A medical research company posted new drug trial results on Tuesday. But in the wake of Alzheimer's recent data fraud crises, have we learned how to be better brokers of supposedly groundbreaking medical intel?
In July, the medical community was rocked by a disappointing reminder of science’s weakest link: the humans doing the work. The journal Science had shared that its six-month investigation supported the findings of whistleblower Matthew Schrag, who first noted altered images in a high-impact paper on Alzheimer’s, published in Nature in 2006. That paper is still flagged on Nature as under review, but the damage has already been done. Alzheimer’s drugs for the last decade and a half have been developed around claims without as much evidence as initially believed—which might also explain why they haven’t been working, leading people to pour false hope into useless and often expensive treatment plans for declining loved ones.
This week, for the first time since that painful disclosure, we’re seeing industry reports of a new Alzheimer’s drug that some news outlets are taking at face value, even as others are noting concerning side effects. But how should we respond to these reports, in the wake of July’s difficult discoveries? This isn’t the first time science journalism has had to deal with the competing pressures of for-profit pharmaceuticals when trying to report accurately on the science itself. Have we learned anything from Alzheimer’s research setbacks?
How do we heal trust in knowledge sectors still working through such recent scandals?
Simufilam: A smoking gun for Alzheimer’s research
In June 2021, the Food and Drug Administration (FDA) approved an Alzheimer’s drug called Aduhelm (aducanumab), developed by Biogen and Eisai, despite scientific advisory panels stating that the drug did not demonstrate enough efficacy and safety to offset the cost (in both senses of the term). Three advisors quit over the FDA’s decision. American Family Physician advocated in March against its use.
But the race was now on in Big Pharm. Cassava Science, a medical science company that had only recently turned to Alzheimer’s research, had its own drug, Simufilam, that it was racing to get into the market while predicted earnings were high. But the drug had its own problematic history with scientific review, and in August 2021 a citizen’s petition to the FDA outlined concerns about the integrity and quality of data used to secure advancement through drug trials. As the legal offices of Labaton Sucharow wrote:
The studies from Drs. Wang and Burns discussed in this dossier were used by Cassava Sciences to garner NIH grants and to open an investigational new drug (IND) application to study simufilam in AD patients. They form the basic science foundation for two completed clinical trials (phase IIa and IIb) which exposed over 70 patients to simufilam. Cassava Sciences is currently recruiting 200 additional patients for a follow-up open-label trial. …
To preface the analysis that follows, no other labs have confirmed this research connecting Filamin A to pain or AD. No other labs have confirmed that simufilam binds or modifies Filamin A or has effects in AD models.”
That’s when two neuroscientists who were petitioning the FDA to halt trials pending audit called in Schrag, a neuroscientist at Vanderbilt University, to help them investigate both Simufilam and Cassava Sciences.
Altruists? Scientists hoping to get the record right?
Not entirely. The people who called in Schrag were short sellers, and stood to profit if Cassava Science’s stock dropped. And Schrag was a great choice to support their claims, because his own Alzheimer’s research countered Cassava Science’s purported findings.
(The FDA rejected the citizens’ petition this year, and more stock market participants have been identified in negative claims and reporting against Cassava. Simon Spichak of The Daily Beast offered a comprehensive run-down of the financial games involved this July.)
That’s why Science‘s corroborating investigation into Schrag’s discovery of signs of data fraud was so important, amid all this economic chicanery.
When the journal reached out to other researchers, they too recognized clear signs of image tampering in first-author Sylvain Lesné’s 2006 paper, “A specific amyloid-β protein assembly in the brain impairs memory”. Nor was this an isolated incident. The concerning research named in the citizens’ petition included two studies, from 2005 and 2010, by first- or co-authors Hoau-Yan Wang and Lindsay Burns, with more signs of data tampering. Wang is on Cassava Science’s advisory board, and Burns is part of the Alzheimer’s research team. Five other studies from Wang’s lab have also been retracted.
While this whole scandal raised significant alarm, some scientists also pushed back on the idea that it completely eradicated 16 years of Alzheimer’s research. This is because Lesné’s paper, while highly influential, was not the backbone of all research into the “amyloid plaque model”: a foundation of modern Alzheimer’s research that views the build-up of plaques made up of amyloid-beta (AB) proteins as the cause of Alzheimer’s symptoms. All it did was focus on a single oligomer (a less complex polymer) of AB proteins, called AB*56.
Some researchers therefore hold that other evidence for the amyloid plaque model still makes it a useful direction for study and the development of drug trials. However, there is no scientific consensus on amyloids as a driver of Alzheimer’s, and the failure of hundreds of drug trials following that research path lends weight to the skepticism.
Lecanemub: The Alzheimer’s drug in recent news
Enter lecanemub, a Biogen and Eisai drug that was accepted under the accelerated approval pathway by the FDA in July, and which posted Phase 3 results on Tuesday, November 29 that showed a 27 percent reduction in cognitive and functional decline at the 18-month mark for the study’s almost 1,800 participants. The 12-month mark showed no significant reduction over the placebo (although Michael Irizarry of Eisai’s research team called the 12-month results “quite predictive” of 18-month changes), so even a small success in Phase 3 is now being hailed as a promising breakthrough by mainstream news media.
(Time might have had the most ridiculous spin. Despite its fairly comprehensive write-up on the twinned approach to research taken by Biogen with aducanumab and Eisai with lecanemub, author Alice Park introduced the research as the next step in an “emotionally thrilling loop” of highs and lows for Alzheimer’s patients and their caregivers. The article treats the two medical companies’ gambles on different research pathways like a game.)
Science journals are more cautious. Nature reported on the findings under the headline “Heralded Alzheimer’s drug works—but safety concerns loom”, because there were two deaths among participants that have given researchers pause. Science reported in depth on the second on November 27, wherein a woman died of “substantial bleeding throughout her brain’s outer layer”, and the scientist who did the autopsy reported “zero doubt … that this is a treatment-caused illness and death”. This is especially concerning because the condition that made her susceptible to brain-bleed is called cerebral amyloid angiopathy (CAA)—and it’s present in almost half of Alzheimer’s patients, along with other conditions that could affect their interactions with anything that has a blood-thinning effect.
The pressing question now facing the medical research community (and the FDA) is how to balance a moderate-at-best improvement in patient outcomes with both the potentially high risks posed by lecanemub, and the paucity of other drugs offering better relief.
Other researchers, as noted by Nature, recognize that this drug trial’s results have only reached “the minimum [shift] to be clinically important”, and that even outside of adverse outcomes, around 20 percent of participants taking lecanemub showed brain-scan abnormalities indicative of swelling or bleeding. This is a step up from aducanamab’s 40 percent, but still not ideal.
The state of Alzheimer’s research and treatment outcomes remains fragile at best.
The driver for poor medical science journalism
Alzheimer’s isn’t the only branch of medical research to struggle with replication and application issues. Preclinical cancer is another field with a Herculean challenge when it comes to corroborating earlier claims. As with so much other scientific research, publication pressures prioritize exceptional breakthroughs over the follow-up studies needed to make their findings stick. It’s just more painful when the work relates to our health.
One core problem for news media on these topics is similar: the desire to report on success. Who doesn’t want to be on the frontlines of breaking word about a treatment that can fix or ease a terrible disease? Why shouldn’t we err on the side of hope?
But science as a methodology works best when everyone who participates in the process—directly as researchers, or indirectly, as people sharing word about their findings—views themself as a filter of incoming intel. This means asking questions especially about news that makes us feel good or affirmed, and approaching all results with a falsifiability mentality. What are the parameters and limits of this latest discovery? Is there a key context in which these findings are emerging? Who are the players reporting on it at all?
Pragmatism isn’t the same as cynicism. If anything, it’s more humane not to get our hopes up ahead of verified and replicated data. People with loved ones suffering from Alzheimer’s, or other degenerative diseases that force us to watch people lose more of themselves day by day, are already highly susceptible to any promise of aid—whatever the cost, whatever the risk—which can easily increase overall suffering over time.
It is our responsibility as outsiders to the messy mechanics of scientific research to call out its gamification by companies and investors. We hold the line for the scientific method when we acknowledge new intel with an attitude, above all else, of “let’s wait and see”.