New documents released Tuesday depict, in greater detail, how staff at the Food and Drug Administration were not in agreement as the agency weighed whether to approve a controversial new treatment for Alzheimer's disease.
For months, it was known that the FDA team responsible for evaluating the statistical analyses used in clinical testing did not believe there was enough evidence to approve the drug, which is now called Aduhelm and was developed by the pharmaceutical companies Biogen and Eisai.
But the statisticians were far outnumbered, according to the internal documents and memos posted by the FDA. Every other office that reviewed the Aduhelm application concluded the drug was worthy of approval, a position that was later shared by senior FDA officials including Patrizia Cavazzoni, the head of the agency's main drug review office.
Aduhelm was cleared on a conditional basis on June 7, a decision that has been met with immense scrutiny.
"This [application], including the data and information submitted to FDA by Biogen, has been the subject of intense public interest, due in part to the unmet medical need and the population potentially eligible for treatment," Cavazzoni wrote, in a memo.
"This [application] has also been the subject of rigorous scientific discussion and deliberation within FDA, including within the review team for this application," she added.
Much of the controversy surrounding the drug stems from the fact that doctors and researchers still don't know how much — or even whether — it actually helps Alzheimer's patients in their daily lives. Aduhelm is meant to clear a kind of sticky plaque that has long been viewed as the root cause of Alzheimer's. And on that measure, Aduhelm works well.
But, it remains unclear whether eliminating these plaques slows down the disease's characteristic decline in brain function. Drugs that work like Aduhelm have failed time and again in clinical trials. Even Aduhelm looked as though it would not succeed, as two large studies testing it in patients with early Alzheimer's were initially determined to be unlikely to show a benefit. Only when Biogen and Eisai ran additional analyses with more data did they conclude that one trial was positive enough to support an approval application.
The FDA, however, was unusually supportive of Biogen, working directly with the drugmaker to determine whether the more complete results were interpretable. Staff on the clinical review and pharmacology teams ultimately took the view the data from the one positive trial was "exceptionally persuasive," the documents released Tuesday show.
Both of those teams agreed the second study failed, but concluded there were sufficient reasons to explain why and that its outcome "did not detract" from the persuasiveness of the first.
Agency statisticians disagreed. Their dissenting view was reinforced by a committee of neuroscience experts who advise the FDA on new brain drugs. Last November, the advisers gave a strong rebuke of Biogen's application in a meeting convened by the regulator, voting almost unanimously against Aduhelm. Some saw the vote as a sign a rejection would be forthcoming, as the agency typically follows its advisers' recommendations.
But in the months following that meeting, FDA officials coalesced around a middle path that was not discussed by the experts, deciding to evaluate Aduhelm for what's known as accelerated approval. Unlike a standard approval, accelerated approval is designed to allow drugs to market that show effects on a biomarker that is likely to translate into clinical benefit.
"FDA, then, is faced with a situation where the available evidence on the clinical endpoints is short of what we would require for standard approval of aducanumab," wrote Peter Stein, director of the agency's Office of New Drugs, in another memo.
"At the same time, patients with [Alzheimer's] have a serious, progressive, ultimately fatal disease and are desperate for treatments that delay progression and prolong well-functioning survival," he wrote.
Stein agreed with the Office of Neuroscience's position that an accelerated approval would be appropriate based on Aduhelm's ability to clear the sticky amyloid plaques associated with Alzheimer's. And, in an April 26 meeting, so did Cavazzoni and a number of other high-ranking FDA leaders, including Peter Marks, head of the Center for Biologics Evaluation and Research, and Richard Pazdur, director of the Oncology Center of Excellence, who had been brought in to consult.
Only Sylvia Collins, head of the FDA's biostatistics office, which houses the Aduhelm statistical review team, argued for rejection of the drug.
The FDA's decision to clear Aduhelm under accelerated approval surprised many, particularly as an official indicated in the November meeting the agency wouldn't be taking such an approach.
Many other drugs aimed at amyloid have also failed in recent years, raising doubts about whether the plaques are a good target for treatment.
The memos show FDA staff put heavy emphasis on the amyloid reduction reported in Biogen's trials of Aduhelm, and the correlation between that reduction and measured changes in patient cognition and function. That correlation became weaker, however, when results from individual patients, rather than from the group, were examined.
"Recognizing the limitations of the results, such as missing data, the consistency of these findings across the studies still supports a conclusion that the change in amyloid plaque is reasonably likely to predict clinical benefit," Stein wrote.
Additionally, FDA staff relied on an analysis of six other amyloid-targeting drugs for Alzheimer's that supported a relationship between plaque reduction and clinical benefit, although that review wasn't published.
Some other technical documents typically released by the FDA following a drug's approval also weren't disclosed Tuesday and the full dissenting statistical review was not posted, either. The agency moved somewhat faster than it usually does, however, in making the summary review available.