas Weiler:

So you've been focused a lot on the genetics of Alzheimer's as a way to look for better therapeutic targets. A lot of our listeners may know that there's a strong genetic component with Alzheimer's.

Could you start by giving us a sense of how strong the genetics of Alzheimer's are? People may have heard of APOE4, for example, which is this risk variant. How big of a deal is that?

Mike Greicius:

Yeah, so E4 is a big deal. Overall, we think if you take all comers, early-onset, late-onset Alzheimer's disease, probably something like 60% of the variability will have a genetic component. In a very small percentage of the population, less than 1%, the genetics drive everything. It's a single gene mutation, these are these even less common genes.

Mutations on three genes: presenilin-1, presenilin-2 and amyloid precursor protein. All three of those are incredibly, tightly linked to the amyloid protein. So that's where the amyloid story really took root at these early-onset mutations. These people get sick in their late 30s, early 40s or mid-40s, so it's a much, much earlier onset.

The rest, the vast majority of the genetic variability in Alzheimer's, isn't due to a single gene necessarily, although APOE4 does account for a lot of that risk. But we think it's more polygenic, meaning three or four different variants might determine somebody's risk. And in late-onset Alzheimer's after age 60 or 65, APOE4 is definitely a big risk factor.

But age becomes probably as strong a risk factor as you get up over 75 or 80. Age really drives risk for Alzheimer's disease. So APOE4 is very common in a European ancestry background. Maybe 15%, 20% of people will carry at least one copy of it, so it's not a rare variant at all. And it's not a yes-no variant like Huntington's disease or something like that. It's just a variant that can increase risk.

Nicholas Weiler:

Yeah. I think I remember reading that it maybe doubles your risk of getting Alzheimer's disease. Is that about right?

Mike Greicius:

Yeah. So that's about right, and it depends a lot actually on the ancestral background. This is where trying to enrich the diversity of our genetics studies in the US has really helped a lot scientifically. Not just because of healthcare equity, which is a huge issue in its own right, but also the biology differs a lot between people of African ancestry, European ancestry.

So on a European ancestry background, you're right, one copy of APOE4 bumps your risk up about two or threefold. About a twofold increase with one copy, more like a tenfold increase with two copies, so very strong. On an African ancestry background, interestingly, the risk is much lower for E4. It's still a significant risk factor, but it's not as powerful.

So maybe 20%, 30% increased risk with one copy and more like fivefold or sixfold with two copies. And that's a big area of research interest now to figure out why is the E4 risk so much less on an African ancestry background compared to European ancestry? Actually, greatest on an East Asian ancestry background.

Nicholas Weiler:

Interesting. So is the idea behind looking at all of these and the ethnic background differences to try to understand what is the beginning of the disease? Where does this start?

And what are some better targets where if we could interrupt the process, we might actually make a bigger difference than what we're seeing with these amyloid clearance drugs?

Mike Greicius:

Yeah. I think those are a lot of the questions that the field is asking broadly. So even for the non-early-onset genes that we talked about where people get sick in their 40s. For APOE4 in particular, it really already drives misprocessing of amyloid even before people show any symptoms. So it's yet another indication that amyloid is a big player in all this.

We can't get around it. It's probably not the plaques but there's something going on with the amyloid. But part of what we've done and what a lot of other groups are thinking about now too, is can we look at people that carry one or two copies of APOE4, this high-risk variant, who are still healthy in their 70s and 80s and 90s?

And ask, "What is it about these people that's seemingly protected them from this, if not totally predetermined genetic fate, at least very strong risk for Alzheimer's disease?" So using that sort of approach, getting dozens or hundreds of people that are at high risk for Alzheimer's because of APOE4, but that are still healthy in their 70s and 80s.

We can start to hopefully pick apart genetic variants, or even environmental exposures or lack of exposures that might have protected them against Alzheimer's disease.

Nicholas Weiler:

You mentioned earlier on in our conversation that the field got very focused on amyloid plaques early on, and basically gave that to the pharma industry and said, "This is our target." They succeeded in getting rid of that and that doesn't seem to help very much in your analysis of the data.

What are the new targets that are coming up that you see as being more promising? Clearly, we don't know yet or we would've heard, but what are some of the other approaches that to you are looking like, "Oh, maybe this one is going to be our next big hope"?

Mike Greicius:

Yeah, so there are a couple targets right off the bat. Tau is one, so we talked about amyloid that you see at pathology and now we can see with amyloid PET scans. These are these big clumps of protein that are actually outside of brain cells. They're in between brain cells, they're not actually in the cells themselves.

Tau is a second protein that tends to track much more closely with symptoms in a patient. So where we see tau in the brain, those are the parts of the brain that aren't working well. They don't use sugar the way they should, they're just sick. So the hippocampus, the memory centers get tau tangles early, early on.

So tau has always at some level seemed like a more meaningful target. So there's, let's call it, an antisense oligonucleotide or ASO. This is a little collection of the pieces that we use to put DNA or RNA together, nucleotides, and they can be perfectly targeted to a target in the brain. A company has developed an ASO against the protein tau, and that's in human trials now.

It definitely looks like it's hitting the target so tau levels are coming down. And I think what's most exciting about it, is not only are tau levels coming down in the spinal fluid, but on tau PET where we can also see tau collections on PET scans in living humans. And these tau PET studies in this trial have shown that the tau PET's actually going down.

Not that it is accruing more slowly compared to placebo, but it's coming out in the brain. And we haven't seen that before in any studies, so I think that's promising. It's very early. We don't really have clinical data yet. This is a bit more invasive so it's not just an infusion into your arm. It's actually a lumbar puncture and the drug is infused directly into the spinal fluid.

But if this works, I think I would have patients lined up around the block to get their lumbar punctures.

Nicholas Weiler:

Do you know when more results are expected from those trials?

Mike Greicius:

Yeah, they're trickling out. They've wrapped up phase one. It looks quite safe. There are side effects associated with lumbar puncture, mainly headache and stuff like that. But there are already approved drugs that are given via lumbar puncture. It slows things down a little bit, would make it hard to scale it across the United States, but it's quite easy to do in the clinical setting. But that looks very promising to me.

And then APOE itself is potentially a target. There's a lot that APOE4 seems to be involved in. There's a notion that maybe E4 is doing more than just affecting the misprocessing of amyloid. It probably has some role at the synapse, that's the connections between neurons. So it's conceivable that E4 is doing more than one thing, not just the misprocessing of amyloid, so I think APOE4 is actually a pretty compelling target.

Nicholas Weiler:

Okay. And then you mentioned earlier that aging itself is a risk factor. I know that people have been very interested in inflammation, for example, as a potential driver in Alzheimer's.

Is there anything else that you think is exciting and upcoming counteracting the disorder?

Mike Greicius:

Yeah, so inflammation's been a tough target. It's tough for a few reasons. One reason is that at some level, we think inflammation is probably useful. So it's possible that there's a period of time where you wouldn't want to reduce the inflammatory response. Maybe it's important for getting rid of plaques or handling neurons that are sick, that are filled with tau that you do want to get rid of, for example.

So there's some notion that inflammation's a protective mechanism, and then, of course, it can tip the balance and become hurtful. So everybody knows about autoimmune conditions like rheumatoid arthritis, lupus, multiple sclerosis. And it's conceivable that Alzheimer's has some of that tipping of the balance where there's at some point too much inflammation.

Or in some particular region of the brain, too much inflammation to where we'd want to knock it down. But teasing that apart in when you'd want to intervene and where in the brain, which cells would you target is a little tough. But I do think the brain's homegrown immune system that these microglia are like the brain's immune cells.

It's clear they have a huge role in Alzheimer's disease. Again, I think at some level, some of what they do might be protective, and at some other level, some of what they do is probably detrimental. So people are very interested in proteins that are made by microglia. So I do think it's a rich vein to mine, but it's going to come with some learning costs I think, for sure.

Nicholas Weiler:

All right. Well, thank you so much for taking us through your concerns about these amyloid plaque clearing drugs. Giving some guidance to listeners who may be thinking about these things, and people who may just be interested to know.

Some of the hype may be overplaced, but we do have some very promising lines of evidence, some things in trials, some things not yet in trials. So hopefully in the next five to 10 years we'll get something that's really effective.

Mike Greicius:

Yeah, I think that's a good, optimistic note to end on. I am quite enthused about a couple of these new therapies that are already in humans.

Some about to, like efforts to knock down APOE4, for example, should be starting in the next year or so. So we're going to make a lot of headway I think in the next, I'd say, five years there. How's that for optimistic? Not even 10.

Nicholas Weiler:

I'll take five, that's fantastic. All right. Well, thanks, Mike, so much for coming on the show.

Mike Greicius:

Thanks, Nick, for covering the topic. I really appreciate it.

Nicholas Weiler:

Thanks again so much to our guest, Mike Greicius. Greicius is the Iqbal Farrukh and Asad Jamal professor of Neurology and Neurological Sciences at Stanford School of Medicine. He's also a steering committee member of the Knight Initiative for Brain Resilience here at the Wu Tsai Neurosciences Institute. To read more about his work, check out the links in the show notes.

If you're enjoying the show, please subscribe and share with your friends. It helps us grow as a show and bring more listeners to the frontiers of neuroscience. As always, we'd love to hear from you. Tell us what you love or what you hate in a comment on your favorite podcast platform, or send us an email at NeuronsPodcast@Stanford.edu.

From Our Neurons to Yours is produced by Michael Osborne at 14th Street Studios, with production assistance from Morgan Honecker. I'm Nicholas Weiler. See you next time.