[HN Gopher] Decades of Research Misconduct Stalled an Alzheimer'...
___________________________________________________________________
Decades of Research Misconduct Stalled an Alzheimer's Cure
Author : sorokod
Score : 104 points
Date : 2025-02-23 17:46 UTC (5 hours ago)
(HTM) web link (www.sciencefriday.com)
(TXT) w3m dump (www.sciencefriday.com)
| winterbloom wrote:
| should execute whoever played along with this
| nobodyandproud wrote:
| No, but the careers and reputation of those who lead this
| avenue _and undermined_ good science must be conclusively
| ruined.
|
| We can start by letting the truth come light in stark and
| unambiguous terms again and again, along with the billions of
| dollars in damage and hundreds of thousands if not millions of
| families left to suffer thanks to prestige and fraud.
|
| Better yet, also reward the risk takers then and now; trying to
| chase good science as hazardous as it was.
| readthenotes1 wrote:
| Like the former President of Stanford? It took years for him
| to be given the golden parachute
| nobodyandproud wrote:
| If the shoe fits.
|
| It's a big scandal and I suspect researchers in businesses
| and universities be found complicit; if we have the
| appetite and headcount to chase it.
|
| I have little tolerance for fraud, and this is an example
| of why trust in science (the institution) is at an all-
| time-low.
| throwawaymaths wrote:
| good news! he's doing a vc-backed startup now
| cortesoft wrote:
| I know this is a flippant comment, but do you really think this
| would solve the problem? I feel like this would just lead to
| academic misconduct never being reported. It is hard enough to
| get someone to report a fellow academic for fraud when the only
| consequence is professional reputation; do you really think
| ANYONE would turn in a colleague when the consequence is death?
|
| Plus, people would be so scared that they will make a mistake
| that will be interpreted as fraud that no one will want to
| publish research.
| ein0p wrote:
| Maybe not "execute" but some real-world accountability would
| be nice. As in, perhaps, a lengthy prison sentence.
| mschuster91 wrote:
| 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...
| sorokod wrote:
| The book is new though.
| SirensOfTitan wrote:
| Scientists like Sylvain Lesne 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.
| voisin wrote:
| There should be some laws that were broken. Fraud in order to
| obtain research grants, fraud in order to maintain labs, etc
| etc.
| delusional wrote:
| In my mind this very much depends on who was tricked here. If
| the research was funded by companies, then arguably they
| should just have better oversight and the loss of their
| investment is already "punishment" to incentivice that better
| oversight. If it was public money, then it depends more on
| the nature of the fraud. I could concieve of a circumstance
| where I would also argue for a lapse in oversight and
| therefore a sort of "diffusion" of the responsibility.
|
| I don't think it's certain that anything illegal happened
| here. It was certainly wrong, but wrong how?
| wakawaka28 wrote:
| >If the research was funded by companies, then arguably
| they should just have better oversight and the loss of
| their investment is already "punishment" to incentivice
| that better oversight.
|
| This is not at all how the law works. You can't fool
| investors on purpose then shrug it off as "I win, they
| should have done their due diligence and second-guessed
| every aspect of my work that I intentionally falsified"
| lol. What you're arguing for is equivalent to a diffusion
| of responsibility for accounting fraud, as if the victims
| deserve it because auditors didn't detect it immediately.
|
| I don't know the whole story but deceptive practices
| (deliberate wrongdoing) need to be punished, if that is
| what happened here.
| onemoresoop wrote:
| That happens all the time in the stock market and rarely
| anyone gets punished.
| matthewdgreen wrote:
| There is an incredibly valuable and interesting comment ranked
| below yours that gives enormous amounts of context on the
| Amyloid hypothesis and the state of the field. I don't have the
| power to make that the top-ranked comment and all the "we need
| to punish" comments lower ranked. But I do wish the posters and
| other readers on HN would help to make that happen.
| RobotToaster wrote:
| They have potentially condemned hundreds of thousands of people
| to suffer with this disease unnecessarily. The level of harm
| they have caused may be far worse than anything Josef Mengele
| did.
| ein0p wrote:
| 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?
| pedalpete wrote:
| 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...
| bartathe wrote:
| 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.
| anonnon wrote:
| > 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
|
| Am I unreasonable to think that funding _ought_ to be re-
| directed elsewhere, given that we 1) already have effective
| anti-amyloid monoclonal antibodies, and 2) they don 't seem to
| work that well, and 3) there are alternatives, like the chronic
| inflammation hypothesis, that have supporting evidence? (e.g.,
| https://www.nature.com/articles/s41591-024-03201-5)
| margalabargala wrote:
| > 1) already have effective anti-amyloid monoclonal
| antibodies, and 2) they don't seem to work that well
|
| I fear you may be falling into the exact trap that the person
| you replied to, is warning against.
|
| There is not just one thing called "amyloid", so not only are
| the "anti-amyloid monoclonal antibodies" not effective
| against all amyloid, the amyloid against which they are
| effective may well not be disease-contributing amyloid.
|
| The state of the field is much more complicated that deciding
| between pop-sci summaries of "amyloid bad" and "amyloid
| irrelevant" and directing funding accordingly.
| _zoltan_ wrote:
| as someone who is not in this field (purely doing CS research):
| haven't recent advances in ML "solved" most protein structures
| and made x-ray crystallography obsolete?
| margalabargala wrote:
| Very much no.
|
| https://www.nature.com/articles/s41592-023-02087-4
| mattkrisiloff wrote:
| Would you be up to chat sometime? Would love to fund a research
| startup to work on some of this more -- matt@scifounders.com
| matthewdgreen wrote:
| I read these paragraphs. I'm incredibly grateful for the
| context and glad people are working on this. Scientists aren't
| perfect but science is all we've got when it comes to this
| shithead if a disease.
| pessimizer wrote:
| > Who will read nine paragraphs that I think are necessary
| context for all of this?
|
| They are very interesting, but I'd have to be convinced that
| they were necessary context. Because the question that I think
| is important is whether this was true and widespread:
|
| > 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.
|
| And you don't know. I don't know either.
|
| We can't say that anything would be farther along if these
| frauds had not happened, because that is a counterfactual. We
| can just guess that if we spend resources in areas that didn't
| have as much fraud surrounding them, we would be more likely to
| be farther along. Any argument otherwise is an argument that
| the direction and funding of research doesn't ever really
| matter.
|
| I don't think any avenue of research should be abandoned if
| anybody still sees any possible value in it. But I know that
| funding decisions heavily influenced by fraudulent research are
| not going to be better made than decisions not influenced by
| fraud; and that if we were making decisions based on fraud over
| a long period of time, it is safe to assume that there was a
| loss. If we want to be less likely to repeat this loss, we
| probably need to change how we evaluate where to allocate
| funding.
___________________________________________________________________
(page generated 2025-02-23 23:00 UTC)