If the U.S. Food and Drug Administration (FDA) wants to approve the first new drug for Alzheimer’s disease in 17 years, it will have to do so against the overwhelming recommendation of the experts it turned to for advice on the matter. An independent advisory panel convened by the agency today to review data on the antibody drug candidate, called aducanumab, concluded that even the strongest available clinical trial data don’t support its effectiveness.
FDA, which is expected to decide about aducanumab by March 2021, doesn’t have to follow the advice of its advisory committees, but it typically does. If approved, aducanumab would be the first Alzheimer’s drug prescribed to slow cognitive decline and would likely bring in tens of billions of dollars in sales for its developer, Biogen. It might also vindicate the battered theory that clearing the brain of the sticky protein called beta amyloid can effectively treat the disease.
During a public comment section of the meeting, people with Alzheimer’s— including some who participated in Biogen studies—and their caregivers strongly urged FDA to approve the drug. But many researchers, including most of the advisory committee members, weren’t convinced by the two large clinical trials of aducanumab—only one of which found evidence of benefit. And the committee was uncomfortable with rosy interpretations of Biogen’s data that FDA presented today and in documents it released this week.
“Alzheimer’s treatment is a huge, urgent, unmet need, but I also think if we approve something where the data is not strong, we have a risk of delaying good treatment,” said Joel Perlmutter, a neurologist at Washington University in St. Louis who is one of 11 voting members of the committee, which included physicians, biostatisticians, a patient representative, and a health care administrator.
Biogen has shown its antibody can bind to and clear away a toxic form of beta amyloid. But two identical, international clinical trials of the experimental drug in patients with early stage disease, referred to as studies 301 and 302, have caused confusion and controversy. Based on a disappointing interim analysis of patient data, Biogen announced a halt to both trials in March 2019, sending aducanumab the way of many other failed antiamyloid drug candidates.
Then, in October 2019, the company shocked the field by announcing it would seek FDA approval after all. Patient data not included in the earlier analysis revealed that after 78 weeks, participants getting the higher of two aducanumab doses in study 302 had 22% less cognitive decline on a standard dementia assessment than those in a placebo group, the company said. However, study 301 was still negative: People in the high-dose group actually had a slightly worse decline than the placebo group. Many researchers insisted that Biogen should conduct another trial to justify approval.
At today’s meeting, scientists at both FDA and Biogen made a case that study 302—with support from a small earlier clinical trial—offered strong evidence that aducanumab works. “We are not ignoring study 301,” Biogen’s head of neurodegeneration development, Samantha Budd Haeberlein, assured the advisory committee. “We have worked diligently with the FDA, and we sufficiently understand why 301 failed.”
She cited two major reasons. Because of midtrial changes to dose regimens, more patients in study 302 than 301 got a consistently high dose of the drug. And compared with 302, 301 included more patients with unusually rapid cognitive decline—18 patients versus 13. Removing those people from the analysis swung the result in 301’s high-dose arm from a 2% deficit to a 6% improvement in cognitive decline compared with placebo.
Billy Dunn, director of FDA’s Office of Neuroscience, echoed those explanations, and said the issues with 301 don’t “meaningfully detract from the persuasiveness of 302.” In his presentation and in FDA documents, the agency suggested the drug might still have a clear path to approval.
Committee members pushed back. Several cited an analysis by FDA’s statistical reviewer, Tristan Massie, who wrote in a briefing released this week before the meeting that “there is no compelling substantial evidence of treatment effect.”
Several panelists also balked at the first question FDA put to them: Does study 302 provide strong evidence of the drug’s effectiveness “without regard for Study 301”? University of Washington, Seattle, statistician Scott Emerson noted the two trials were intended to be evaluated together. “We may never, ever, ever ignore the fact that study 301 was done.”
“It was tacitly accepted [by FDA] that 302 represented truth and 301 did not,” added neurologist Michael Gold, head of neuroscience development at AbbVie, a nonvoting member of the committee. “I just didn’t understand why there seemed to be this unilateral kind of effort to discredit one study.”
Committee members also raised concerns about 302 itself. Some worried that a relatively common side effect of aducanumab—a swelling of the brain—had revealed to patients and caregivers whether they were getting the study drug or placebo, which could bias the results. And committee member Madhav Thambisetty, a neurologist at the U.S. National Institute on Aging, said that in practical terms, the benefits documented in study 302 were “extremely small.”
On the agency’s final question, whether it was “reasonable to consider Study 302 as primary evidence of effectiveness of aducanumab for the treatment of Alzheimer’s disease,” 10 members voted “no.” (University of Virginia neurologist and panel chair Nathan Fountain voted “uncertain.”)
“It’s the right decision,” says Robert Howard, a psychiatrist at University College London who has run clinical trials of potential Alzheimer’s treatments and didn’t participate in the meeting. “Approval of aducanumab would be really bad for my field—and really bad for the patients,” he says. “We’ve got to keep a foot on the gas to find a treatment that works … with much more spectacular effect sizes.”