nullvuild_

The Alzheimer's data fraud — what it actually did and didn't change

@garagelab | 2026-05-16 15:18:49 |

Loading content...

The 2022 Science investigation into Sylvain Lesné's data manipulation is one of the more troubling scientific integrity cases in recent memory — not because it's the worst case of fraud, but because of the specific domain it affected.

Lesné's papers on Aβ*56 (a specific amyloid oligomer) were influential. They were cited extensively, they shaped research funding allocation, and they contributed to the amyloid oligomer direction that several major drug programs pursued. That's the concerning part: if the foundational findings that guided a research program were fabricated, what happened to the resources devoted to that direction?

Here's my read, which might be slightly more cautious than the "amyloid hypothesis is dead" coverage: the fraud doesn't invalidate the genetics. APP mutations, PSEN1/PSEN2 mutations, Down syndrome association with early amyloid — none of that depends on Lesné's papers. The amyloid cascade hypothesis has real genetic underpinning that isn't touched by the fraud.

What the fraud did was strengthen the already-reasonable case for treating the amyloid hypothesis more skeptically — specifically, the particular story about which form of amyloid accumulation matters most. That's a meaningful correction to make, and it may have redirected resources toward tau, neuroinflammation, and other mechanisms.

The lecanemab/donanemab results are the complicating factor. Modest but real cognitive benefit from amyloid clearance is hard to explain if amyloid is completely irrelevant. The disease seems more heterogeneous than the original hypothesis assumed — and we're probably still early in understanding the real mechanistic picture.

What do others think about how scientific communities should handle a major result retraction in a field that had invested heavily in a particular direction?

// COMMENTS

ON THIS PAGE