“The whole thing is befuddling, and a series of inexplicable surprises,” says Dr. Caleb Alexander, professor of epidemiology and medicine at the Johns Hopkins Bloomberg School of Public Health. Alexander is a member of the U.S. Food and Drug Administration’s advisory committee that provided recommendations to the agency on whether or not to approve the Alzheimer’s drug aducanumab.
On June 7, the FDA did approve the drug (under the brand name Aduhelm), in a highly controversial decision that continues to divide the Alzheimer’s community. Patients and advocacy groups, including the Alzheimer’s Association, support the approval, noting that the drug is the first to address the underlying causes of Alzheimer’s disease. While it’s not a cure or a home run therapy, they say, it’s a needed start.
Neurology and dementia experts, as well as statisticians who have looked at the data the drug makers submitted for approval are more skeptical. They note that the two large studies of the drug were hardly convincing in showing effectiveness. One study showed no significant benefit among people who received the drug compared to those getting a placebo, while the other showed only marginal improvement in cognitive tests among people getting the drug.
Adding to the contention around the FDA decision is the way the agency made it. Before approving a drug, the FDA conducts a thorough review of the company’s request, which includes all the data on effectiveness, and convenes an advisory committee, made up of independent experts, to also look over and weigh in on the data. The independent panel votes on whether to approve the drug or not, and the agency takes that vote into consideration in making its decision.
In aducanumab’s case, 11 members with expertise in neurology, Alzheimer’s disease, epidemiology and statistics, met in November to review the data presented by the makers of aducanumab, U.S.-based biotech Biogen and Japanese pharmaceutical company Eisai. After reviewing that data, 10 members of the committee voted that the evidence was not sufficient to approve the drug, and one was uncertain.
But on June 7, the FDA decided to approve the drug anyway. Following the approval, three committee members—Dr. Joel Perlmutter, professor of neurology at Washington University St. Louis, Dr. David Knopman, a clinical neurologist at the Mayo Clinic, and Dr. Aaron Kesselheim, professor of medicine at Harvard Medical School—resigned, citing concerns over the FDA’s decision-making process for this drug.
“As an advisory committee member, I am extraordinarily disappointed that our unbiased advisory committee review was not valued,” Perlmutter wrote in an email explaining his decision to TIME. “For these reasons, I resigned my membership on this committee.” Kesselheim did not immediately respond to requests for an interview, but said to STAT that the drug’s approval is “probably the worst drug approval decision in recent U.S. history.” Knopman declined to be interviewed by TIME for this article but said in an email response to CNN that “I resigned from the … committee because if I ever were asked to serve on a future panel, I wouldn’t have wanted to be treated in the disrespectful way that the aducanumab external advisers were treated.”
In addition, the committee members were surprised by a number of other apparent liberties the agency took in making its decision. The committee concluded that the evidence was insufficient for approval because while the studies showed that aducanumab could reduce levels of amyloid in the brain, there wasn’t strong enough evidence to show that reduction led to clinical benefit in the form of improved scores on cognitive tests, which are a proxy for real-world benefit. “There is insufficient evidence that A-beta [amyloid] clearance predicts clinical benefit,” Perlmutter said. “This is a major problem.”
Yet in approving the drug, the FDA decided that the amyloid reduction could serve as what scientists call a surrogate endpoint—a stand-in for clinical benefit. “The committee was never consulted regarding the prospect of using amyloid as a surrogate,” says Alexander. In fact, he says, at the November meeting, the committee specifically asked FDA representatives if they would consider the amyloid reduction as a surrogate in making its decision. “The FDA advised us that it didn’t plan to do so. There was a specific query to FDA as to whether they were considering amyloid as a surrogate and they explicitly said they were not.”
Dr. Patrizia Cavazzoni, director of the agency’s Center for Drug Evaluation and Research, was not available to speak to TIME, but an FDA spokesperson provided written responses to questions about the agency’s decision. In explaining why the agency accepted amyloid as a surrogate marker, the spokesperson wrote, “The advisory committee’s view was that there was insufficient evidence that the drug provided clinical benefit. Taking the advisory committee’s input into account, we considered the application further and determined that although there was residual uncertainty about clinical benefit, as the committee told us, Aduhelm does reduce amyloid plaque. FDA further concluded that it is reasonably likely that this reduction in amyloid plaque will result in meaningful clinical benefit to patients. Given the unmet needs for patients with Alzheimer’s disease—a serious, progressive, and ultimately fatal disease—the Agency chose to use the accelerated approval pathway to allow earlier access to patients while we continue to acquire data on the drug’s benefit.”
That wasn’t the only surprise. The FDA decided to use the Accelerated Approval pathway to approve aducanumab, something that also was not discussed with the committee. “The decision to use the Accelerated Approval Pathway was made at the eleventh hour, without consultation with the advisory committee,” says Alexander. “And it’s unclear why.”
Unmet need for treatment vs. the need for more studies
Technically, the FDA can make these decisions unilaterally, but it’s unusual not to consult an expert panel when it’s been convened and whose purpose is to provide expertise on exactly such matters. In further clarifying the agency’s decision, the spokesperson said “FDA considered the committee’s input, weighed the overall evidence in the application, and considered the dire situation of patients with [Alzheimer’s] who have few treatment options. Based upon this, the FDA concluded that an accelerated approval was appropriate.”
As part of the Accelerated Approval, Biogen is required to conduct a new, randomized controlled trial to confirm the efficacy of its drug. “We are happy to do this; we don’t want there to be any lingering doubt in anyone’s mind about the efficacy of aducanumab,” Dr. Alfred Sandrock, head of research and development at Biogen tells TIME.
Launching that trial, which Biogen anticipates will occur in 2022, won’t be easy since there is little incentive for people to join the study and run the risk of getting placebo when their doctor could prescribe them aducanumab, since it’s approved. While people volunteering to join the study would receive the drug for free, if they are randomly assigned to get a placebo, they would delay any benefit they might receive from getting the drug. So patients will have to weigh those factors in deciding to participate. Among strategies that Biogen scientists are discussing to incentivize people to volunteer include skewing the ratio of drug to placebo in favor of the drug so people volunteering for the trial will have a higher chance of getting drug over placebo.
In any case, there’s no guarantee the results of this trial will be any more illuminating than those of the earlier large studies. “Confirmatory trials aren’t a slam dunk,” says Alexander, adding that this trial should have been required before approval of the drug, and not after. “The FDA missed an enormous opportunity by not requiring a third pivotal trial before market approval.”
While the FDA appears to have relied heavily on the unmet need in the Alzheimer’s field for an effective treatment in giving aducanumab the green light, Alexander questions whether that should trump scientific evidence—or lack thereof. “Unmet need should drive research and investment in drug development, not lower standards of safety and effectiveness,” he says.
Will the approval help or harm future drug development?
How FDA treats surrogate endpoints like amyloid has implications far beyond this one drug. It’s possible, for example, that taking a cue from this approval, other makers of drugs that reduce amyloid levels in the brain but have not shown significant clinical benefit—and there are at least a dozen that would fall into this category from recent years—could flood the agency with requests for approval.
In addition, says Perlmutter, if a treatment for the disease exists on the market, “new Alzheimer’s disease studies may be required to no longer compare an investigational drug to placebo—as would be the standard—but rather may have to compare to aducanumab.” If aducanumab turns out not to be a strongly effective drug, that could lower the bar for future drugs and open the door to increasingly less effective therapies.
More importantly for patients and their families, the approval, and rollout of aducanumab, may muddle Alzheimer’s treatment in unproductive ways. The drug has only been studied in people with the earliest signs of Alzheimer’s dementia, but in its approval, the FDA did not limit use of the drug to this population. In a June 7 briefing, Dr. Peter Stein, director of the Office of New Drugs at the FDA, defended that decision, noting that “the mechanism of [aducanumab] is relevant to all stages of Alzheimer’s disease. It’s not expected that this is only relevant necessarily to the early stages. But because amyloid is a key part, a hallmark of the disease through the entire course of the disease, the expectation is that this drug will provide benefit across that spectrum.”
The benefit that patients at different stages of disease will experience, however, could vary greatly, with those starting aducanumab soon after their diagnosis potentially seeing greater slowdown of their cognitive decline than people starting the drug years after they have been living with the disease. Experts don’t know—there’s simply no data yet.
Patients and their families, however, may understandably have a hard time setting their expectations appropriately and expect the drug to improve cognitive symptoms for all comers. “That’s a tough conversation,” says Dr. Ronald Petersen, director of the Mayo Clinic Alzheimer’s Disease Research Center about how he will talk to his more advanced patients about the drug. “I’m worrying about that.”
In fact, much of the responsibility for ensuring that aducanumab goes to the right patients at the right time will fall on doctors, since the FDA did not specify any requirements for use of the drug. Dr. Stephen Salloway, professor of neurology at Brown University and one of the lead investigators studying the drug, notes that aducanumab was studied in people with mild Alzheimer’s disease who had amyloid plaque already building in their brains (which can be identified using brain imaging.) Yet the FDA did not require patients to demonstrate such a baseline of amyloid to qualify for the drug, so it will be up to doctors to conduct that testing and prescribe it only in those with amyloid that the drug can target. While that might seem obvious, some previous studies of anti-amyloid drugs failed because it turned out that while patients showed signs of dementia and memory problems, they did not actually have amyloid plaques in the brain.
Then there is the cost of the drug. Biogen announced that aducanumab would run about $56,000 a year for each patient, with many potentially needing to remain on the drug for years to prevent amyloid from building up in the brain. It’s not clear yet how insurers will reimburse for the treatment, but they will likely consider the financial burden of long term care of Alzheimer’s patients, which includes extensive monitoring and increasingly complex medical care as more body systems become affected. In any case, the cost to the health care system for the drug will be enormous. “The implementation of aducanumab therapy will potentially cost billions of dollars, and these dollars may be better spent in either developing stronger evidence for aducanumab or other therapeutic interventions,” said Perlmutter.
For its part, Biogen believes the drug will be cost effective, even if patients need to take it indefinitely. Some participants of the early phase trial have now been on the drug for four years, and they continue to see reduction in amyloid plaques in their brains, says Samantha Budd-Haeberlein, vice president of Alzheimer’s disease discovery and development at Biogen. “We do believe the treatment does continue to benefit, so we need to learn more about how long the optimal treatment duration is for different patients.” According to the conditions of the approval, Biogen has up to nine years to complete that required study.
“Patients and caregivers deserve to know that the treatments they are receiving actually work. And that’s quite unclear in this instance,” says Alexander. “And it’s unacceptable that we may have to wait nine years to find out.”