Leeza Osipenko:  55 million people suffer from dementia worldwide based on WHO estimates, and 60 to 70% of them suffer from Alzheimer's disease. And the current definition of Alzheimer's disease or characterization of Alzheimer's disease is the presence of beta-amyloid plaques in the brain. Right now, over 4,000 clinical trials in Alzheimer's disease are found in the registries. These might be terminated trials, ongoing trials, completed trials, and right now about 120 new potential medicines are in development.

Unfortunately to date, every phase 3 trial studying a disease modifying drug has failed, and it's been incredibly disappointing for all parties involved--for the research community, for basic researchers who are questioning the actual models for targeting the disease, to the participants of the trials, to clinicians but most importantly, to patients and carers. This is an extremely important topic which, in one way or another, will concern all of us through our lives. However, over the last few years, the landscape has been changing and we've got some new drugs that we would want to talk about now.

So, Biogen made huge headlines a few years ago it received accelerated approval from the FDA in July of 2021, and then in December of 2021, it actually received a refusal for marketing authorization from the European Medicines Agency (EMA). So this drug is available on some terms in the United States, and we'll discuss that and it's not available in Europe. It's it was studied in two Phase 3 trials, one of which was negative and very controversial evidence package. Both studies were stopped prematurely because of interim analysis suggested they were not going to meet their endpoints, and now, fortunately, there is a phase 4 confirmatory trial, but results are far away. The second drug is Lecanemab from Eisai, which received an accelerated approval from the FDA in January 2023, and the decision on this drug is imminent from the FDA. The European Medicines Agency is expected to make the decision in the first quarter of 2024, and the results of the Phase 3 trial “CLARITY” is what we'll be talking about today as well. And finally, Donanemab, which is slightly further away, did not receive accelerated approval from the FDA, but it may enter the market soon as well. Its results are based on the Phase 3 Trailblazer trial (TRAILBLAZER-ALZ-2).

So today we put together an exceptional panel for you. We have Professor Aaron Kesselheim from a Harvard Medical School, we have Professor Frank Harrell, biostatistician from Vanderbilt University, we have Professor Peter Whitehouse, who is a clinician from Case Western Reserve University, and we have Dr. Richard Oakley, who is joining us from the UK Alzheimer's Society, and I will be chairing today's event. So, without further delay, I would like to now invite Professor Aaron Kesselheim to introduce himself.

Aaron Kesselheim: Hi! Well, thanks for inviting me to this panel. My name's Aaron Kesselheim. I'm a primary care doctor, a lawyer, and a Professor of Medicine at Brigham and Women's Hospital, where I run a research group called the Program on Regulation Therapeutics and Law (PORTAL), and I’m also a member a faculty member in the Division of Pharmacoepidemiology and Pharmacoeconomics, which is a division within the Department of Medicine at Brigham and Women's Hospital.

So, I care for patients with Alzheimer's disease in my primary care practice, and I have been involved in this field as a member of the FDA Peripheral and Central Nervous System Advisory Committee that reviewed Aducanamab, and I was a member of it from 2015 until 2021, when I resigned in the wake of the FDA approval decision, because I felt like it was the wrong decision, and that the process, (the Advisory Committee process) was not being used optimally in that and other circumstances that I'd observed, so I resigned. So I I'm glad to be a part of this conversation today, and I look forward to your questions.

Frank Harrell: I'm Frank Harrell. I'm a professor of Biostatistics at Vanderbilt University, which is in Nashville, Tennessee. I also work part of my time with the FDA Center for Drug Evaluation and Research, but I'm coming today strictly in my university role. I have a long-time interest in clinical trials, design and analysis and interpretation, and I'm very interested in clinical outcome scales, and their clinical interpretation, and making outcomes scales more clinically relevant and I have several concerns about that, and I think I may bring a fresh perspective, because I'm unburdened by any experience with Alzheimer’s clinical trials. My work has been in cardiovascular and critical care medicine primarily, but I'm very glad to be here today.

Peter Whitehouse: Sorry to be off camera but delighted to be with you. I'm a geriatric neurologist and I have been practicing for over 30 years, taking care of people with these kinds of conditions. Although I'm working more in public health because I think we have some really hard social decisions to make about where we invest in brain health. I was involved on the FDA panel when the first round of drugs were approved, Cholinesterase inhibitors some several decades ago, and was involved in the development of some of the instruments that we use today to still assess patients and try to determine whether these drugs are efficacious. So thank you very much again for the invitation, I look forward to the questions as well.

Richard Oakley: I'm Richard Oakley, I'm the Associate Director of Research Information at Alzheimer’s Society, so that's the biggest charity in the UK focused on dementia. So we fund a portfolio of research trying to help academics and clinicians fund research but also try to help small startups who have good products but maybe need access to people affected by dementia to make sure their product is really focused on the needs of people with dementia.

So we have a group of about 16,000 people living with dementia and carers, who we kind of partner up with academics in order to help their research develop and we also have a huge influencing arm where we try to work with the UK government to try and make sure they're listening to the voice of people affected by dementia to make sure our policies are relevant and are focused on the people at the center of everything we're going to be talking about today.

Leeza Osipenko: To kick off the discussion, I would like to invite our experts to share their personal opinions on the current situation with these new Alzheimer's drugs coming to the market? What do you make out of the headlines? What is your personal opinion on potential of these drugs being used in the market? And, as you see our audience is, our panel is, very much US-based, but I’m also very happy to have Richard from the UK and we're happy to hear questions across the board not just focused on the US perspective. 

Aaron Kesselheim: So I mean my opinion, which I expressed also at the (FDA) Advisory Committee, for Aducanumab is that there did not appear to be good evidence that Aducanumab had efficacy. There was a large trial that was conducted and stopped at the interim analysis point. At that point, the 2 trials making up that combination review were then separated, and one of them appeared to be positive, and in the high dose arm, the other was negative, and so there was conflicting evidence at best, that that drug had efficacy, and in that circumstance it makes sense to continue to try to test it to determine if the efficacy was real and not to rely on post-hoc analyses of the trial to try to make a clear determination about the drug. In addition to that, of course, there was the very substantial risks associated with the drug; upwards of 35 to 40% of patients had brain swelling with bleeding, which was well controlled in the context of the clinical trial, but I think that there was a lot of concern at the Advisory Committee that outside of the context of the very closely monitored clinical trial setting that that risk would have presented a very substantial population level risk to patients and so on in the context, of unclear evidence of the drug works and clear evidence of a very substantial risk, it didn't make sense to me to approve the drug, and it certainly didn't make sense to price the drug initially at $56,000 a year. Which, if even 10% of patients on Medicare in the US would have gotten the drug, we would have spent more money on it than we spend on the entire budget of NASA for just that one drug. The manufacturer recognized that craziness after a little while and cut the price in half, but really at any price, a drug that doesn't work, and has substantial risks is not a good idea.

Lecanemab is a slightly different story, Lecanemab does have a trial that is positive, although it is very marginally positive. The amount of difference after 18 months of therapy on the clinical scale that was used is very meager. The risks of the Lecanemab are less than the risks of Aducanumab, the rate of area of brain swelling and bleeding is less so the risks are slightly lower. And in that context the FDA thought it was appropriate to do accelerated approval, and it looks like the FDA is going to give the drug full approval in in July. That drug is also priced at a very high level, and it will be incumbent on physicians and patients at that point to have clear conversations about what people can actually expect on this drug, particularly since these are otherwise patients with very mild cognitive impairment who might otherwise still be very independent. And we just don't know a lot of things about Lecanemab, about the effect on things that matter to patients like institutionalization or independence. We don't know if this very small effect, after 18 months of therapy, is the most effect you'll ever see, or if that effect continues to progress, there's still a lot of questions about this drug, but it does look like it is on more sure footing clinically than Aducanumab was, and for those reasons I think that it is why most people believe that the drug will get approved in the summer.

Leeza Osipenko: I'd like to give word to Peter now to either reflect what Aaron has said, or just share your perspective as a clinician, and perhaps address a very specific question of the population that all 3 drugs are meant to target. And the comment that Aaron made about these being relatively fit people with early signs of Alzheimer's and potential side effects for them of these drugs.

Peter Whitehouse: So as usual, I agree with what Aaron said, and this time, and else wise and congratulate him for his courage. He actually participated with me in a panel of experts who suggested that the FDA actually withdraw Aducanumab, because the lessons we learned from that was the malignancy of the politics associated with these compounds. The fact that people exaggerated and distorted data, and this involved Congressional criticism, so it was a debacle, as I would refer to it. So that those politics and those conflicts of interest still exist.

Let me say a couple of different things. I don't believe the amyloid hypothesis actually is a hypothesis. It's more of a political statement. Rudolf Virchow, who described amyloid back in the late 1800s, essentially, then went to social medicine and said that medicine is just politics at a different scale.

Naito, the president of Eisai, who I've had the personal opportunity to meet, said his drug proved the amyloid hypothesis, because the question is, what is the amyloid hypothesis? What is the proof? Did the approval of tacrine and cholinesterase inhibitors and Eisai’s Aricept prove the cholinergic hypothesis? No, these, these are statements that frame the conversation ways that aren't helpful. Another and I'll conclude and answer your question about mild cognitive impairment and early dementia--we also tend to misconstrue things when it comes to diagnostic labelling, the concept of mild cognitive impairment is actually quite controversial and now we have another diagnostic label, subjective cognitive decline. So this is a continuum, so is mild cognitive impairment early dementia? or is early dementia different than mild cognitive impairment? So the problem with this is that people with mild cognitive impairment by definition don't have any impairments or very mild impairments in activities of daily living. So they're really not yet affected by the disease, and it may not progress. People with MCI don't always progress to dementia.

So we have a lot of diagnostic confusion that is partially clarified by biomarkers, and we have great concerns about giving these medicines to people very early in a process that I think, is brain aging, where the side effects, as Aaron said, are significant, and I think, will be a much greater problem when it comes into greater use. So I'm a sceptical about the value for society.

Leeza Ospienko: I'd like to hear the response from Richard on the perspective of what's going on in the Alzheimer’s community with patients and carers who are desperately waiting for a solution. And currently, you can see a lot of headlines, even echoing the results of the Eisai trial, saying, if I get those 5 extra months to be like I am now, I want to have those 5 extra months and not to progress. Are we getting those 5 months, or is this some data massaging to which I'll get to Frank in a moment, and then back to Peter, but Richard, please, what would be your perspective?

Richard Oakley: Yeah, thanks so much. I mean, I think we're taking that a strong stance in the UK, we're not trying to lobby the MHRA or EMA, or NICE to make any decision, we want to give them the time and allow them to look at the data and make the decision to make the recommendations, and then we will respond accordingly. But I think what we've been talking about a lot is that the first treatments for any condition are not the ones that you'd like to hope for. We, I hope, and I believe other treatments will come along. So the question for me is, can these treatments be a steppingstone? And like I say, I have to defer to colleagues who are on the panel at MHRA.

But if we look into things like cancer, the first chemotherapies that came through had awful wave reviews. If we look at things like MS and conditions like that, the first treatment shows similar slowing in their first clinical trials into one 33% slowing up to climb. But we know in real life it was much larger. So we are hopeful, and I think absolutely, who is to say what is clinically meaningful? A part of me says that it should be an informed decision by the patients. A commission and a patient should sit down. There are absolutely risks. The risks are huge, for any treatment, if it's unmanaged and Aaron mentioned in the clinical trial it was well managed, it is very different, obviously, in the population, of course it is, but we still, with conditions like MS but in the UK, the entire diagnosis system was re-modelled to ensure that people get into and on the on Beta, and the moment in UK we have 62% of people getting a diagnosis of dementia. No timeframe. There's no accuracy, so we accept that a third of people don't even get one, and if this can be a step towards us, getting people and actual diagnosis at an early stage, I think that would be a really good one. But that is only if the evidence and the data that comes from, and if MHRA is allowed, which is the UK equivalent of the FDA, it's allowed to look at the data and the experts make that decision.

But coming from other conditions, we're in cancer we can extend life for 3 months or 6 months, we celebrate that. I would like a little bit more talk around the positives and the progress we have made where there's been nothing for 20 years, and in 6 months we had two, we think, positive trials, but we await those conditions, that has to be seen as a sign of progress, and we want to talk to our politicians around how they should get involved, because they have always told us why there's nothing you can do. And I'm just really worried that we carry on that narrative and actually, the narrative is we can do something, we are doing things, we are making progress, and I think that's a good positive message. But we have to make sure it's the right treatment at the right time, and that's why we have to defer to my colleagues to make that decision and tell us if that's right.

Leeza Osipenko: I really liked your comparison comment to cancer. How getting extra 8 weeks of progression, pre-survival, or with, if you’re lucky your overall survival, gets the registration of the drug at the FDA at the 100k price tag per year and is a very, very interesting parallel. 

Frank Harrell: Couple of random comments just speaking about Aduhelm. If you look at all of the spheres in which people are looking at treatments and medical procedures and business, only with the regulated medical environment where you have a two trial rule, do people continue to try to look at 2 trials separately and say did they pass a certain bar? And I think in every other field the common-sense approach, would be to say, if you have 2 trials that are almost identical to each other, and the 2 trials used to support Aduhelm were uncannily similar to one another, that the best estimate of the drugs effect is to average the 2 trials, so if FDA wasn't involved I'm sure most anyone would do that, and their interpretation of the 2. These are averageable trials, cause they're almost interchangeable. And so the result from one trial probably overestimated the efficacy, and the other one probably underestimated. But I think the average, and the confidence interval that goes with that would be very reasonable. 

The other comment, much different from that is that we have lots of experience in other fields, such as in medical rehabilitation and spinal surgery. So in spine surgery, where I've had a lot of research experience, there's a lot of great data collected about what matters to patients. So patients with back injuries, they have very serious issues. It's very debilitating, but if you ask patients what matters to them in terms of the success of their treatment, one year after spinal surgery, what matters to them is where they end up, how they got there is completely irrelevant to patients. So they don't care about trajectories; they care about the final result a year after surgery, and they interpret that, independently of the starting point of their amount of debilitation. So, if that carries over the Alzheimer's research, then what I'm left with is a great disappointment, in what I've seen so far, and say the papers that you send around Leeza, in that Alzheimer's research failed to correct the problems that are so common in clinical trials with patient outcomes scales in which it's taken for granted that the change from baseline is a relevant measure that captures what's important to patients. And I'd go so far to say that the measurement scale that's used as a primary outcome of these trials it seems to be well-established and validated, but the change on baseline I would be very surprised if that's ever been validated to capture the right thing, and I'm pretty sure you can invalidate it with not much data, because it has floor and ceiling effects. So things that have floor and ceiling effects, it's a restricted range from 0 to 18 when you have a foreign ceiling effect, they change from baseline means something different, depending on where you started. So you can't really interpret the result clinically in any meaningful way. It's something that could have been corrected early on in Alzheimer’s disease research, but the clinical trial designers chose to follow tradition where this approach has been used in countless other medical fields and treatments unfortunately, and so I'm disappointed that that problem was not corrected. And so the result is the clinical interpretation of what's the value to patients is to me very, very cloudy. Talking about a mean change from baseline, when you cannot interpret that change without knowing what the baseline was, may mean in effect, that that mean change from baseline only applies to patients who started at the mean on baseline, not in general, and so there's a lot of issues where the analysis and the interpretation of these trialss could be way more clinically relevant to patients than they have been analyzed to date.

Leeza Ospienko: Thank you so much, Frank, and I'm looking forward to the questions from the audience, so feel free to raise your hand to ask your question, and the while we are waiting for your questions, I think we started this discussion already, but I'd like to share the 3 slides summarizing the actual results of these trials, and what they mean on the talking through these results and their meaning on the scale would be an important next step before we receive your further questions. So if we start with the Aducanumab, I invite either Peter or Frank to comment on what it means on the results of the EMERGE trial of the difference of 0.39 and the 22% decrease.

Peter Whitehouse: Why don't I mentioned the clinical scale itself, as the clinician, and ask Frank to comment about the statistics? This is a scale that's been used for a long time. It basically has 3 ratings, 0, questionable or no evidence. One mild dementia, to moderate and severe. So it's based on 0 to 3. But then there are 6 domains in which the clinician is asked to evaluate other patients doing in that arena. So, 6 by 3 gives you the 0 to 18 that you're seeing on the slide. I would say to you that after the introduction of the CDR as it's called, 0.5 became mild cognitive impairment, so they kind of inserted the condition that we're that I mentioned is somewhat controversial as 0.5. The issue, that, I think is relevant is, how do you take a score on this and convert that into months of additional life? From my mind, those claims are really unsubstantiated, and exaggerations beyond what the data allows us to really draw. And, as Aaron said, we don't really know what happens after 18 months, and I would like to say what last thing before I ask Frank to come on the statistics, Richard, I think we can do things for people with dementia, with mild cognitive impairment. There is dementia friends, there's art, there's music, there's dance, there is prevention for the decline by exercising and helping. So it's just not a case of we don't have a drug we do have a drug, it's a question of “what's the relative value of any drug compared to all the other good things we can do for people with dementia?” in my opinion. So I don't like to say that's not treatable. It's not we may not have the drugs we want, and I certainly agree with you there! But there are other social approaches that are equally import, or, more important perhaps than these drugs that have such a minimal effect as Frank's already mentioned. So, Frank, do you want to elaborate on that or over to you?

Frank Harrell: I can elaborate on part of it, I think, a scale that such as this one that has a 0 to 18 or 0 point 5 to 18 range. When you look at the components of scale, it seems to be quite reasonable, and seems to have a lot of face validity. The problem is, the scale is not being analyzed the way it was derived, it’s just being tortured, using subtraction. And so this is the sort of scale if you analyze it statistically, you'll find a subtraction doesn't actually work in this context.

So just, numerical subtraction of a scale that's not blood pressure. An interval scale measurement like blood pressure, you can actually subtract. But a scale like this doesn't work with subtraction. So you're actually making it very hard to interpret and violating a bunch of statistical assumptions which no one seems to have checked in any of the analyses that have been published, such as a normality assumption.

Once you take the scale and put it into intervals, you're taking a scale that has some promise, and then immediately ruining it by labelling things as mild, moderate, or severe. I think that's a very dangerous practice. It's very misleading. It loses a lot of information. It's not. It's not respectful of the scale, and how the scale was created. So those categories just get in the way of understanding.

And then there's one final problem with the research done today. Well, when you talk about 36% slowing of decline, I don't know that that means anything, and I don't think percentages work well, percent changes, work well on this sort of scale.

But the other problem, and this is a very subtle one, is we've already talked about side effects it's just brain swelling-when you ask, how would you do an analysis that incorporates a benefit risk trade-off, which is all important to do, how do you do that when the scale is a change from baseline? There's no way to equate a certain change from baseline with a certain amount of brain swelling. But there is a hope of equating a certain absolute outcome of the patient on a 0 to 18 scale, with brain swelling. So, if you respect the scale, not only do you get more clinical relevance, but you're able to do a risk-benefit trade-off to get a net clinical benefit analysis that would penalize for brain swelling over time and so the difference between 2 ordinal scales is no longer ordinal, and you cannot equate it with certain side effects. Whereas the absolute scale is something you could base a comprehensive risk-benefit analysis on. So I think that I think I've covered enough points there. Just anytime people subtract baseline, it's usually highly, highly problematic. And I'm disappointed the FDA is still allowing that to be done.

Leeza Osipenko: You have a excellent question from Jack Scannell, and Richard I'll give you a chance as soon as I read Jack's question, because it's completely related to what Frank just described. So (attendee question) “Frank’s endpoint, scale, and clinical relevance, comments strike me as a relevant across a wide range of diseases. Other therapy areas where improved disease measures have made a real difference who has the right incentives to break the inertia here? And I can understand why companies are reluctant to take risks with new scales.”

Frank, any comments?

Frank Harrell: I don't have any direct examples, but we have learned in amazing amount with COVID-19 where the patients it's just short term follow up, unlike Alzheimer's, but a lot of the issues come to the fore, and what we've been able to show is, if you have a disease severity measure, such as the WHO COVID-19 outcome scale, and if you treat that as longitudinal and ordinal, you get so much more power than using traditional analysis, which is usually based on an oversimplified idea of time to recovery.

So just trying to estimate the time to recovery for each patient and using that one number it's a different issue than what we're talking about here. But the power of that compared to the power of using all the raw data, which means you get your hands on all of the longitudinal measurements, however often the patient was measured on the outcome scale, you use all of those measures, and you don't subtract baseline.

So in the COVID sphere we have lots of experience and lots of primary analyses based on optimal statistical analysis and have been able to show you gain a lot of power and clinical interpretation, such as getting what is the mean amount of time that a patient is in a certain state. So a state could be you're at a certain cognitive level or better, and that level or better, could be 18 different levels, it could be not something you do just once, by estimating the time and cognitive level X or better, using all of the raw data and not using change from baseline, you can get a time-related clinical interpretation that has also optimal statistical power which has great ramifications in terms of lowering the sample sizes that have been needed in the trials design to date.

Richard Oakley: Yeah, I just respond to the Peter’s very good point that there are absolutely things we can do to help you live better with dementia but I mean, we are not changing the effect of disease. You still, on average, die within 7 years, and in UK it takes 2 years to get diagnosis. And so that is part of the challenge we have is that it takes so long currently and there's no will in the system to change things but to DMTs would. We've seen it with Beta, we've seen it with thrombolysis and stroke. That when you look at those treatments that came through, they changed the diagnosis system and people got earlier and earlier diagnoses either to or not to get to those treatments and it changed the care landscape and then someone these treatments are we're talking about Aducanumab aside where obviously the FDA entire academic panel did not recommend it, which, as I'm saying, is different from the now on the reduced panel. If we get these through, then we are going to hopefully see you have to get that early and a more accurate diagnosis, or even get a diagnosis, and that may change the care pathway. And then, if you live longer, you'll still need more, and more care, more support.

So I think that's really important that just to mention that as part of the reason, it's not because these treatments are perfect, far from it, but we believe they will start the change in the health care system, that currently like say a third of people in the UK don't even get a diagnosis. We believe this could be changing just I mean the data you show them the CDR, I'll defer to Frank and experts to talk about that. But one stature, not holding up is that things what it meant for carers. So on different measures of quality of life, in Lecanemab, one was 56% higher for carers with people who are on this treatment. For Donanemab, is 40% for people with dementia still having conversations about current affairs, managing their finances, driving. These are some of the measures they're beginning to get in. And again, as I mentioned in MS, it was a similar slowing, but it actually a greater way more, because, as we all know, these trials around 18 months, we don't know what's gonna happen, and there seems to be a lot of assumption that they may not be as effective after 18 months we don't know that we don't know. They could be more so.

It's and we're only beginning to see some benefits for carers and people affected. So again, I think it's really important that you are showing the sum of boxes and is it clinically meaningful? Again, I think there has to be a conversation between clinician and a patient. In the same way some people choose not have chemotherapy. It's like at the end of life they can go when have another 6 months of good quality life rather than 12 months or 18 months, and that is the decision of the patient determining do they want to go through that given the risks and benefits to them, they make that informed decision. And again providing it goes through the regulatory authorities which has to be a really important caveat, that people like Aaron and others look at the data and tell us, yes or no. Once that comes through, even if it's small, I don't think it's our right to say we're not to offer it. It is our right, to say the benefits are small, the risks are large, but it may do this, and that is your choice and that is your choice as a patient if you or don't you not want to do that. I think that, combined with the changes to the healthcare system, is a really important moment in dementia.

Leeza Osipenko: There's another part of the story is the marketing of pharmaceutical company, and a very strong push by the market and procure in that promise in the ability of the patients and clinicians and very pragmatically and reasonably understand what's communicated, rather than take information from the sales reps, and actually to have this very informed discussion. So Peter, I'd like to ask you a question or comment on what Rich just said. What do you think about this joint decision-making? Where should it take place? and how should be dealt with these drugs versus the reality, especially in the US, where drug marketing might kind of surpass this step?

Peter Whitehouse: Well, certainly, I think shared decision-making is at the core of the individual doctor or nurse, and patient and family, so that is true. It is also true that the pharmaceutical industry spends an enormous amount of money in the United States and New Zealand, for example, with direct-to-consumer marketing convincing people that these interventions are much more effective than they really have been demonstrated to be scientifically, so both points that you raise are important. I think at the middle of that is what I raised before, and what Aaron also mentioned at the community in the population health level. In the UK, you're more civilized, you have NICE, which does work on pricing. We (in the US) have a very complicated system between the FDA and CMS, which I think we continue to try to change politically, because there's a social and a community in a population issue here, which is where do we want to invest our resources in keeping our brains healthy? Is the answer to Alzheimer's disease, or is the are there better approaches to the focus on public health and care than these rather small effects? Effects, that many in Europe who advise EMA, I think overly, not clinically significant, and I don't think, Richard, frankly, we can let any little things sneak through, and then let individual patients and family families to decide because that space is why we have to do science in the first place, we have to respect that relationship but we have to inform that relationship with science that really tells us this is something we, as a society need to invest in.

Leeza Osipenko: Aaron, I'd like to invite you to comment on the discussion. We just had plus on the slide, because US also has ICER, which was almost worded from NICE.

Aaron Kesselheim: Right, and I also wanna just say that I really agree with Rich's comments. I thought that they were really thoughtful about the need for our high-quality decisions and discussions between doctors and patients, as after Lecanemab, as we most people assume, will happen, gets approved, there is gonna be a lot of pressure on those conversations, and as policymakers, as people of the public health community, as people in our in our local institutions, we all need to try to take steps to try to ensure that those conversations are as informed as possible, because when you ask someone, when somebody considers taking this product, they're going to need to think about, the need to come in from multiple MRIs, the risk of infusion reactions, the possibility that every time they trip over a curb then they might need to have another MRI to see if there's swelling going on because it's hard to tell the difference between area and sort of progression of disease. So I think that there, there is a lot that people are gonna have to deal with, and so it's gonna be very important to get patients and clinicians accurate unbiased information and as you point out, Leeza, that can be a problem in the context of pharmaceutical promotion- not only the promotion direct paid to patients, but the 20 to 30 billion dollars a year that the pharmaceutical industry spends promoting their drugs to physicians in ways that might overestimate the benefits and underestimate the risk, so I think that it's going to be really important to make sure that there is fair information out there about these products to try to ensure that patients can make as accurate a decision as possible.

And then you wanted me to just mention talk about the pricing.  The pricing issue, the reason that we have the pricing issue is because in the United States we allow pharmaceutical companies to charge whatever they want for their drugs. ICER is, I like the people who run ICER, I think they do a good job, but they are a small nonprofit entity that is, not does not have any real power, and it has limited resources to undertake its analyses, and so I think that there should be an organization like ICER that operates on a national level to try to provide timely analyses of drugs and their prices that they're made available at to try to understand what the relationship is between the drug’s benefits, and the price and the impact on the budget for some of these drugs, because we definitely get to a situation where payers are challenged to provide other sorts of care by if they're spending too much money or patience, for that matter, if they're spending too much money on a drug that offers the sort of possibility of a very limited benefit.

Leeza Osipenko:. And just the follow-on question about this headline from the Medicare, that plans to pay for Alzheimer's drug that when full FDA approval. So in case like, Lecanemab gains it this summer, and where do the Lecanemab/Aducanumab fall in this? And what's going on with pricing which are presented on this slide? About 28,000 per year for Aducanumab and 26 for Lecanemab and the estimates of ICER how much they should cost per year based on the clinical trial data.

Aaron Kesselheim: Well again, I love ICER, or I should say I appreciate ICER’s work, but, like ICER has no power, and they've got this nice little estimate, and they've got a lot of good data behind it. And if you're interested in it, they publish massive appendices of all the work that they do, and they do really solid work. But that doesn't matter. What matters is in the US is the price the company sets, and Medicare also doesn't have any authority to negotiate drug prices except under the Inflation Reduction Act for drugs, after 9 to 13 years on the market for a small number, and subject to other exemptions as well.

So, we're gonna be in a situation where the price is gonna be what the company determines it wants the price to be and that's just kind of where we are right now, with drug pricing policy in the US. And so I feel like there are things that we can do to talk about that. But I don't know that that's going to change anytime soon.

Peter Whitehouse: I would add to that I mean the VA and Medicaid United States can negotiate prices and private insurance doesn't necessarily have to follow Medicare, and in Europe pricing is more involved in the regulatory environment, and I am not sure that the EMA is going to prove this drug or at least we're gonna put more conditions on it than the FDA has done.  So, I think, hey, as Aaron said, the pricing is not just a drug: it's all the clinical time, and the MRIs, and all the other things that go along with it. So I do think pricing is going to continue to be in in the equation in the United States. But basically, Aaron was quite right. Medicare doesn't have the ability to do it, to really do that.

Leeza Osipenko: A very important comment, and very interesting to watch the space and the Medicare's decision, and how they are going to deal with their budget if they follow what they have stated on June 1st. So I'd like to go back to the question that we've had from Till Bruckner for Richard.

So “Some have argued that approval of even clinically, marginally effective ad drugs helps to encourage companies to stay in the AD research field, else they might give up on looking for effective treatments, and we will never see effective drugs. What is your take on that?” and the I would be happy if others want to comment on that as well.

Richard Oakley: Yeah, I mean, I, it's a really difficult one. I work in that industry. I mean, I hope that's not the case. I hope people are looking at where is the greatest need and dementia is, I think, one of the biggest biomedical challenges we face, and I hope that people see that and want to do that. Of course, that is not the only consideration in their mind and there are others, and that may factor in if you've had no new treatment of 20 years in Europe, and that is hard to keep on persuading people to invest the money there where there are a lot easier avenues actually to get through treatments in other conditions, but I have heard that from speaking to pharmaceutical colleagues in the UK that's their decisions are helping to encourage people to do that as we are. But yeah, it's difficult for me to comment. I don't work in pharma industry, and I hope they are driven by the want to tackle this challenge.

Leeza Osipenko: Let's assume these drugs get into clinical practice or let's assume we had a drug with longer than 18 months and much more impressive improvement from the clinical perspective. How are these drugs meant to be used in clinical practice? Because from clinical trials we have data on 18 months and (do we know) if the patients should be taking these medications up until first adverse events? Until clearance of amyloid plaques in their brain or clearly documented disease progression? Do we have knowledge of that?

Peter Whiterhouse: I don't think we have enough knowledge. The analysis of the CDR, some of boxing sports suggested that there are subtypes, for example, women did not seem to respond as well as men, and so there's a whole the mesh of things we don't know. The labeling apparently, if it's approved, will allow people in with certain kinds of blood thinners, but in general the issue of should we or shouldn’t we be giving these drugs to people that are on anticoagulants, for example. So, the issue of whether you can stop these drugs if the amyloid is clear, I think it's also uncertain, it's one thing, It's clear about these drugs is that they do clear amyloid. So should that be your endpoint, or should be purely a clinical endpoint? These are all very good questions, and there will be practice guidelines that are coming out and those are best estimates from people that have used medication and trials and we need to follow those very clear carefully.

But I do think, as usual, when you bring something out of rarefied environment to trials where people are very clear, very clearly a healthy population, into older folks that have more comorbidities, the dangers of the safety issues have not been adequately addressed. I'll say one last thing which is kind of ironic, we talk about brain swelling which is ARIA, the amyloid-related imaging abnormalities, all these medications also cause brain atrophy- shrinkage of the brain out of portion to any expectations you would have around the removal of amyloid. Like you say, okay, if you remove amyloid, the brain's gonna shrink a little bit. But in actual fact, looking puzzled hat these are. But that has been the background of all these compound. What does it mean that the brain actually shrinks as the amyloid is being removed? If you, if you don't have areas, it's active. So there's a lot we don't understand about safety, in my opinion.

Aaron Kesselheim: Also, let me just sort of add on to that. And first of all, echo Dan Allendor's comment, based on our conversation earlier about whether or not drug companies are just gonna give up or researchers are just gonna give up a drug doesn't get approved for this field I also think that that really doesn't pass, the smell test because there are in the US there are 5 or 6 million people with Allzheimer's disease, this massive unmet medical need in this area. And there's a lot of public funding going into looking into new types of biomarkers and other pathways that cause Alzheimer's disease. And I think that those are going to provide insights into treatment. So I don't see this. If there are people out there that are arguing that unless we approve these drugs all innovation will stop, that seems like a relatively specious argument.

But then I also want to piggyback a little bit on what Peter said, which is that I think it's going to be really important, assuming that Lecanemab does get approved, to follow closely the experience of patients who receive the drug and that may be what CMS does in the context of covering the drug, is cover the drug on the condition that patients enroll in a registry to follow their clinical course. Of course now there are registries and there are registries, and my hope is that if that does happen, and a registry is set up, that it is a very well managed and closely maintained registry, that sort of keeps close watch on patients on their clinical course and their side effects and outcomes, and to allow the government and physicians to respond quickly to it. But that sort of follow up is going to be essential to get at some of these issues.

Peter Whitehouse: If I can, just but I just add quickly, Phase 4, that is post approval is really critical. And this is a role that CMS could play in asking for these kinds of registries to be set up. Unfortunately, with most Phase 4  studies, even when the FDA has required them, even when the post marketing data has been collected to the drug has been demonstrated to be ineffective, the FDA has a very poor track record of actually removing drugs, that they've asked for more data on haven't been able to demonstrate that their efficacious. So I think the FDA needs to be this post approval monitoring process.

Leeza Osipenko: They're good comments, very interesting comment from Jack Scannell in the chat. And while you're reading it, I'd like to ask our panel the final question mostly targeted to Frank and to Peter, but others obviously feel free to comment. All 3 drugs had quite similar clinical trials, and their design. Are you happy with how they've been done? Besides, of course, stopping for the trials prematurely, for Aducanumab.  But another question is, what trials would you like to see? Is there any chance, maybe it's a question for Aaron, that we can see further controlled trials if the drug gains full approval? Or are we likely to rely on real world evidence in the future? So Frank, perhaps you can start.

Frank Harrell: Yeah, I'm not happy with the lack of respect for the raw data that's present in all of the analyses that I've seen. So I think the analyses could be a lot better, and I think the p-values that are reported in the papers are not necessarily very accurate. As a result of having data that doesn't meet the assumptions that are used in the analyses. In terms of design, I think the designs are pretty good, except there's a very common problem in clinical trials that people treat the sample size as a constant, and if you were to say which thing that as a design parameter, needs to be treated as a random variable because of tremendous uncertainty of the outcome rates and standard deviations, the sample size is what should be treated as a random variable. We pretend that we can calculate a sample size and then we aim to complete that number of patients enrolled, whether it provides enough evidence or provides too much evidence. We have to move past that to more sequential and adaptive trials for efficiency and for more honesty in what the sample size ultimately needs to be.

Peter Whitehouse: Well, what is interesting about some of these drugs is some secondary outcome. measures were also positive. But I don't think we've looked at the relationships within the outcome measures and the amyloid imaging for people who reduce their amyloid more, do they get more clinical benefit? A critical question. My own focus on outcome measures has been activities of daily living and quality of life. And then I sometimes say, the patients and families don't really care whether the doctor thinks the patient's better. The question is, what is the function? Richard mentioned this, and there was some data. What's the functional improvements? And (is there) less decline in functional disabilities? And how does that affect the quality of life with the patient? The family? That's what I would like to see more.

Leeza Osipenko: Any final comments from Aaron or Richard?

Aaron Kesselheim: Yeah, suggested something about whether we'd like we could see additional trials and I think that would be ideal, like if we had a health care system that would support that. And if the drug would be approved and people would continue to be enrolled in clinical trials, that would be ideal. And that I think actually to go back to the Covid epidemic as sort of early days, the UK health system was great in helping identify some early interventions for Covid by virtue of having a system that could support the sort of practical clinical trials in which people would be enrolled as they were presenting into the health system. We don't really have that in the US. The signals that the FDA is giving off, is suggesting that they're not going to require any additional or sorry that CMS is giving off, is that they're not going to require those kinds of additional trials that they're going to require a registry, which is why I said it's going to be really important for that registry to be super-duper, well organized, and we'll have to see.

Richard Oakley: Be really quick just to back. I mean, Aaron's point around registry is so important. That’s the only way we're gonna know what's happening and improve in second generation. You mentioned about improvement on trials. The vast majority of these trials were White, and that is not the populations at wide, so are there going to be variants? We know Peter mentioned, there may be difference in how women responded. Would there be differences in terms of yeah, ethnicity as well? I think we've got data show. That's probably going to be the case sometimes. So we didn't get audiences or participants that were reflective of our societies. And I think that's something we need to make sure starts happening in the future.