Experts in the treatment and management of patients with Alzheimer disease and dementia, including Marwan Sabbagh, MD; Lon Schneider, MD, MS; Anton Porsteinsson, MD; and Robert Howard, MD, MRCPsych, weigh in on the FDA's controversial approval of Biogen's anti-amyloid drug, aducanumab.
Following a long and winding journey to the FDA, Biogen's anti-amyloid agent aducanumab (Aduhelm) crossed the finish line victorious, winning approval for the treatment of Alzheimer disease as the first therapy with the potential to slow disease progression.
The decision came down after nearly 2 years of debate among members of the healthcare, research, and patient communities who grappled with the treatment's promising—but questionable—clinical trial data.
Undoubtedly a monumental decision that has major implications not just for the Alzheimer disease community but for clinical research and drug development as a whole, aducanumab's approval is driving a very apparent wedge between opposing parties who support or question its clinical impact. Much of the same sentiment was expressed by the clinical community following the FDA’s Peripheral and Central Nervous System Drugs Advisory Committee meeting in November 2020, which voted not to recommend regulatory approval of the intravenous infusion drug. Adding further fuel to the fire, the Institute of Clinical and Economic Review (ICER) released a statement following the approval, calling the FDA's decision a failure in the regulator’s “responsibility to protect patients and families from unproven treatments with known harms.”1
In a statement, the FDA said that the data from EMERGE (NCT02484547) and ENGAGE (NCT02477800) suggested aducanumab "consistently and very convincingly" reduced amyloid plaques in both a dose- and time-dependent fashion and that it anticipates that this reduction will result in a corresponding reduction in clinical decline.2 As such, the agency's approval is “contingent upon verification” of this potential clinical benefit in a phase 4 confirmatory trial.3
Keen to take the pulse of the clinical community following the approval, NeurologyLive spoke to several physicians who treat these patients and have varying opinions on this news: these include Marwan Sabbagh, MD, director, Cleveland Clinic Lou Ruvo Center for Brain Health, and investigator in the aducanumab trials; Lon Schneider, MD, MS, professor of Psychiatry and the Behavioral Sciences, and Della Martin Chair in Psychiatry and Neuroscience, University of Southern California Keck School of Medicine; Anton P. Porsteinsson, MD, director, Alzheimer's Disease Care, Research and Education Program, University of Rochester School of Medicine and Dentistry, and site investigator in the aducanumab trials; and Robert Howard, MD, MRCPsych, Professor of Old Age Psychiatry, Faculty of Brain Sciences, University College London.
Marwan Sabbagh, MD: This is a seminal moment in the field. I was surprised at how polarizing it was—we really had strong opinions for and against—but if you take the pro-patient aspect and look at this in the context of the fact that many drugs are starting to show similar signals, I think this is a great step forward. People have to understand that contextually, medically speaking, we need the first drug in the class to get the class going forward. Otherwise, there's going to be no progress. I would remind you, of course, that we always have a first before we get to the third, fourth, fifth—we don't know if aducanumab will be the final drug, but we need the first drug. And we know that it has a consistent signal, we know it can remove amyloid very robustly. I'm overjoyed. I cannot tell you how excited I am to actually have a treatment option for my patients. I've been waiting 2 decades to get a new treatment option.
Anton P. Porsteinsson, MD: It was a relief and excitement for those of my patients where aducanumab is indicated. The relief is that I do believe that having aducanumab approved, basically, is going to slowly usher in a new era in Alzheimer disease. It's not so much because aducanumab is available, but because we have a treatment that targets the underpinning of the disease. Because of that, there will need to be a change in clinical practice. We will have greater access to and greater utility for biomarker validation, be that through cerebral spinal fluid (CSF) analysis, or amyloid PET scans, or hopefully soon, plasma biomarkers. Now, we have something to offer if a patient has elevated amyloid burden. We have a drug that is approved for amyloid reduction, and in the right patient population, this can be a reasonable treatment.
Lon Schneider, MD, MS: The FDA did what they started out to do and what they were going to do, even before the Advisory Committee in November 2020. Perhaps they were surprised by the Advisory Committee's negativism, but it's now easy to understand why they weren't deterred by it. Remember, the Advisory Committee voted unanimously with 1 abstention, to say, there is no evidence for efficacy or effectiveness. And specifically, that the so-called positive study could not be primary evidence of effectiveness and could not be primary evidence of effectiveness with support from the negative study or the phase 1 study, or with support of any pharmacodynamic effect. The FDA asked several times about pharmacodynamic effects, which is kind of unusual, as the applicant presented strong evidence for a pharmacodynamic effect on Alzheimer disease physiology. There, the committee said, yes—though they voted 5 yes and 6 uncertain and they had a lot of discussions—it has a pharmacodynamic effect, the drug hits the target that you wanted it to hit and does what it wants.
This is an antibody; these are antibodies that are designed to break down and lower amyloid plaque. So now, what was the basis for this approval? The basis for this approval was not efficacy, as the FDA statement itself says. It was an effect on a biomarker. The basis of the approval was that this thing broke down plaque, which is what we all know and totally agree on. This is an effective antibody for breaking down plaque--and that's all it is. The FDA has, from the beginning, been relying on their own draft guidance for early-stage Alzheimer disease or dementia, which was from 2018, that said that biomarkers, if valid, could support efficacy, and we're open to considering biomarkers at the time that a new drug or biological licensing application is submitted. They said that.
Then, the next part of this is that the FDA advanced this on the basis of Accelerated Approval. Accelerated Approval is a full-approved pathway—there isn't such a thing as conditional approval—but provisional approval. But very specifically, and as Dr. Cavazzoni said in her letter, the approval of aducanumab is based on a surrogate clinical end point—in this case, the reduction of amyloid plaques in the brain. A surrogate end point is a biomarker that is thought to predict clinical benefit but is not itself a measure of clinical benefit. They're saying essentially, “Never mind the lack of effectiveness, aducanumab lowers plaques and we think plaque lowering is reasonably likely to predict clinical benefits.” This is the whole basis for approval: it lowers plaques, it's reasonably likely to predict clinical benefit, and now we want you to do a study postmarketing—not necessarily a clinical trial, we don't know what the study is—to validate that. Efficacy is optional, effectiveness is optional. This is essentially what is being laid out here, and that, in some ways, is a stunner.
Robert Howard, MD, MRCPsych: They’ve used this Accelerated Approval category, which, to my understanding, is generally used when there isn't really enough data to know whether something works or not. But there's evidence of engaging with biomarkers and good reasons to think that, given more data, we’ll know if it works or not. I suppose my point would be that we've already had 3000 patients go through the aducanumab clinical trials, and there is a large amount, really, of efficacy data available from those trials that indicate that there isn't efficacy—or not significant efficacy. If you set the bar for approving dementia drugs to the point where all you have to do is show that you’ve engaged with a biomarker by reducing amyloid rather than actually having a significant effect on cognition and function decline, that's going to be terribly damaging for the field—and is essentially what they've done. This decision will now open the floodgate for those who've got drugs that have failed to show efficacy but have engaged with the amyloid marker. Almost all of the anti-amyloid antibodies have shown a reduction in amyloid in the brain, even though, of course, they failed on the clinical outcome measures.
We need more blinded, placebo-controlled data. We don't need phase 4 data. Phase 4 data is not blinded—that's useless—and it's going to take 10 years. Basically, what we could have discovered in 3 years, if we've done more trials, is actually going to take us 10 years now, before we realize, actually, that this is useless, isn't it? In our clinics, we're going to see this, and it's such a shame.
Marwan Sabbagh, MD: It opens the door. In the future, the option is going to be like chemotherapy, right? Just like cancer, it's going to be not 1 drug, but an applied drug cocktail. When you look at aducanumab, people forget the fact that, contextually, we're talking about this is in addition to background therapies. If you actually put it from that paradigm, that aducanumab plus amantadine, plus donepezil or a cholinesterase inhibitor, and then you'll see something else added. Our future is going to be a 4- or 5- or 6-drug cocktail, and the consequence, of course, is that we’ll slow the rate of decline.
Anton P. Porsteinsson, MD: First of all, aducanumab can be used with currently available treatments, those medications were not restricted in any way, shape, or form, at entry into the clinical trial. As well, as we know that Alzheimer disease is kind of a multidomain disease. There are amyloid plaques, there are tau tangles, there is neuroinflammation, there is oxidative stress. In order to do combination treatments where you have an amyloid-modulating agent and you add something on top of that, the amyloid-modulating agent needs to be able to stand on its own. I do think that this opens up the possibility for combination treatments be that amyloid-lowering and tau-lowering treatment, amyloid and neuroinflammation, amyloid and oxidative stress, et cetera. And we do know that Alzheimer disease has all of these components.
Do we begin to stage Alzheimer disease patients? If you look at, for example, the TRAILBLAZER study with donanemab and the whole TRAILBLAZER development program, in that study they looked not only at amyloid burden but also tau burden. That was kind of a first and I think we're going to see a continuation of that, where maybe it's going to be the so-called ATN network—amyloid, tau, neurodegeneration; or at amyloid, tau, neurodegeneration, and neuroinflammation. I think that ultimately, and not everyone is happy with this, but I think that the treatment of Alzheimer disease is going to look quite a bit like oncology. In that, your Alzheimer disease will be staged or typified, and you will have validated the amyloid burden, tau burden, oxidative stress, neuroinflammation, et cetera. Then, we'll develop a medication cocktail that treats what has gone awry for you. We are a certain ways away from that in clinical practice, but it wouldn't surprise me if, in the next couple of years, we'll see this emerge in clinical trials.
Robert Howard, MD, MRCPsych: [Combination therapy] is based on the way that the treatments for AIDS and TB have worked, which is that 1 drug doesn't do much, but you put together a combination of 3 of them and you get an amazing effect. That's what I think this has come from. What we need, first of all, is drugs that work. I can buy that we might have drugs that do a little bit, and if you combine them and you hit different parts of the pathway, you get a bigger effect. But we haven't really established that aducanumab works clinically. I just think this has become a sort of mantra that people can use that, of course, we need to do combination drugs. It's very easy to say that, isn't it? No one's got any evidence that it is going to work, so it's a way of sort of comforting ourselves when none of the drugs that we have, at the moment and in development, appear to be working. We think, “Well, maybe if we were to put them all together, it would somehow be better and work.” But until we do the trials, and we actually demonstrate a significant sort of effectiveness or efficacy, we just don't know.
I've been in this field for 30 years. We've always been so optimistic. We've always over-promised that within 5 or 10 years, we're going to have treatments that are going to revolutionize and be life-changing. It hasn't come. I think we should just have a little bit of modesty and just accept that it may take a long time to find treatments that work. Obviously, that's frustrating and difficult for people who've got dementia now. But, you know, what's far worse, I think, is to go down an avenue of false hope. Or basically licensing a treatment that essentially doesn't work. I said this on Twitter, and I don't know if it's an accident or not, but I really feel that Biogen has pulled off a heist with this approval. It's the great dementia heist. Hats off to them. It's amazing what they've done. They've had their enablers and helpers. But considering what they've actually got in the way of data, it's absolutely amazing that they got this far. It was weaselly of the FDA, to say that the Advisory Committee wasn't asked about potential efficacy. Really, really weaselly. It's incredible. The way Biogen staged this, the way my scientific colleagues have allowed them to do it, the way they were allowed to present the data and not take open questions, they've not had to publish the data--it's incredible. Even when they went to the FDA, most of us are able to see what the data really showed. It’s extraordinary.
Marwan Sabbagh, MD: I have patients literally waiting—literally, I have a list of patients ready to go—for as soon as they get the approval, so I can get them infused in a matter of weeks. Patients are desperate. Imagine having mild cognitive impairment, knowing you have amyloid in your brain, and knowing that one direction you're going to head for is terminal dementia and that there was nothing that we could do until now to prevent that from occurring. How great it is, as a doctor, where this has been my whole career, to say, “Maybe that's not your inevitable future.” That's what's so exciting about this moment.
Anton P. Porsteinsson, MD: What I'm worried about, after having read the label, is that the label is fake. It's a provisional approval of a drug for amyloid lowering. There are not a lot of specifications about the severity of the patients. There are some general discussions about how to manage adverse effects. Basically, they say it's indicated for 10 mg/kg and extended use, so there's no stop date, per se. I think that the biggest challenge right now, is how do we operationalize treatment with aducanumab? This is not for everyone. It's not a cure. There is a meaningful potential for adverse effects. Those adverse effects are, thankfully, pretty narrow, but not uncommon. We have about the 35% to 40% that will have ARIA-E and ARIA-H. Whereas for most patients, it's mild to moderate and asymptomatic, there is about 25% of patients that get ARIA-E that will have symptoms, and a small percentage—maybe only 1% or something like that—that will have substantial symptoms that have to be dealt with.
I believe that ARIA-E and ARIA-H are manageable adverse effects, but it definitely doesn't mean that we can be nonchalant about it. There needs to be careful clinical monitoring, there needs to be clinical MRI monitoring, there needs to be monitoring that the titration is done correctly, that you basically screen for any clinical symptoms that would suggest the emergence of vasogenic edema, and that if there is enough suspicion that an unscheduled MRI might be done. We have the recommended MRI monitoring, basically month 6 and month 12, roughly.
This is not a simple process—not for the clinicians, and not for the patients. In my opinion, it has to be done well. We have to select the right patients that are most likely to benefit from treatment, we have to treat them in the right manner (which is a dose-escalation, where the dose is steadily pushed and towards 10 mg/kg), and that everything is done as long as it's safe for ongoing treatment, 10 milligrams monthly for an extended period. Those are the patients that did best.
Clearly, having a treatment that differs from all other available treatments, because it targets the underpinning of disease, and may work to normalize plaque burden, and appears to be associated with a modest of meaningful delay in progression, is exciting and we should make it available. But I really think that there is a high need to have a careful process and careful conversation with the potential patients and family about what aducanumab does and what it doesn't do, and the possible tolerability issues, the cumbersomeness of the infusion, and the monitoring, so that nobody is surprised.
Lon Schneider, MD, MS: There's a complete lack of evidence that this is beneficial for a patient. That the drug is directly beneficial. Can it be beneficial in some way? Well, it can be beneficial in a lot of indirect ways. You can get people excited; you could increase hope; you can lay out the hypothesis to a greater extent—the hypothesis is yes, of course, that lowering amyloid plaques will lead to a better clinical course. But there's no evidence of that. So, you can get excited, of course, and people could feel excited. But in order for my colleagues to honestly be excited, they have to be agreeable to this. They have to believe that the reduction of amyloid plaques does predict cognitive improvement, or they have to believe that the clinical trial showed cognitive improvement. Notice that the FDA doesn't even stay on the clinical trials showing improvement.
Robert Howard, MD, MRCPsych: I don't think it's good for patients. I mean, first of all, it doesn't work, and secondly, it is not a drug I’d really want them to take. We don't know what the long-term effects of those MRI hyperintensities and those hemorrhages in the brain are, but they can't be good. If you'd asked me 5 or 10 years ago, “How do you think you'd be feeling on the day that the first Alzheimer drug got approved?” I would have been walking on air. I’d thought I’d be punching the air; I’d be so happy. I actually just feel a little bit sick on this news, I'm afraid.
Marwan Sabbagh, MD: There are a couple of thoughts about that that are worthy of consideration. Number one is that I think this is going to energize the field. At the macro level, I think there'll be huge investments. If you look at, contextually, [other agents] could have similar evidence of efficacy. I think people are going to kind of warm up to the idea that monoclonal antibodies have some role. The second is that this is going to change how we practice medicine. You're going to now see biomarker confirmation with in vivo biomarker diagnostic CSF and PET. Everything that's been around but never been used, except for by the handful of people like me. We've never used these routinely, and I think this is going to force our hand to use them.
Fundamentally people have to understand that these drugs aren’t making you better, they just make you less worse. I think we’re going to have a real reckoning on what this drug does, how it works, how we manage it, how it affects our clinical practice, how do we get patients infused, et cetera. It's funny, we should have these conversations. I know companies have been thinking about this for years—they're not worried about whether drugs are going to be approved. They're worried about the fact that doctors cannot confirm the patient has amyloid, cannot confirm the clinical diagnosis, and that nobody's going to get these kinds of drugs because nobody's capable or experienced.
We're going to have many conversations about this. I think this portends well for a lot of people. But I also think that unintended consequences will slow clinical research because the patients will prefer to get an approved drug over a research study drug.
Anton P. Porsteinsson, MD: First and foremost, I want to see the other amyloid-lowering agents that appear to have a similar profile of efficacy potential move forward. Some people have said, “Well, if they are also well-established as amyloid lowering agents, could they be approved on a provisional basis?” I don't know where we're at with that. But I would love for those drugs to have undisputable data sets. By that, I don't mean that we can't have a discussion about the response and the effect size and all of that, but that there is no premature stopping. That the studies ran their course and that they were executed without flaw, that we can then have confidence that the data sets that are submitted don't have some sort of execution flaw, and then the efficacy data will have to stand on their own. Hopefully, we'll have 2 data sets that support each other where both show a somewhat similar degree of benefits.
Beyond that, I hope that we continue to expand the treatment targets beyond amyloid. Amyloid is clearly a potential therapeutic avenue. Now that we know that and we have an approval, maybe we can start to expand on quote, what is available. But I hope that we will see better, simpler, safer treatment options for targeting amyloid plaques and amyloid oligomers. I hope that we will come up with products that make sense in terms of reducing tau tangles. We are starting to develop drugs that target neuroinflammation, and there are multiple other pathways that are being explored. I hope that we continue to see us branch beyond amyloid and that the approval of aducanumab will not lead somehow to a concentration of efforts on amyloid—[I hope that] the broadening of potential targets will continue and even proliferate.
Lon Schneider, MD, MS: On the one hand, it's a stunner that [the FDA] would do this. But on another, looking back on it, this had been their evaluation all along. That is, this is the basis for what they're going to approve on because that's also how they asked the questions. There was substantial political pressure, and this is what it is. Now, the devil is in the details. How exactly is this going to be launched? And how exactly is it going to be provided and paid for? And then to whom will this be given?
This Stage 3 and Stage 4 conceptualization is the FDA’s conceptualization in their 2018 guidance for early-stage Alzheimer disease. So again, they're using their own language to help define this. The study itself didn't say we're enrolling Stage 3 and Stage 4 people. The clinical trial said we're taking people with mild cognitive impairment due to Alzheimer disease, or mild dementia due to Alzheimer disease. In one sense, it's totally remarkable. In another way, it was in retrospect, of course, totally predictable.
Robert Howard, MD, MRCPsych: It is a very murky world, now, with FDA approval. Whereas 10 to 15 years ago, if the FDA approved something, you could be confident that was a signal of efficacy and effectiveness. I don't think that's the case anymore. I honestly think this is going to set the field back. I've already seen some of my colleagues saying this is great, this is going to really encourage drug companies, it's going to push drug development. It's not. All it's going to do is push drug companies to set the bar terribly low. All the drugs like solanezumab, which failed a few years ago, are just as good as aducanumab in terms of being able to clear amyloid. So, I just expect we're now going to see the approval process and clinical trials gummed up with drugs that are already failed. But if you use the aducanumab standard, you might succeed.
Transcript edited for clarity.