And may have saved billions of dollars in Alzheimer’s research funding
Over the summer, a criminal investigation was opened against Cassava Sciences Inc., a Biotech company accused of fabricating pre-clinical data and clinical trial results. Their problems began in 2021 when two scientists, David Bredt and Geoffrey Pitt, noticed inconsistencies in the published data surrounding the billion-dollar company’s flagship product (Simufilam) claiming to miraculously reverse Alzheimer’s. These scientists were also unusually motivated: They realized that they could make a buck by taking a short position on their stock (betting on the stock price dropping) when they exposed Cassava’s fraudulent science. This led to Bredt and Pitt filing a “Citizen’s Petition,” which is a process provided by the Food and Drug Administration (FDA) for individuals and community organizations to make requests to the FDA for changes in policy, with the aim of halting any further clinical trials (tests on human patients). The Citizen’s Petition was published by the FDA, presenting evidence of data forgery in Cassava’s pre-clinical research, but was eventually denied, likely due to the suspect motives on the part of the filers. Reuters, in the article breaking news of the criminal probe, reports that “Cassava Sciences vehemently denies any and all allegations of wrongdoing.”
On the fringes of the Cassava drama was Dr. Matthew Schrag, a neuroscientist and physician at Vanderbilt University, who was paid by the two short-selling scientists’ attorneys to interrogate the publications underlying Simufilam. In the process, Schrag started to unearth a different and likely more significant case of misconduct regarding a 2006 paper published in Nature, one of the world’s most prestigious scientific journals. Schrag identified several images that appeared to be falsified in a highly cited publication which led Science, another of the world’s most prestigious journals, to launch a six-month investigation that they reported on in an article in July titled “BLOTS ON A FIELD? A neuroscience image sleuth finds signs of fabrication in scores of Alzheimer’s articles, threatening a reigning theory of the disease.”
Amyloids have a long history of being linked with Alzheimer’s, starting at the very beginning with the 1906 finding by Alois Alzheimer that plaques and other protein deposits were in the brain of a deceased dementia patient. In 1984, amyloid beta (broadly—not the specific one in question) was identified as the main component of the plaques identified in Alzheimer’s and eventually became accepted as the most plausible target for therapy. Unfortunately, the hundreds of clinical trials targeting amyloids have largely been unsuccessful, with only one underwhelming therapy (Aduhelm) gaining FDA approval.
The paper in question (“A specific amyloid-β protein assembly in the brain impairs memory,” Lesné et al., Nature 2006) came as a star of hope to amyloid-based research that was facing doubts at the time. The paper has a large part in the current focus on amyloid-based Alzheimer’s research—the National Institutes of Health (NIH; America’s medical research agency) spent about $1.6 billion on projects that mention amyloids in this fiscal year, about half its overall Alzheimer’s funding—at the expense of competing hypotheses. Unfortunately, the authors of the paper are now accused of tampering with a key piece of data upon which the methods they employed in their experiments are validated.
The experiment in question involved a genetically modified mouse that produces a human amyloid beta precursor protein, where they found that the mouse developed memory deficits in middle age and found plaque in its brain as it got old. They then found that a specific form of amyloid beta (Aβ*56) purified from the brains of impaired mice disrupts memory loss when administered to young rats. The red flags center around a handful of Western blots, which is a well-established method to detect specific proteins in a blood or tissue sample. The Western blots in question aim to detect the Aβ*56 that was central to both the conclusions drawn from these experiments and the drug action of Cassava’s Simufilam. The authors are accused of faking the detection of Aβ*56, which, if true, would render their whole experiment meaningless as they had not successfully isolated the protein. Harvard University’s Dr. Dennis Selkoe, who examined Schrag’s dossier at Science’s request, is quoted in the Science article saying: “In science, once you publish your data, if it’s not readily replicated, then there is real concern that it’s not correct or true. There’s precious little clearcut evidence that Aβ*56 exists, or if it exists, correlates in a reproducible fashion with features of Alzheimer’s—even in animal models.”
How has the science world responded? Far too slowly for Schrag’s liking—hence the report to Science. The NIH replied to Schrag’s whistleblower report that they would send the case to the Office of Research Integrity for review, followed by a process that could take years involving university investigations and a further final review. Journals contacted by Schrag regarding other suspect publications by the authors of the Aβ*56 paper have indicated that papers are under investigation. Nature published a note saying it was investigating the paper and advising caution about its results.
If he is correct in his findings, Schrag may have saved the public an enormous amount of money (in the form of NIH grants) because, as he notes, you can’t cheat science: “You can cheat to get a paper. You can cheat to get a degree. You can cheat to get a grant. You can’t cheat to cure a disease. Biology doesn’t care.”