All things Amyloid, including Aducanumab and Amyloid PET scans with Gil Rabinovici

Eric: Welcome to the GeriPal podcast. This is Eric Widera.

Alex: This is Alex Smith.

Eric: And Alex, I’m very excited today. I haven’t been this excited in a while. Who do we have with us today?

Alex: Today we have Gil Rabinovici, who is Professor of Neurology and Radiology at UCSF and director of the NIH Alzheimer’s Center at UCSF. Welcome to the GeriPal podcast, Gil.

Gil: Hey, Eric and Alex. Great to be here.

Eric: We’re going to be talking about all things amyloid and dementia today. Really excited, topics like PET amyloid scans and all these new drugs out there that amyloid antibodies-

Alex: mAbs.

Eric: The mAbs as Alex calls them. Is that just Alex or is there more people besides Alex, Gil?

Gil: It’s much easier to say than monoclonal antibodies. And bonus points if you can pronounce all the antibody names correctly today.

Eric: We’ll see how many me and Alex can actually do. But before we get into that topic, Gil, do you have a song request for Alex?

Gil: I do. I’m going to request the song I’m Not Going to Miss You by Glen Campbell.

Alex: And tell us why.

Gil: So I think as many listeners probably know Glen Campbell suffered from Alzheimer’s disease. And actually a documentary was filmed documenting his last tour as he performed while he was symptomatic with the disease. And this is a song he wrote with a close friend expressing his feelings about what it’s like to have Alzheimer’s disease, what some of his thoughts are. I think it’s very poignant. It’s really a beautiful love song, and it gives a voice to people with Alzheimer’s, which I think is very important. So thank you for indulging me.

Alex: Yeah, I love that. And it’s a beautiful poignant song as you say, but it’s also a little tongue and cheek. “I have dementia. I’m not going to miss you.”

Gil: I know. It could go dark, but I think it’s really more reflecting his sense of humor as I read on how he wrote the song.

Alex: Right? Okay, here we go. A little bit. (singing)

Eric: Dang. That’s kind of heartbreaking

Gil: Gets me teary eyed every time.

Eric: So Gil what we like to start off with all these podcasts is how our guests became interested and passionate and the work that you do, in particular around dementia. How did you get interested in this field?

Gil: For me, it really started, Eric, with falling in love with the brain as an organ. And this started actually even before I went to college. A good friend of mine was in a head-on collision with a bus and sustained a traumatic brain injury. He’s made a fantastic recovery, but after he was hit, he had a really dense retrograde amnesia where he recognized his family but he really had very little memory about basic elements of his life. And then as he recovered, he had this amazing experience where memories would come flooding back to him. So he would describe walking back into the room where he had grown up and having a flood of memories about things that had happened in that room from a birthday cake to a first kiss. And he described how he would take a book off his bookshelf and had no memory of ever reading the book, would read the first few pages, and the whole plot would just come flooding back into his brain.

Gil: So I thought this was just about the most fascinating thing I’d ever heard. And I started to read Oliver Sacks, which I think is true of many neurologists, a fabulous writer if listeners haven’t read his work talking about brain behavior. And then in college, I had this fabulous mentor named Bob Sapolsky at Stanford who gives unbelievable talks about brain behavior relationships. And when I was a senior, he taught a seminar called Mechanisms of Neuron Death which was basically a seminar about neurodegenerative diseases and what was known about them back in the late nineties. And I look back just a few years ago on some of the notes that I’d taken about Alzheimer’s disease, and actually our knowledge about the disease hasn’t changed that much since I took that class in college. But that kind of set me on that trajectory. So I fell in love with behavioral neurology before I even knew that I wanted to be a doctor. And then there was a strong feeling that I wanted to also take care of patients. And so the path to neurology and behavioral neurology was paved by those early experiences.

Eric: And we just did a podcast, that was about a month ago, Alex, I forget with Jason Karlawish about his new book. And man, we talked about how maybe things haven’t changed a lot in the last 20 years, 30 years with dementia but you get a sense that the next 10 years maybe ushering in potentially a new era or not. How do you feel where we are right now with dementia, dementia diagnosis, dementia therapeutics? Do you feel like we’re at that precipice?

Gil: I do. I do feel like we’re at a precipice. So when I think back just on how much has changed in the 15 years that I’ve been training to take care of patients with dementia and caring for people, a lot of the breakthroughs have happened in the biomarker space. So our ability to identify the pathology of plaques and tangles in living people initially with spinal fluid tests and then with imaging tests, PET scans, and now in the last few years, I think really exciting progress in blood biomarkers that can very reliably detect the amyloid and tau in a much more scalable cost-effective way. And so these biomarkers are incredibly important for a lot of reasons. First of all, they help us understand the human disease. And so a lot of what we know about Alzheimer’s is really what we know about mouse-zheimer’s disease. It’s based on animal models that are based on-

Eric: Did you just call it mouse-zheimers?

Gil: Mouse-zheimer’s. Yeah. You haven’t heard that thing before? Yeah.

Eric: I was like, wait, what?

Gil: Yeah. I don’t want to belittle…I think animal models are really important, they teach us about molecular mechanisms. But it is important to understand that the animal models that we have for this disease are based on mutations that cause familial Alzheimer’s. And those mutations cause 1% of cases of Alzheimer’s. And so the changes that are seen in the mice may represent 1% of the totality of the human disease. And what we can now do thanks to advances in biomarkers, is we can follow people prospectively, longitudinally, and understand more about how amyloid and tau are forming as the brain ages, how they interact with the brain structure and function, how they ultimately lead in some people though not all, to cognitive changes. And so having that human model to study the disease has been incredibly important.

Gil: We don’t have biomarkers for everything that happens as the brain ages and we might get into that, but that’s another problem, which is that we can only see part of the story of what’s going on in the brain at a molecular level. The other thing is it’s really transformed how we do drug trials. And so in the past, when people did a clinical trial say for an amyloid lowering therapy in Alzheimer’s disease, we included only people at the dementia stage because we couldn’t really reliably diagnose Alzheimer’s before that. We included them based on clinical criteria, which when you compare them to autopsy results are only about 80% accurate even at expert centers. And then finally, we actually had no idea if the drugs that we were giving in a drug trial were engaging the target. Were they modifying the biology of the disease? And so we were really running blind.

Gil: And to me, these advances that we have in biomarkers are a prerequisite for being able to do better drug trials. I think we are doing better drug trials now. And that I think, is over time, inevitably going to lead to effective therapies. So I do see major progress. You commented on my background, I am a Cubs fan so by nature I’m a very optimistic person, so take everything I say with a little bit of a grain of salt. But that said, I think we really are at a precipice and at a very exciting time to be involved in this disease as a researcher and as a doctor.

Alex: I was going to say, one of the other points that Jason Karlawish raised in his book and in the podcast was we’re discovering… We’re now pretty certain that Alzheimer’s is a heterogeneous group of diseases. And that we will be discovering ways to test for different forms of Alzheimer’s and to treat different forms of Alzheimer’s. I wonder if you have any commentary on that observation as well?

Gil: Yeah. I mean, you could think about that on two levels, right? So one is, dementia is heterogeneous. And so what we see and what matters to patients of course is when they develop changes in their cognition and their function, but that is a result of a variety of different molecular processes. And plaques and tangles are part of that story, but we know that there are many other important things that contribute to cognitive decline; vascular changes in the brain, other proteins that accumulate in the brain like alpha-synuclein and TDP 43. And so the reason that any individual might develop dementia might differ quite a bit. And we can only measure some of those molecular changes now, amyloid and tau, yes, the others not so much.

Gil: The other thing about heterogeneity is, is Alzheimer’s defined as a plaque and tangle disease, is that a heterogeneous disease? And I do think that probably we will learn that it is. In some people, we talked about 1% of people who have autosomal dominant mutations that cause them to overproduce this Abeta peptide that forms plaques, these people get Alzheimer’s disease with about 100% penetrance with these mutations. So that’s 1%. We know that about 60% of people have a genetic risk factor called APOE e4, and that may contribute to the risk of a disease through a variety of different mechanisms. But there are also probably other pathways that lead to plaques and tangles, and they may differ from patient to patient. So in some patients, it may be related to some inflammatory event that led to activation of the innate immune system in the brain, and that might’ve occurred and precipitated the pathology of plaques and tangles.

Gil: In other people it may have been a traumatic brain injury that maybe set some of these things in motion. In other people it may be vascular disease that leads to changes in the inflow and outflow of some of these proteins. And so I think there are probably different upstream pathways that are leading to this common phenotype of plaques and tangles. And then the final angle you might think about it, is the clinical phenotypes of the disease are heterogeneous. So most people present with memory loss, but other people who have plaque and tangle disease present with primary problems with their language function or their spatial function, or in some people even changes in their behavior and personality. And so there’s a lot of layers to heterogeneity in this disease and we’re just starting to peel some of those layers off now.

Eric: Given that, how important right now, given that… I mean, We won’t dwell a lot on the cholinesterase inhibitors and Namendas out there of how effective they are. I think in general, my feeling is meh! Is there a good summary of the evidence around them? It doesn’t really help that much.

Gil: Yeah. It’s interesting though Eric, because some people do respond in the sense that they actually do get better. These tend to be people, especially who have a Lewy body disease where acetylcholine deficiency may be a big driver of some of their functional brain changes. For most people I agree, they don’t notice an improvement. It doesn’t stop or decline over time, and it’s very hard to know if these drugs even work because people don’t feel better and over time they get worse. And there’s nothing you can measure to say the drug is or is not working. So absolutely no argument that we need much better therapies.

Eric: Yeah. And it’s really exciting to think we can target these things like amyloid. But before we talk about that, right now with let’s say targeting from a diagnostic perspective with PET amyloid scans, where do you think we are right now with the use of both these PET amyloid and these biomarkers in helping us determine what’s going on with this heterogeneity of this person that’s in front of us that may have memory complaints, or we may have concerns around their memory? Where do you feel we are right now?

Gil: So I’m coming at this from a perspective of a dementia subspecialist, right? So people who come and see me or others in our center are really looking for answers. And the biomarkers in and of themselves are not sufficient to diagnose Alzheimer’s, but in the right clinical context just like any lab test that we use in medicine, they can actually provide a lot of certainty about what you are or are not dealing with. And so a common question is, “I’m experiencing some mild symptoms, are these normal for age or is there something more insidious that I need to worry about?” A lot of patients who come to see us have atypical symptoms that are alluded to a little bit with the atypical Alzheimer’s phenotypes. These folks are dismissed a lot of times as just having depression or a midlife crisis. They tend to occur in actually younger people in their fifties and sixties, where Alzheimer’s may not be high on the radar of the physician who’s doing the first assessment.

Gil: The people who develop visual problems often end up seeing an optometrist, an ophthalmologists. They might have cataract surgery before someone actually figures out that it’s a brain problem, not an eye problem. And so in those patients who are really seeking answers, when we can recognize these phenotypes and then we have a biomarker that can support the presence of the key pathology, that can really I think empower patients and families to understand what they’re dealing with at a relatively early stage, and it gives them choices about future planning, the ability to guide the next few years when the patients are still in an early stage and can actually make these decisions and have a voice in these decisions early on.

Gil: A lot of people actually amazingly find it a relief when I show them an amyloid scan and there’s a lot of red that you can see, and that represents the amyloid in the brain. And they say, “Because you have these symptoms and we have evidence of amyloid and you’re relatively young, so we don’t think the amyloid is just incidental there.” They feel like these symptoms that have been sometimes for years dismissed as nothing to worry about, are validated with a true biology. And now they’re like, “Okay. Now I can end this diagnostic odyssey and move on to the next step.” Which is, what do I do? Do I decide to be in a clinical trial? Do I retire from my work? A lot of really core decisions and also having that objective evidence. I’m sure you guys have seen this quite a bit, but sometimes there’s a lot of controversy where the patient and their spouse may not agree on what’s going on. And the child that lives in New York, or in California who talks to mom once a… This is a familiar story. I can see you smile-

Eric: When I was in New York, it was always the daughter from California.

Gil: Okay. So having that biological evidence can end a lot of these kinds of disagreements. And everyone can come to the same place of, “This is serious and this is what we’re dealing with. And now let’s move on and think about the next step.” So I found that the amyloid PET can help. Tau PET scans are something that we’ve been able to use also. And they’re very helpful because unlike amyloid, there’s a really strong correspondence between where tau tangles are in the brain and the specific symptoms that patients have. So people who have memory loss, show tangles in memory areas in the temporal lobes, et cetera. So looking at those pictures also can help patients and families understand the disease.

Gil: Now that said, these biomarkers are expensive, they’re not accessible everywhere, and that’s where I think blood markers are really going to transform the field. Because if you have a blood marker that can tell you about amyloid and tau in the brain with pretty high accuracy, or even if it can just rule out the presence of these proteins in a large number of people, reducing the number of people who need to have a PET scan or a spinal fluid test to confirm that these proteins are present in the brain, then I think we’re dealing with something that’s much more tenable from a public health point of view, from a global health point of view. As opposed to saying everyone who is concerned about symptoms is going to need a very expensive invasive test.

Eric: And for a PET amyloid, how much does the amount of amyloid in the brain actually correlate with the severity of disease, or potentially where they are in the progression?

Gil: It depends on the stage. So in what’s called the preclinical stage of the disease, where there’s evidence that the pathology is happening but people don’t yet have symptoms, or even in the early symptomatic stage, people with higher levels of amyloid are more likely to progress than people with lower levels of amyloid. But then it reaches a relatively early plateau, even in the stage that we call Mild Cognitive Impairment or MCI where people have cognitive impairment but are still functionally independent. And so at that point, you lose any correlation between how much amyloid is in the brain and the symptoms. So at that point, it becomes more is the pathology there or not?

Gil: But one misconception I think that people have about amyloid PET scans, is that they only represent the amount of amyloid. And it’s true, the tracer is binding to amyloid, but actually amyloid is very strongly related to tau tangles and tau tangles are very strongly related to symptoms. And by the time there’s enough amyloid in the brain to have a positive scan, there’s actually a lot of tau in the brain as well. And so, even though it’s not directly measuring plaques and tangles together, once you have a positive PET scan, it tells you not only is there significant amyloid, but very likely there’s also significant tangles. And that information has much stronger prognostic value. And so these scans aren’t super sensitive to early stages of amyloid, and once they’re positive, we found that they are about 90% accurate in our center for example, not just in predicting amyloid at autopsy, but who actually ends up with an autopsy diagnosis of Alzheimer’s disease as the cause of their dementia syndrome.

Eric: Yeah. And I remember from another one of your studies, I think about a third of patients who we thought, and these were in neuro centers, that we thought had Alzheimer’s disease had a negative amyloid. And about a half of those that we thought didn’t have Alzheimer’s disease actually had some amyloid in their brain. Is that about right?

Gil: Yeah. As clinicians, we can only get so far in terms of characterizing people clinically. And at the end of the day, it’s the biology that’s going to matter probably, at least in terms of some of the therapies that we’ll be discussing today. And so it reminds me of in med school when you were auscultating the heart and you were learning to do all of these provoking maneuvers to try to distinguish aortic stenosis from mitral regurgitation. And at the end of the day you do the echo and you figure out which one it is, right?. So that reminds me of that, trying to use our clinical acumen. We’re pretty good at it, but not good enough. And especially not in early stages where we’re trying to identify a biology based on pretty mild clinical symptoms that can be difficult to distinguish just from aging.

Eric: Do you think, and this is my last question about amyloid scans until maybe after we talk about future therapies, do you think in a wider practice outside of where the-

Alex: Alzheimer’s centers.

Eric: Alzheimer’s centers, where people actually are thinking this through and working through the diagnosis, do you think there is a wider role right now for these scans in the community?

Gil: The current guidelines for appropriate use that have been published by the Alzheimer’s Association and Society of Nuclear Medicine really state that at the moment, these should be used by specialists. That they don’t have a role, I’m talking about PET scans and spinal fluid tests, they don’t have a role in primary care. They really should be considered an adjunct to a clinical evaluation by a specialist, including all the typical things that we do in labs and structural imaging, CT, or MRI. And if a specialist at the end of that workup is still uncertain about the cause of impairments and an amyloid scan is going to change their diagnosis and management, which it may or may not depending on what their differential is, that’s really the utility right now.

Gil: And I think one of the concerns about blood tests, which will be accessible to everyone and will probably be marketed, in fact, we have some evidence now that they’re being marketed directly to consumers. There’s a risk not only that these biomarkers will be used and interpreted by the wide population of physicians, but actually the patients themselves might be getting these tests and then coming to the doctor with the test in hand, asking for an interpretation. And so blood tests really changed the game here. And while I think that these are best used by specialists who understand the nuances and the limitations of all of these tests, the reality is just as with direct to consumer genetic testing, we’re very likely to be looking at direct to consumer Alzheimer’s blood biomarkers in the next five years. And we’re going to have to adjust to it in terms of our clinical practice

Eric: And do the blood biomarkers supplant the PET scan, or do we use them in conjunction? How do you think they’ll be used in the future?

Gil: The data so far, and these need to be replicated in larger and more diverse cohorts and more in real care than an academic centers, so with a lot of caveats, the markers seem to be pretty good at telling who has amyloid and tau in their brain and who doesn’t. They don’t tell you what the imaging tells you, which is where or how much, but they could be very useful in ruling out amyloid and tau in a large number of people. And then if you really need to confirm their presence, you can do a spinal fluid test or a PET scan, which may be more specific. But you could set a threshold that would be sensitive for these blood tests that would allow you to rule out the disease in the large number of people. And I think that would be very helpful for practice but inevitably, it’s also going to detect amyloid and tau in people who don’t have symptoms. And they’re going to be looking to find out if they’re at risk for Alzheimer’s. And I think we as a field, are going to have to adjust to that because it’s coming.

Alex: Are we going to pivot to talk about treatments now?

Eric: Yeah. Let’s talk about treatments.

Alex: Can I do the pivots? I’ve got two pivots.

Eric: All right.

Alex: Okay. Here’s the first pivot. So every year, Eric and Ken Covinsky do a summary of groundbreaking articles in geriatrics at the American Geriatrics Society meeting. And then I do parody songs for selected articles. And so they’re going to present an article about aducanumab this year. And so as a preview, I’ve worked up some lyrics for this. I’m interested in your thoughts on them. This is based on supercalifragilisticexpialidocious.(singing) I’ll stop there. That’s my first pivot.

Gil: So you’ve passed the first test, which is perfect pronunciation of aducanumab which can be quite the tongue twister.

Eric: Aducanumab?

Gil: Aducanumab.

Alex: Okay, the second pivot is that generally we don’t do screening tests without having a treatment. And with the proliferation, as you’re talking about of all these imaging blood tests, many of us can foresee a day in which as you’d say, these will be widely available. What’s in the pipeline as far as treatment and aducanumab in particular?

Gil: Yeah. So I’ll just say one word about that, which is that I think there are even pharmacological therapies. There are a lot of things that people might change in terms of their behavior and lifestyle if they learned that they’re at risk for developing Alzheimer’s in the next few years. And studies have shown this, so even though we all know that we should eat healthy and sleep well, and avoid stress, and exercise regularly, studies show for example, that people who learn that they have a genetic risk for Alzheimer’s, the APOE e4 gene, are actually much more likely to make those changes than people who find out that they don’t have that risk factor. We all know that we should get our advanced directives in line and get everything ready in case we’re not able to make decisions for ourselves in the future, but there’s a powerful motivator when you learn that you might be at risk of losing that capacity. And that, that may be happening much sooner than you would hope.

Gil: And so I do think that there are some very actionable results of finding out. Now would I want to know that I have amyloid and tau in my brain if I have no symptoms? No. But actually we pulled our lab members who are of a different generation, and many of them said they would want to know as they get older if they were at risk. This comes a lot towards, are you seeking information? There’s almost a generational gap I think, in how we think about these things. We just recently had a grand rounds in neurology about blood biomarkers, and we asked whether people would want to know. Okay, that was the last question, it was a Zoom poll. And the split was 51-49. And so I think that there’s no correct answer and no easy answer.

Eric: And just because I have, let’s say amyloid in my brain, does that mean I’m going to at some point develop the symptoms?

Gil: It means you have a higher risk of developing cognitive impairment, it doesn’t mean that you will develop it inevitably. The studies that follow people prospectively show that the risks separation between people who are positive or negative for amyloid starts to occur around three years after the biomarker assessment. And then it gets bigger and bigger as you follow people over time. Such that if you follow people for 10 years, maybe 70% who were positive are likely to develop some clinically meaningful cognitive impairment, but not everyone. Others might die of other causes, some people might have resilience in their brain and never develop symptoms. So it’s a risk factor and it should be thought of as that, not as a determinant of developing impairments.

Alex: And also I want to say thank you for pointing out that there are things we can do. Some of our other guests on a prior podcast about comprehensive dementia care said, “We need to stop saying there’s nothing we can do about this, because we know there are things we can do.” And many of these are non-pharmacological treatments. And a dementia care program, the care ecosystem, all of the factors that you mentioned just a moment ago. But I think we want to pivot to aducanumab before-

Gil: Yes, I’m stalling. Okay.

Alex: The idea of the amyloid hypothesis, that we are going to look for amyloid and if amyloid is there, we’re going to get rid of amyloid potentially through these amyloid antibodies. And the next rational thing is if we get rid of it, people’s cognitive issues either will stabilize or improve as compared to if we did nothing. So where are we right now with specifically the antibodies towards amyloid?

Gil: Yeah. So the antibodies have cured mouse-zheimer’s. They all have cured mouse-zheimer’s very definitively.

Alex: Yeah. Cured the disease, right?

Gil: Yes. The mice are delighted. But none of them so far have proven efficacious in human disease. Now it’s quite an interesting story. So initially the antibodies that were tried in people were dosed too low, there’s no doubt about that. Eventually, as we got to higher doses, we learned there is a significant side effect where if you dose antibodies high enough to remove amyloid in the brain, that actually leads to a risk of developing inflammation or hemorrhage, something called Amyloid-related imaging abnormalities, ARIA. And that comes in the form of edema more commonly or hemorrhage. And so there was a lot of concern about this specific side effect. And that caused a mess in some of the clinical trials that were done because the dosing… Now I’ll make a long story short and say that what we’ve learned is that this is something you have to monitor with MRIs, but in most cases, two-thirds or more of cases, it’s entirely asymptomatic.

Gil: And then in other people, it causes very mild symptoms. And it can actually be treated through if you reduce dose and then slowly escalate back up. But there was a lot of worry about do no harm. And so the dosing of these drugs got messed up in a lot of trials. And it got messed up in a very specific way because people who carry APOE e4, we said about 60 or 70% of people with Alzheimer’s have APOE e4, were at higher risk of getting this side effect and so they systematically received lower doses. And that made a lot of trials that are difficult to interpret. And that’s part of the story of aducanumab.

Gil: So let me tell the story of aducanumab, it’s a really fascinating story. This is a drug that made the cover of Nature, my lifelong dream, because in a phase 1b… It’s a humanized, I should say, monoclonal antibody. It targets both aggregated and soluble forms of the Abeta polypeptide. And on the cover of Nature, it showed in a phase 1b that it can very dramatically remove amyloid PET signal. And so people who had a lot of red signal, and you treat them in a dose dependent matter, you can just watch that amyloid clear. And so it generated a lot of excitement. This was the first drug that very definitively hit the target and lowered amyloid. And then the next question was would that translate into clinical efficacy? Would it slow decline? And so that led to the launching of two phase 3 studies called emerge and engage. And these were randomized control trials that were done in parallel, a little bit staggered of either placebo low dose or high dose aducanumab.

Gil: And there were, I think about 1600 patients in each trial that were randomized to receive one of these three treatments. And in March of 2019, there was a sudden announcement by the sponsor that the trial was terminated because of a pooled futility analysis that was done in both trials together, that showed that there was a low likelihood of hitting the clinical endpoint. The clinical endpoint was a statistically significant difference on a scale called the Clinical Dementia Rating scale Sum of Boxes, which is a measure of functional impairment in Alzheimer’s disease. And so a lot of patients were actually really upset. There was a lot of excitement about this trial. Some patients felt like they were benefiting, who knows if that’s true or not. They were devastated when this trial was terminated.

Gil: And then life moves along and about seven months later, the company says, “Wait a minute, we actually have more data now. We’ve reanalyzed it, and we actually are calling one of these two phase 3 studies, the emerge study positive. It hit the end point, and the second one is negative.” And I remember this was close to Halloween because one of our fellows made a meme of a skeleton arm rising from the grave and called it aducanumab. So this drug that was dead in the water was suddenly resuscitated. And the sponsor Biogen, I think we can say it on the podcast, went to the FDA with these data and asked for FDA approval based on what they felt was a very compelling clinical trial results. I talked a lot about dosing because they’ve tried to explain why one study was positive in the high dose group. I should clarify that, both were negative in the low dose group. In the other trial, the high dose group did no better than placebo. In fact, it did worse than placebo.

Eric: It’s flat out negative –

Gil: And so you got these very discrepant results. One of the differences was the dosing of the drug in APOE e4 carriers. And in the positive study, they were able to escalate the dose because they made a protocol change and that protocol change occurred earlier in recruitment for one of the trials than the other. And so the trial that was negative, fewer e4 carriers were dosed at high dose. The trial that was positive, more of them were. So in a post hoc analysis, this is something that has been used to try to explain the discrepancy between the results. And then the company worked with the FDA for months trying to understand the data. And some have said that this relationship might have been too close between the sponsor and the FDA.

Eric: There was a brilliant public citizens document that I read that really went after the FDA process.

Gil: Yeah. Depending on your perspective, is the FDA really working hard to try to understand if there is a signal for a disease that desperately needs a treatment, or are they getting too cozy with the sponsor? it depends. And so all of this culminated in an FDA advisory committee that happened on November 6, 2020, a calm week in our country, not much going on that week. But after the election on Tuesday, Friday, there was this FDA advisory committee. And there was a presentation from the sponsor and the FDA that tried to show the data in a very positive light. There was another review by the FDA statistician that was very skeptical about the results of the trial and the post hoc analysis that were used to rationalize the different results. That was actually not presented at the meeting, but the advisors read the FDA statistical review. And at the end of quite a contentious advisory committee meeting, basically unanimously the advisory committee recommended against approval of the drug based on the current data.

Gil: We still don’t have an FDA ruling on this. As we record this on April 1st, the FDA ruling will occur I believe on June 7th or earlier. And I think there are some people who still think the FDA might offer some sort of approval or conditional approval, even though the advisory committee was very much against it. I am skeptical about post hoc analyses. I do think there has been a trend and we don’t have time to go in detail into all the data that suggests that amyloid lowering therapies have a modest clinical benefit in early stages. This is based on more than one trial. So I would not be shocked if the trial that was positive was correct and there were some issues. But I think the only way to really resolve that is to do another well-designed phase 3 RCT, dosing the drug correctly for everyone. And then you’ll see basically, are you able to replicate these results or not? And to me, that is the most logical path forward.

Gil: I do worry that if the drug is approved based on the current data, there’s really going to be a polarization in the field where some people are very gung ho pro and some are very gung ho against. And patients are going to be stuck in the middle. And depending on which doctor you talk to, people are going to be very confused about whether this drug is beneficial for them, whether they should be treated with it. And again, as we talked about, the side effects are not inconsequential and the cost is likely to be high. And this involves monthly infusion so it’s not a benign intervention.

Eric: You’re Right. I heard it’s probably like $50,000 a year for the medicine. You’ve got to get your your PET amyloid beforehand or your biomarkers. And then I think about 40% had some imaging abnormalities, so you get your pre-MRI, you get your post-MRI and potentially monthly, or do you think it’s going to be Q3 month MRIs at least in the beginning? Because it sounds like this type of edema happens earlier on, is that right?

Gil: It does, it happens earlier on. So not quite monthly and we don’t know exactly what the label will say in terms of safety, but it would require some monitoring. That’s correct.

Eric: So a lot that’s going on. And I guess my question to you is, if we take a big step back and let’s pretend engage trial never happened, you didn’t have this negative trial. We have this emerge trial where the low dose didn’t seem to work, but the high dose did, a 0.4 difference on CDRSB. It’s a 18 point scale. I always thought one, one and a half points a clinically meaningful difference, is 0.4 something? How do you think about… Obviously there are probably some people do better and some people do worse, this is a population based average. Your thoughts on, is this a clinical meaningful difference?

Gil: I think it’s very hard to measure at early disease stage, which is where we’re treating now clinically meaningful differences because these scales are really designed… they accelerate in the dementia stage. And right now a lot of the focus on treatment is in the MCI stage or earlier. And so if you just look at the progression in placebo, I think it was around one and a half points on the Sum of Boxes. And so you really need a powerful drug, which is basically going to stop any progression in order to measure a larger change. And so the challenge is when you’re doing these trials over 78 weeks or whatever, 18 months is pretty typical, how do you measure change in a way where you can really judge?

Gil: And one thing we don’t know is if you were then to extrapolate that, say someone might live another 10, 12 years from their diagnosis of MCI, would that be a 0.4 change per year that would over time separate and turn into something that might be very meaningful in terms of preserving quality of life? Or is it going to be like the cholinesterase inhibitors where those curves are kind of parallel after some time, and there’s not a cumulative benefit. We don’t know. Someone very smart, I won’t say who it is said, ” One problem with this disease is that at the time that we might actually be successful with treatment, we won’t be able to measure that with clinical measures. And by the time we can measure things with clinical measures, it might be too late to treat.” So we’re caught in this catch 22.

Gil: And so I understand what you’re saying about clinically meaningful, but I think that setting that context of how much does this scale or another scale change over the duration of a clinical trial, it’s important to set that expectation.

Eric: And I guess another question, because it sounded like this drug did a really good job of getting rid of amyloid, but we didn’t see a bit… Even if you trust the one of four positive findings, so the low doses didn’t work, the high dose in the other study didn’t work. And the emerge high dose did, not a whole lot over the course of the 18 months. Given that we can cure, what do you call it again? Mouse-

Gil: Mouse-zheimer’s.

Eric: Mouse-zheimer’s. We can cure mouse-zheimer’s-

Gil: My basic science colleagues are going to kill me.

Eric: We can cure mouse-zheimer’s but we can’t really dramatically change their symptoms, at least over the course of the 18 months. How much does that say about the whole amyloid hypothesis too?

Gil: I don’t think that the amyloid hypothesis is likely to be entirely incorrect because of the genetic data. We know that mutations that lead to overproduction of amyloid are sufficient to cause disease. But I think what we don’t know is how much amyloid lowering is likely to change the course. So I can give you my opinion because I know we’re running short on time-

Eric: Yeah. Give us your opinion, Gil.

Gil: Yeah. I mean, I think that if amyloid lowering therapies are going to work, they’re going to work very early on and try to prevent the spread of tau, which amyloid does seem to facilitate. And it’s really that spread of tangles through the brain that correlates with decline. And so if these are going to work, they might work very early on. Ideally you would have an oral medication if you’re thinking more about prevention than treatment, you don’t want to be treating people preventively with mAb infusions. But I think that drugs that are going to slow the progression of tau are more likely to be efficacious at the stage of clinical impairment because there is that very strong correlation between how tau spreads and the symptoms that people develop.

Gil: And with PET scans, we’re learning that it’s not incidental. People used to talk about the tangles as the tombstones of the dying neurons; that is not correct. You can look at where tau is in the brain and anticipate which brain areas are going to fail in the coming years. And there are studies that have shown that. So I think where the field will ultimately end up is in a combination therapy where we are treating more than one aspect of this biology. I think amyloid lowering therapies are likely to have some role especially at an early stage in that cocktail. But I think that other drug targets are more likely to be more robustly efficacious in later stages. And to think back to early in our conversation, if we can really understand it in individual patient’s level, why are they developing the symptoms? and the biology, we’re going to… Kind of where cancer is, where you can take a biopsy and understand the specific mutations that are driving the tumor in that patient, then I think we’ll be able to make really powerful gains in our ability to treat the disease.

Gil: So we’re just at the beginning stages of that. I’ll leave you with one analogy, someone compared aducanumab to AZT, in thinking about HIV treatments. Horrible drug, lots of problems, lots of toxicities, but it was the first one that opened the door to much more effective therapies, ultimately for people living with HIV. And I don’t know if it’s going to be aducanumab or another one of these, but I do think that that’s where the field is right now. We’re at the AZT stage. And one of these drugs will probably be approved eventually and it will open the door, however, to more effective approaches. So stay tuned, I think it’s going to be quite a roller coaster in the next few years.

Eric: I am eagerly awaiting for all our wall street bet folks out there. Biogen, I’m not sure which we’re going to go on that one, but it sounds like the next couple of years are really going to be exciting. Gil, I could talk to you for another three hours about this. I really appreciate all your time. I am just absolutely fascinated what this will bring over the next couple of years and appreciate your thoughts and sharing it with us today.

Gil: Yeah, my pleasure. Thanks for inviting me.

Eric: But before we end, wait, which song are you going to sing Alex now? Are you going to sing the aducanumab song? Or the-

Alex: No, that was just a preview. We’re going back to I’m Not Going to Miss You.

Alex: (singing).

Eric: My heart just breaks. Gil, a very, very big thank you for joining us today. I learned a ton and it’s great having you on the podcast with us.

Gil: Yeah. Thanks a lot for inviting me. It was a lot of fun and hopefully we can continue the dialogue-

Eric: Yeah. We’ve got to have you back in a year once everything is approved by the FDA and the world is a very different place.

Gil: It might be. I hope that FDA doesn’t approve it. But the Donanemab study which we didn’t have time to get into is a much cleaner study. And it did show that lowering amyloid reduced the spread of tau. So there’s something there, but it’s very-

Eric: That was a phase 2 study?

Gil: Phase 2. Yeah. So they’re going to proceed with a larger replication study.

Eric: Hopefully they don’t stop it midway through.

Gil: Yeah. Well, the other thing that was great about that study is it was like induction therapy. So they treated people until the plaques were gone and then they stopped treating and just monitor it to see if the plaques come back. And so that is again, much more appealing if you’re going to have these mAb treatments, having them be not everlasting but really that’s-

Eric: In part that my work, because I know it’s going to happen. It’s like 10 years on, they’re still on aducanumab. They’re in hospice. They’re still getting their aducanumab monthly injections with at best, a small signal that it works. As also a palliative care doctor and a hospice doctor, that also scares the bejesus out of me.

Gil: Yeah, no, that would be a really bad outcome, I totally agree. But let’s see, I hope the FDA does the right thing and does a phase 3, but at this point who knows? All that’s well…

Eric: Yeah, right. We’ll see, put your money down. Well thanks everybody for joining us for this podcast. And a big thank you to Archstone Foundation for your continued support. Goodnight everybody.

Alex: Goodnight.