Decades of Research Misconduct Stalled an Alzheimer's Cure

Scientists like Sylvain Lesné whose alleged misconduct cost us over a decade investigating a dead end hypothesis should face substantial prison time. Considerable people have faced the horror of this disease with no hope for a cure in sight because charlatans like this likely manipulated data in the name of making their own careers.

This is my field. There have been three major scandals in the "amyloid" field -- Sylvain Lesne, Eliezer Masliah, and lecanemab trials not informing patients they had the APOE4 variant, which is associated with cerebral amyloid angiopathy, a vascular condition that the same scientists who led those trials previously noted was correlated with cerebral hemorrhage side effects. Sylvain Lesne produced shitty research that was not that highly cited, below is my take on the field and my concerns regarding journalists poorly communicating the science in this story. My perspective on why we are behind other fields is in paragraph five. The next three paragraphs are just context as to why I think there is no singular “amyloid hypothesis,” and why this kind of journalism threatens our field despite a desperate need for dealing with fraud, too. I realize this is long, I am not a journalist, I am not a good communicator, I am a scientist. If anyone has advice, please share it with me, likewise with questions.

Not being clear about "amyloid" nomenclature threatens to throw the baby out with the bathwater, which will stall an Alzheimer's cure even more. Most proteins are "globular," think kind of round balls of scrunched up string, arranged in alpha-helices and some beta-sheets. "Amyloids" are spine-like fibrillar protein aggregates, where each vertebrate is a flattened version of a protein folded in beta-strands. The vertebrate is created by these beta-strands stacking into beta-sheets. Google "amyloid fiber" versus "globular protein" to see what I mean by this description.

"Amyloid BETA" is a usually disordered protein which can aggregate into ONE of the amyloid fibrils seen in Alzheimer's and other dementia patients' brains. Tau can form amyloid fibers, so can TDP-43, TMEM, alpha-synuclein, etc. This is a good link -- https://people.mbi.ucla.edu/sawaya/amyloidatlas/ -- if you want to see the cross-sections of all of the amyloid fiber structures the field has solved with useful annotations. One "amyloid hypothesis," for example, is that TAU hyperphosphorylation (the addition of a lot of very negatively charged post-translational modifications, think chemical ornaments you can add to a tree) leads to tau amyloid fibers, which then lead to amyloid-beta amyloid fibers. There is lots of speculation about the mechanism, whether amyloid fibers can also have enzymatic function that lead to metabolic dysregulation, whether certain amyloid fibers are actually functional and exist in everyone but that certain types or a certain amount is associated with disease, the catecholamine hypothesis is something that can’t be discounted, maybe amyloid fibers are just a downstream effect of a true ultimate cause (in which case, amyloid fibers are still an important clue, lots of people are doing experiments now where they see how they can get certain amyloid fibers in vitro using co-factors which may be one step back to the root cause).

Another "amyloid hypothesis" is that amyloid beta OLIGOMERs, some kind of non-fibrillar aggregate that we don't know the structure of but that we know contains proteins that usually also make amyloid fibers , causes Alzheimer's. This is what "amyloid beta *56" is, by the way, an oligomer, and what Lesne’s work argued. We find oligomers to be extremely hard to work with and I could write a paragraph or two about why, but the fact that this is true makes OLIGOMER research, some of which is probably legitimate, an easy target for fraud. When molecules are well-known to be extremely difficult to work with, if you try to replicate someone's experiments and you can't, it could be because the molecules are extremely difficult to work with or because the authors whose experiments you tried to replicate committed fraud. It’s easy to think “well, it must be me,” which is how people get away with it for so long.

So, why are we so far behind? Something important to note is that amyloid fiber structure on an atomic level has only recently been cracked. Consider that we've known the structure of DNA since the 50s, but we didn't know the structure of the MOST common amyloid fiber specific to Alzheimer's patients until 20 freaking 16 -- https://pubmed.ncbi.nlm.nih.gov/28678775/. The reason for this is that the method used to solve the structure of most proteins is x-ray crystallography, but nobody has ever successfully crystallized amyloid fibers except for very small fragments of them -- https://pubmed.ncbi.nlm.nih.gov/15944695/. People think this makes physical sense for reasons related to crystal symmetry. The 2016 structure was solved using cryo-EM, which is a relatively recent development and only won the Nobel in 2017. Prior, it was derided as “blobology” because you would get very coarse structures from cryo-EM -- https://pmc.ncbi.nlm.nih.gov/articles/PMC2726924/. So, the field had to wait for cryo-EM to improve, a big part of which was waiting for better computer vision algorithms (i.e. we use YOLO and various CNN-based algorithms, too, we had to wait for that shit just like Waymo), among other things.

Why is structure so important? Drugs have to physically interact with some kind of protein target. So, you should probably know the atomic model of the target so you can make use of computational modeling techniques that can help you figure out what binds to proteins. Alternatively, if you want to do a ton of biochemical screens, i.e. make a bunch of the target protein, treat it with various compounds, and see what sticks, you have to make sure you are correctly making the target protein. In other words, you don’t know you’ve successfully reproduced the target if you don’t know what the target is. And the exact fiber structure matters, since different structures appear to be correlated with different diseases. Then, you have to figure out a method to make that protein, which people also have done recently for various types. The alternative is waiting for brains from brain banks, which is very slow and doesn’t provide a lot of material. Keep in mind this also means you can’t verify your mouse models have the same structures, too.

I do not think these scandals significantly stalled an Alzheimer's cure. It's a genuinely tough field -- you can't do brain biopsies, amyloids are a tricky protein to work with, we didn’t know the structure until 9 years ago. I think there are a lot of people who feel like they were locked out of funding opportunities because of the focus on amyloid-beta. Maybe this is true, I don’t know. The lab I’m in has worked on tau for decades.

Another point -- neurodegeneration starts far before symptoms show up. A lot of the recent drugs were designed for what is basically metastatic cancer. Learning more about earlier stages of neurodegeneration, which we can do with PET-ligands designed to bind to fibers, the recent p-tau blood test, etc. is necessary to uh... treat stage 1 or 2 cancer equivalent.

Finally, I have to get political here, considering recent events. I’m not a professional communicator. Most scientists aren’t. It is things like this that are the reason we are being threatened with funding cuts. Who will read nine paragraphs that I think are necessary context for all of this? But it is much easier to read an article like this, where the same point is repeated multiple times, with some “they said this,” “they said that,” etc. than to understand even a small portion of a field. More people will do the former, and then apparently call for executions without any way to judge who will be executed. And I sit down to write code for my experiments, click on one link, and see what I perceive as harmful information, and there goes apparently half an hour. I can either let this kind of stuff lead to my funding being cut, or reply to it and slow down my research.

I work in neurotech/sleeptech and work with Alzheimer's researchers.

This is sensationalist and one-sided.

Alzheimer's is the most common form of dementia, and it is not clearly understood. It is potentially the most common form because it is not understood and there are many, myself among them, who believe that multiple different diseases are being assigned a single title because we don't understand the disease enough.

This isn't to say the Amyloid hypothesis is correct, it may be correct in a number of cases, but the Type 3 Diabetes may also be correct.

The Amyloid Hypothesis has not been disproven. Yes, misrepresented data in a few studies have significantly put into question the reliability, however hundreds of other not "doctored" studies have shown the hypothesis holds significant promise.

The pharmaceutical companies did not make up the format of the disease in order to sell drugs, and the current drugs are not very effective, but that doesn't mean the theory is wrong. Just because amyloid build up is a marker of the disease does not mean that removing this metabolic waste would reverse the disease. That's like saying blowing your nose would cure the common cold.

In the case of AD, the suspicion is that the Amyloid/Tau build up damages the neurons ability to signal within the brain, like a blockage in a pipe that causes the pipe to break. Removing the blockage does not repair the pipe.

So why bother removing it at all?

A researcher we work with has posited a theory (which is fairly new, but I'm not sure she is the first) that amyloid build up reduces the power of delta slow waves, and the related glymphatic flushing of metabolic waste. This is why the drugs might be valuable (I'll say might be, rather than are), increasing the removal of waste slows further progression of the disease.

Having said that, there is a point in time where slowing progression is just lengthening time in poor health, and is not valuable, but prior to that, it can be valuable (I'll leave it up to individuals to decide what is right for them or their family members).

We've been developing slow-wave enhancement technology at Affectable Sleep [1] for the last 5 years, and similar tech has shown early promise in people with AD [2] as well as potential preventative in elderly people [3].

[1] https://affectablesleep.com

[2] https://pubmed.ncbi.nlm.nih.gov/39048400/

[3] https://academic.oup.com/ageing/article/52/12/afad228/750330...

FYI: This topic has been extensively discussed here over a loooong time [1].

[1] https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...

What do doctors prescribe nowadays? Same discredited meds?

How about SSRIs, for which much of the research does not provide supporting evidence? E.g. there's a 2022 meta-study by Moncrieff et al that shows there's no consistent evidence tying serotonin to depression. An earlier 2008 placebo-controlled study by Kirsch et al shows that SSRIs are barely any better than placebo.

IOW, SSRIs are useless horseshit for the vast majority of people (especially in the absence of e.g. CBT) and the theory behind them is not supported by evidence. Why are they still prescribed?

should execute whoever played along with this