For several years, doctors and patients have been closely following the development of the drug aducanumab, hoping that it might finally be the first medication to stop the cycle of failure that has been the fate of dozens of Alzheimer’s drug candidates that have come before. If approved, it would become the first treatment for Alzheimer’s; current medications address symptoms of the disease but do not tackle the root causes of the neurodegenerative illness. Those hopes were slowly quashed during a seven-hour meeting of a Food and Drug Administration (FDA) advisory committee last week.
The committee, made up of independent experts in Alzheimer’s and statistics, heard from Biogen, the makers of aducanumab, as well as from FDA scientists who reviewed data submitted by the company. They focused their attention on three major studies the company provided for consideration: one showing the drug was effective in patients with mild Alzheimer’s disease, a second that showed the opposite, and a third, early-stage trial that wasn’t designed to measure effectiveness but hinted at benefit. The FDA’s review of the data was divided. A clinical review described the data as “robust” and “exceptionally persuasive” when it came to the drug’s effectiveness, while the agency’s statistician was more skeptical.
But committee advisers expressed a litany of concerns about what was presented and the FDA’s interpretation of those data. In the group’s final vote, members were asked whether it was reasonable to consider the single positive study along with the suggestive findings from the early-stage trial, as evidence of the effectiveness of aducanumab for treating Alzheimer’s. Ten of the 11 members voted no, and one was uncertain. “I don’t think the evidence from [the positive study] provides substantial evidence on the efficacy or the effectiveness of this drug,” said committee member Dr. Aaron Kesselheim, associate professor of medicine at Harvard Medical School and Brigham and Women’s Hospital, who voted no.
Throughout the day, committee members raised concerns about the unusual nature of that single positive study, and the lack of a satisfactory explanation for why two large, identical studies produced opposite results. “If you have a product that clearly works, and you do two identical trials, you would expect both to be positive,” Dr. Caleb Alexander, professor of epidemiology and medicine at the Johns Hopkins Bloomberg School of Public Health and one of the committee members, told TIME after the meeting. “It’s pretty puzzling to try to figure out why both trials wouldn’t be positive in a product that works. Going into this committee, there is nothing I wanted to hear more than evidence that makes me confident that the product works. And it simply wasn’t there in this instance.”
The studies that were under debate
Aducanumab, developed by the Massachusetts-based multinational Biogen, has followed a bumpy and unprecedented road to this point. In 2016, the drug excited dementia specialists when early results from its first human studies showed that the treatment could improve cognition among people with early signs of Alzheimer’s. The drug is an antibody that was developed based on antibodies made by older people who had mild cognitive impairment but did not go on to develop Alzheimer’s. The antibody binds to clumps of amyloid protein (which build up to toxic levels in the brains of Alzheimer’s patients) and makes it easier for immune cells to then clear the clumps away. The early studies showing improvements in patients were supported by further studies, including images of the brain, that confirmed that the drug reduced the amount of amyloid plaques in the brain, as well as other research showing people on the drug improved people’s performance on brain tests.
But in March 2019, Biogen abruptly terminated two large, ongoing late-phase studies of the drug after a planned analysis of the data showed futility. In that analysis, which combined data from both studies, people receiving the drug did no better than people getting placebo when it came to amyloid in their brains or their performance on cognitive tests.
Then, several months later, after a more careful review, the company said data from one of the trials were positive, while the data from the other were negative. After consulting with the FDA, which agreed that the positive data were valid, despite the truncated study, Biogen applied for approval for treating early-stage Alzheimer’s disease. During her presentation, Samantha Budd Haeberlein, senior vice president and head of the neurodegeneration developing unit at Biogen, detailed the reasons behind the company’s conflicting studies and the about-face decision to apply for approval. The two studies, while identical, enrolled patients at different rates. Researchers knew that people with a genetic risk for Alzheimer’s disease known as the ApoE4 allele, are more likely to develop an inflammatory side effect (called ARIA) from the drug. But people who are genetically predisposed to developing Alzheimer’s are also most likely to benefit from a drug that reduces amyloid, so these people were allowed to enroll in the studies as long as they received lower doses of the drug.
Once the studies began, doctors learned that people with ApoE4 could likely tolerate a dose of aducanumab just under that which people without the genetic risk received. They decided to move everyone, including those with ApoE4, to the higher dose. However, by this time—close to the interim analysis—most of the people enrolled in one of the studies had already received the lower dose, which turned out to not be effective.
That’s why, when the company looked at preliminary results from both of the studies combined, they found little benefit of the drug over placebo. But when they delved deeper into the data and analyzed the studies separately, they saw that in the study enrolling patients later, in which more people with ApoE4 received the higher dose of the drug, people receiving aducanumab did better on cognitive tests than those getting placebo.
In its report to the committee, the FDA said it determined that Biogen’s decision to declare the drug as ineffective was “not an accurate reflection of the individual studies” and that even the study that enrolled more slowly, “would have had a reasonable chance of success if run to completion.” In a June meeting with the company, the FDA advised Biogen that “it would have been more appropriate if futility had not been declared and that it was possible that [the slow-enrolling] study might not only be interpreted as being supportive of the efficacy of aducanumab but might also be considered exceptionally persuasive.”
More importantly, the agency accepted the pharma company’s virtual modeling of the studies that offered a post-hoc analysis of what the results would have been had they been completed. The FDA scientists agreed with Biogen that the termination of the studies did not compromise the data, and that those data could provide the basis for the company’s application for approval of the drug. “Overall, the results of the [slow-enrolling] study are highly persuasive and capable of providing the primary contribution to a demonstration of substantial evidence of effectiveness of aducanumab,” the FDA said.
Why the FDA committee was concerned
The FDA’s statistician, however, noted some concerns about how the company interpreted the data from the positive study, and some experts on the advisory committee panel also expressed confusion about how to interpret the conflicting results from the two major aducanumab studies. They also noted that while the studies clearly showed that the drug reduced amyloid in the brain, and that the more drug people received, the more amyloid was reduced, the evidence that this reduction then translated into quantifiable improvements in patients’ brain functioning was less solid. “The committee spent remarkably minimal time discussing the clinical difference,” Alexander says. “We could have spent more time on that.”
Based on these gaps in the data, some of the committee members criticized the FDA for its rosy review, noting that its materials were biased toward taking for granted that the single study firmly established aducanumab’s effectiveness. “The briefing book materials were not at all symmetric…and there is no question all of this is terrifically one-sided,” said Dr. Scott Emerson, professor of biostatistics at University of Washington. “I’m highly critical of the fact that the FDA presentation today was heavily weighted to giving the same conclusions as the sponsor [Biogen].”
Dr. Pierre Tariot, director of the Banner Alzheimer’s Institute, who is not a committee member but treats Alzheimer’s patients, was surprised that there wasn’t more discussion about advising the FDA on how to move forward given the conflicting results. “I’m not an FDA expert, but I theoretically would be supportive of another round of discussion with a different emphasis, on what else could we do to get more answers while beginning to make this [drug] available in a constrained way,” he says. There might, for example, be a way for the drug to be only prescribed and administered by dementia specialists, with tight control over which patients receive it, and extensive monitoring and collecting of data to continue to add to the knowledge about the drug. “Call me Pollyanna, but I like to believe that there is some other path forward than ‘forget about it.’”
That view was on strong display during the public comment period as patients, patient advocates and doctors who treat Alzheimer’s disease urged approval of the drug. They reflected the other reality of Alzheimer’s: that there has not been a new medication to treat its symptoms in nearly two decades, and that there has never been a drug to treat the underlying causes of the disease. The Alzheimer’s Association supported approval, noting that the burden of the disease is shouldered by not just patients but their caregivers, and that any drug that can help people to live independently or function on their own for even a little bit longer could have a huge positive impact. That immediate benefit is what patients are most eager to see, even with a drug that may not be as effective as doctors hope.
“If we wait for another study, there is 0% chance that people will benefit,” said one Alzheimer’s patient whose family has battled the disease for three generations. “We’ll take those odds now. For us, waiting for perfection is not an option. We are losing the ability to recognize family members now. We are becoming agitated now. We are losing ourselves now. We simply cannot wait.”
Now it’s up to the FDA to decide whether it agrees with those patients, or whether it agrees with the experts who advocated for an additional trial to settle lingering questions about the drug’s efficacy. The agency has until next March to make that decision.