Skip to Main Content

Like many people, I was shocked when the Food and Drug Administration ignored the advice of its neurological drugs advisory panel and broadly approved Biogen’s new drug, Aduhelm, even for populations never included in the clinical trials to assess the drug.

I am not a casual bystander to this controversial decision. I am a physician who has been treating people with Alzheimer’s since 1982; an early researcher into the biology of amyloid, the brain protein that Aduhelm targets; someone with a strong personal family history of dementia — I have shared the responsibility of caring for seven relatives who died from Alzheimer’s disease over the past 40 years; and have personally undergone biomarker assessment for an amyloid-lowering drug trial.


From all of these perspectives, the approval of Aduhelm was a mistake.

Slightly more than a week has passed since the FDA made its decision public. Since then, three members of the FDA’s advisory committee have resigned in protest over the decision to approve the drug for far broader use than intended and without as much as courtesy notification of its expert advisers. The FDA was not acting here as a tiebreaker: 10 of 11 advisers voted to reject Biogen’s application and the 11th was undecided on how to interpret the data.

As I formulate my responses to my own patients, here are six recurring themes:


Pull of desperation. Even the most sophisticated patients in my private practice who have enrolled in clinical trials of donanemab or lecanemab, experimental Alzheimer’s drugs that work in similar ways as Aduhelm, are unable to resist the pull of desperation potentiated by the false hope engendered by Aduhelm’s approval. They are rushing to resign from their current trials and queuing up to start Aduhelm infusions.

I mentioned to them months ago that this might happen, but I never dreamed that the approval of Aduhelm would be as indefensible as has now been revealed. There are no direct comparison trial data upon which to base a comparison of Aduhelm to donanemab or to lecanemab, so it is difficult to begrudge individuals in trials of these treatments who prefer immediate access to Aduhelm rather than the possibility of getting placebo for at least 18 months if they remain in their double-blinded trials. After those 18 months, those on placebo would likely be offered donanemab or lecanemab — assuming they work — but in the meantime, they wake up every morning asking themselves whether they can endure the possibility they are in the placebo arm.

The improper approval of Aduhelm poses an immediate and permanent threat to the discovery of real effective disease modifiers for dementia.

Elbowing aside existing treatments. The harm posed by Aduhelm’s approval to the detriment of more promising drugs is not merely theoretical. This is already happening with galantamine — an inexpensive oral drug approved to treat Alzheimer’s two decades ago — that operates on a different principle than Aduhelm.

A report from Maria Eriksdotter and colleagues in Sweden, published online in the journal Neurology in March 2021, showed that galantamine is far and away the most promising tool in the current armamentarium aimed at slowing the progression of dementia. In a study of thousands of people over many years, these researchers observed that people with Alzheimer’s dementia who took galantamine virtually never declined past moderate-stage dementia. Almost no one taking the drug developed end-stage dementia.

All three major drugs in this class, known as cholinesterase inhibitors, were associated with benefits for cognitive outcomes and death, but galantamine was singular in what looks precisely like what we are chasing when it comes to desirable properties for a drug that slows dementia progression.

The performance of galantamine over decades is far superior to the uninterpretable data package upon which the FDA relied in approving Aduhelm.

The FDA’s baseless decision has served to focus patients, families, and clinicians on how to operationalize the approval of a costly, unreimbursed, intravenous infusion drug as patients by the hundreds or thousands reach out to their doctors asking when and how can they begin receiving it. At almost the same moment, far superior data pointing to a safe, affordable, orally active, immediately available medication are almost totally ignored.

In the interest of full disclosure, galantamine was developed by a member of the board of trustees at the Icahn School of Medicine at Mount Sinai, where I work, but the drug was not developed at Mount Sinai, and I was not involved in its discovery, development, or post-market research.

Amyloid reduction alone is no help. For its approval of Aduhelm, the FDA said that reducing amyloid in the brain could be reasonably expected to benefit brain function. This is the same rationale that drove Pfizer’s development of bapineuzumab, the first anti-amyloid antibody subjected to clinical trials. Those trials were reported in The Lancet Neurology in 2010 by Juha Rinne and colleagues, and I wrote the accompanying editorial summarizing some possible next steps. Fast forward to 2021. Though bapineuzumab (which Pfizer ultimately abandoned) had reduced amyloid accumulation over the course of the trials but showed no meaningful impact on clinical outcome, here we go again: approval of the anti-amyloid antibody aducanumab, which rapidly purges the brain of detectable fibrillar amyloid and for which there is still no evidence of consistent meaningful clinical benefit.

Since then, the failure of amyloid reduction to predict meaningful clinical benefit has been noted by many experts, myself included, as my colleague Mary Sano and I wrote in Nature Reviews Neurology in 2015 when solanezumab, another amyloid-clearing drug, was under consideration for FDA approval. Today we have solanezumab déjà vu — but worse.

The FDA’s opinion that amyloid reduction should lead to meaningful clinical benefit, an opinion shared by many dementia researchers, flies in the face of a decade of failures with bapineuzumab, solanezumab, and others. It is possible that personal polygenic risk scores might point to individuals in whom some early initiation of amyloid reduction might yield meaningful benefit, but that science is not yet in hand. Moreover, current data suggest that this early age of initiation of amyloid reduction might be the third or fourth decade of life, far earlier than the Aduhelm approval anticipates.

It is possible that when mildly impaired patients are treated for five years or 10 years or more with Aduhelm, a meaningful clinical benefit will be obvious. I hope this will occur. But FDA approvals should be based on data not speculation.

Finally, there is convincing evidence that the most readily detectable form of amyloid, known as a fibril, is not the worst type. The title of most harmful form of amyloid goes to what’s known as a toxic oligomer, which is currently undetectable in the living human brain. Developing a means of establishing which oligomer types and how much of each is in the brain of a patient, and how her or his oligomer type and level responds to intervention, will be required to resolve the question of whether oligomers have important clinical correlates and, if so, which ones.

An uninterpretable trial. Biogen launched two identically designed Phase 3 trials for aducanumab, one called ENGAGE, the other EMERGE. One of them, ENGAGE, has been called a “failed” or “negative” trial. A better description is uninterpretable. The main problem with that trial was that the participants receiving placebo failed to decline as rapidly as expected.

The Alzheimer’s Disease Neuroimaging Initiative recently reported that measurements of plasma neurofilament-light chain (NfL) can predict the trajectory of decline. Biogen should report the NfL levels from its various cohorts since those measurements might enable a reanalysis that offers compelling support for Aduhelm.

Trials and labeling disconnect. There is an inexplicable disconnect between the criteria Biogen employed for its Aduhelm trials and the company’s statements about the types of patients for whom the FDA approved the drug, known as the indication or product labeling.

The aducanumab trials were restricted to participants who met these conditions: They had the mildest of symptoms; amyloid biomarkers were present at baseline; each patient’s genetic risk factors, such as APOE4, were known; and periodic brain MRIs were performed to detect the most potentially serious side effect, the brain swelling and bleeding known as amyloid-related imaging abnormalities (ARIA).

In the trials, surveillance MRIs almost always enabled early detection of ARIA before symptoms or brain bleeding appeared. In such cases, when dosing was reduced, skipped, or stopped, ARIA often resolved with no important consequences in response to care that is supportive in nature. Individuals who were taking blood thinners were excluded from the trials because they had the highest risk for major bleeds from ARIA, a potentially fatal complication. In the trials, once the red line of 10 microhemorrhages had been reached, study participants were excluded and offered no further Aduhelm infusions.

The product labeling does not recommend stopping Aduhelm after 10 microhemorrhages, but instead to proceed “cautiously after radiographic stabilization,” another instance of how the Aduhelm drug label, which is intended to educate clinicians and patients about how to use the drug safely, does not provide adequate recognition for a possible serious adverse event. Nor does it advise clinicians about exactly what steps to take after a patient has their 10th brain microhemorrhage.

None of the standard features of the Aduhelm trials have been included in the product labeling. These might soon appear among best-practice recommendations from expert panels, but as the situation stands now, this just looks like another affront to the FDA’s neurological drugs advisory panel.

Decisions that ought to be science-based are being increasingly driven by politics, wishful thinking, and even delusion. A favorite phrase used in several of the past week’s position statements has been “this is a new day in the fight against Alzheimer’s.” What is new is that the FDA has apparently applied what I can only speculate to have been bias, wishful thinking, and/or delusion as substitutes for science and data. Amazingly, major research agencies and foundations are rushing to say that they, too, see the emperor’s new clothes. The American Academy of Neurology’s Neurology Today website is the exception in having characterized the FDA’s decision as mixed at best.

Cost and equity issues. In any discussion of Aduhelm, the linked issues of cost and inequity of access to care immediately arise. Biogen has announced its intention to ship Aduhelm to 900 infusion sites over the next two weeks, sparking families to call for infusion appointments while local stakeholder committees (like the one at my hospital in Manhattan) convene urgently and add Aduhelm to their health system infusion formularies. The cost Biogen announced for Aduhelm, $56,000 a year, represents only about half of the actual yearly cost once physician, infusion center, and imaging costs are factored in.

While some seeking help at our center can afford the cost, most cannot. I hope that Biogen will make good on its promise to help patients seeking Aduhelm therapy find a way to afford it. In the absence of guidance (and in a comparable situation to that in place for amyloid imaging for several years), I must ask my patients to be sure to bring their credit cards or checkbooks when they arrive for their appointments, because prepayment out of pocket is the current standard.

Bottom line. The FDA’s approval of Aduhelm raises more questions and creates more problems than a new drug approval should. It’s time for governmental, professional, and advocacy entities to step in where Biogen and the FDA have failed and explain to patients, caregivers, and clinicians how this drug is not the “new day” in the fight against Alzheimer’s disease and needs to be approached cautiously, if at all.

Sam Gandy is a neurologist; professor of neurology and of psychiatry at the Icahn School of Medicine at Mount Sinai where he holds the Mount Sinai Chair in Alzheimer’s Research; the founding and current director of the Mount Sinai Center for Cognitive Health and NFL Neurological Care; and associate director of the National Institute on Aging-Designated Mount Sinai Alzheimer’s Disease Research Center. He reports founding companies based on immunotherapeutic strategies for dementia; receiving personal or laboratory support from Wyeth, DiaGenic, Pfizer, Baxter, and Polyphenolics/Constellation Wines; and recently joined the scientific advisory board of Ritrova Therapeutics.

Create a display name to comment

This name will appear with your comment

There was an error saving your display name. Please check and try again.