How an Unproven Alzheimer’s Drug Got Approved

Two months before the Food and Drug Administration’s deadline to decide whether to approve Biogen’s controversial Alzheimer’s drug aducanumab, a council of senior agency officials resoundingly agreed that there wasn’t enough evidence it worked.

The council, a group of 15 officials who review complex issues, concluded that another clinical trial was necessary before approving the drug. Otherwise, one council member noted, approval could “result in millions of patients taking aducanumab without any indication of actually receiving any benefit, or worse, cause harm,” according to minutes of the meeting, obtained by The New York Times.

“It is critical that the decision be made from a place of certainty,” the minutes said.

The session, whose details have not been reported before, represented at least the third time that proponents of approving aducanumab in the F.D.A. had received a clear message that the evidence did not convincingly show the drug could slow cognitive decline.

On June 7, the F.D.A. greenlighted the drug anyway — a decision that has been met with scathing rebuke from many Alzheimer’s experts and other scientists and calls for investigations into how the agency approved a treatment that has little evidence it helps patients.

How and why the F.D.A. went ahead and approved the drug — an intravenous infusion, marketed as Aduhelm, that the company has since priced at $56,000 a year — has become the subject of intense scrutiny. Two congressional committees are investigating the approval and the price. Much is still unknown, but an examination by The Times has found that the process leading to approval took several unusual turns, including a decision for the F.D.A. to work far more closely with Biogen than is typical in a regulatory review.

Allegations about the collaboration prompted the F.D.A. to conduct an internal inquiry after a consumer advocacy group called for an inspector general’s investigation, according to documents reviewed by The Times. The agency has not disclosed the inquiry.

Though the decision was considered one of the agency’s most consequential and controversial in years, its leader, Dr. Janet Woodcock, the acting commissioner, was not involved in the deliberations and left the final ruling to the head of the center responsible for drug applications, the agency confirmed.

In written responses to questions from The Times, the F.D.A. defended its decision to approve the drug — the first for Alzheimer’s in 18 years.

“The agency did not lower its standards,” the F.D.A. said, adding, “and at no time considered doing so.”

The decision, the agency said, was “informed by science, medicine, policy, and judgment, in accordance with applicable legal and regulatory standards.”

In written answers to questions, Biogen said, “Biogen stands 100 percent behind Aduhelm and the clinical data that supported its approval.”

Facing mounting pressure, Dr. Woodcock recently called for an inspector general to investigate the agency’s approval process. Dr. Woodcock has publicly acknowledged “process problems,” but has not described what those problems were.

“This incident has shaken F.D.A. integrity quite significantly,” said Wayne Pines, a former F.D.A. senior official who has written histories about the agency. “The F.D.A. is obligated to be sure that all stones are turned over, that every avenue is pursued to make sure that this was a decision that was made on the basis of scientific judgment and not on the basis of anything else.”

While some Alzheimer’s experts did support the drug’s approval given the dearth of treatment choices for patients, many say it was a mistake to approve a medication with such unclear evidence of benefit and that trials showed can cause brain swelling or brain bleeding.

Two nearly identical late-stage clinical trials of aducanumab were shut down in 2019 because an independent monitoring committee concluded that the drug did not appear to be helping patients. A later analysis by Biogen found that participants receiving the highest dose of aducanumab in one trial experienced a very slight slowing of cognitive decline, but participants in the other trial did not benefit at all.

Analysts have predicted that the drug could bring Biogen billions of dollars. But since the approval, some major medical centers have decided not to offer it, and the American Neurological Association’s executive committee told the doctors who are its members that “based on the clinical evidence, Aduhelm should not have been approved at this time.”

Even some scientists who were involved in earlier phases of the company’s aducanumab research said in interviews that they did not agree with the F.D.A.’s decision.

“This approval shouldn’t have happened,” said Dr. Vissia Viglietta, a former Biogen senior medical director, who helped design the two late-stage clinical trials of the drug. “It defeats everything I believe in scientifically and it lowers the rigor of regulatory bodies.”

Because of that, “I felt really deflated personally,” Dr. Viglietta said, adding, “This was not the reason why my team and I did the work we did designing the study.”

In announcing its approval in June, the F.D.A. acknowledged there was not sufficient evidence that the drug would help patients. Instead, it said it was greenlighting Aduhelm under a program called “accelerated approval,” which allows the authorization of drugs without persuasive proof of benefit if they are for serious diseases with few treatment options and if the drug affects part of the disease’s biology (known as a biomarker) in a way that is “reasonably likely to predict clinical benefit.”

The reason the agency gave — that the drug reduces a key protein that clumps into plaques in the brains of people with Alzheimer’s — is one that the agency official leading the aducanumab review had said in an earlier public meeting would not be used. Many Alzheimer’s experts say there is not nearly enough evidence that reducing the protein, amyloid, slows memory and thinking problems.

Many of the questions surrounding the approval of the drug have centered on the close working relationship the F.D.A. and Biogen seemed to have during the application process. That included meeting several times a week in the summer of 2019 to jointly assess the data and chart a path forward, as well as a joint Biogen-F.D.A. presentation to a committee of independent experts.

After receiving letters in December and January from the consumer advocacy group Public Citizen calling for an inspector general’s investigation of the collaboration, the F.D.A. began an internal review focusing on the issue.

The inquiry took place during the spring, as the decision deadline on the drug loomed, and was conducted by an office in the Center for Drug Evaluation and Research. The center includes the office that led the aducanumab assessment. Public Citizen said it was not informed that the inquiry was taking place. Biogen said the F.D.A. had not contacted the company in the inquiry.

It is unclear what the internal review concluded. Michael Felberbaum, a spokesman for the F.D.A., said the agency would have no comment while the issue of close collaboration is the subject of external investigations.

The approval was the culmination of a roller-coaster journey for aducanumab, which seemed to be dead when the trials were aborted in March 2019. Less than two months later, Biogen decided that because its subsequent analysis had found a slight benefit for patients on the high dose in one trial, it would revive its effort to get the drug approved.

That May, Dr. Al Sandrock, Biogen’s chief medical officer, scheduled an informal meeting with Dr. Billy Dunn, director of the F.D.A. office that reviews Alzheimer’s drugs, at a neurology conference they both attended. At the meeting, first reported by the medical news organization STAT and confirmed by The Times, Dr. Sandrock showed the regulator some data underlying the new analysis.

The discussion led to a formal meeting several weeks later on the F.D.A. campus. There, according to minutes of the meeting, Biogen and Dr. Dunn’s team decided that “it is imperative that extensive resources be brought to bear on achieving a maximum understanding of the existing data” on aducanumab.

“Given the wholly unique situation that is the current state of the aducanumab development program,” the minutes say, “further analyses would best be conducted as part of a bilateral effort involving the agency and sponsor, i.e. through a ‘workstream’ or ‘working group’ collaboration.”

Biogen said the idea for collaboration was proposed by the F.D.A. and was “carefully structured and documented, and in effect, allowed for an appropriate deep dive analysis by the F.D.A.”

The F.D.A. said it “often works closely with industry,” especially “where there is a significant need for treatments for devastating diseases.”

When Biogen officials presented what had happened to the company’s board of directors, “people were just blown away that this would be the situation and that aducanumab actually might have a forward path,” said a person familiar with the session.

As the process unfolded, a former employee was surprised by the collaborative workstream, saying “what I was shocked by was just how close the interaction was between the teams.”

While aducanumab was in trials, Dr. Dunn and Samantha Budd Haeberlein, who oversaw the drug’s clinical development for Biogen, worked together on several other projects, interactions that some scientists, former F.D.A. officials and former Biogen employees said they thought blurred the expected boundary between a regulator and an official of a company in that regulator’s purview.

The projects included a framework for understanding the biology of Alzheimer’s disease published in 2018 as part of a work group convened by Maria Carrillo, chief science officer of the Alzheimer’s Association, a patient advocacy group that later pushed for aducanumab’s approval. That effort led to new F.D.A. guidance for reviewing Alzheimer’s drugs drafted by Dr. Dunn’s team.

Dr. Dunn and Dr. Budd Haeberlein also made joint presentations or appeared together on conference panels several times during the aducanumab trials.

William B. Schultz, who served as a deputy F.D.A. commissioner and general counsel for the Department of Health and Human Services, the F.D.A.’s parent agency, said such interactions were ill-advised.

“It is not appropriate for F.D.A. officials to collaborate on publications and presentations with employees of companies with applications pending before those very officials,” he said. “It undermines the essential arm’s-length relationship between the regulator and the regulated industry and destroys the F.D.A.’s credibility as the government agency entrusted with the critical responsibility of deciding the safety and efficacy of drugs.”

The F.D.A. said “it is part of the agency’s role to participate” in the group developing the Alzheimer’s framework, but declined to comment about the joint presentations or say whether Dr. Dunn’s participation in them was cleared beforehand.

Asked about Dr. Budd Haeberlein’s working relationship with the F.D.A. official, Dr. Priya Singhal, Biogen’s head of global safety and regulatory sciences, said: “Relationships do not govern the regulatory process or its outcomes. There is no relationship that would override data gaps.”

Last November, Biogen and Dr. Dunn’s team presented a joint review to an advisory committee of independent experts outside the agency who were tasked with voting on whether aducanumab was ready for approval. Usually, a company and F.D.A. reviewers give separate presentations.

The joint presentation asserted that there was “substantial evidence of effectiveness to support approval” and, in language that a former F.D.A. official said was unusually effusive for a scientific presentation, described the single positive trial — which showed that the high dose slowed decline by 0.39 on an 18-point scale — as “exceptionally persuasive.”

That was not the conclusion of every F.D.A. division. The agency’s biostatistical office had reached an opposite assessment, writing in a separate review presented to the committee that “there is no compelling, substantial evidence of treatment effect or disease slowing.”

At the end of the daylong meeting, the advisory committee overwhelmingly agreed with the biostatistical assessment. To the question of whether there was enough evidence the drug would help patients, 10 members of the panel voted no and one was uncertain. There were no yes votes.

“To have a virtually unanimous vote against approval and then to have the F.D.A. turn around and approve — that’s never happened,” said Mr. Pines, the former F.D.A. official, who now directs the health practice at APCO, a public relations firm.

For months after the advisory committee meeting, Dr. Dunn and his team continued to work toward conventional approval of the drug.

But when the case was presented to the agency’s Medical Policy and Program Review Council, meeting on March 31 and April 7, the vast majority of the 15 members said the evidence did not meet the threshold for “instilling public confidence in the usefulness of the drug,” according to the minutes. Another trial was necessary, the council agreed, but could be “a shorter and more efficient trial,” countering the contention from approval advocates that another trial would take years.

Although the council, which advises on but doesn’t make approval decisions, recognized that some patients would accept the drug “despite the uncertainties,” the minutes say, “the council, however, stressed that this should not influence the regulatory decision.”

The idea of accelerated approval came up briefly toward the end, raised by Dr. Rick Pazdur, head of F.D.A.’s oncology center, who was not a council member. It was not discussed in detail, but after the meeting, given the council’s rejection of standard approval, accelerated approval appeared to be the only way to make the drug available.

On April 26, Dr. Patrizia Cavazzoni, Dr. Dunn’s boss and director of the Center for Drug Evaluation and Research, led a smaller meeting about accelerated approval, which had never been used for Alzheimer’s drugs.

In fact, the F.D.A.’s most recent guidance for Alzheimer’s drugs, issued by Dr. Dunn in 2018, says “the standard for accelerated approval” had not yet been met for the disease, “despite a great deal of research.” The guidance says that is because “there is unfortunately at present no sufficiently reliable evidence” that attacking amyloid plaques or other biomarkers of Alzheimer’s “would be reasonably likely to predict clinical benefit.”

And at the November advisory committee meeting, Dr. Dunn said that in considering whether to approve aducanumab, “we’re not using the amyloid as a surrogate for efficacy.”

Under accelerated approval, while a drug is on the market, a company must conduct an additional trial, a costly undertaking. Biogen said its goal was standard approval, which it believed its data warranted.

At the April 26 meeting, Dr. Cavazzoni invited two officials not involved with neurological drugs who had used accelerated approval frequently: Dr. Pazdur and Dr. Peter Marks, the top vaccine regulator. They and Dr. Cavazzoni voted to grant such approval to aducanumab, as did Dr. Issam Zineh, director of the Office of Pharmacology, and Dr. Jacqueline Corrigan-Curay, who led the internal review of the F.D.A-Biogen collaboration.

The director of the office of translational sciences, Dr. ShaAvhrée Buckman-Garner — who supervises both the pharmacology and biostatistics offices — did not vote yes or no, saying she understood both arguments. The only clear no vote, F.D.A. documents say, was the director of the office of biostatistics, Dr. Sylva Collins, “stating her belief that there is insufficient evidence to support accelerated approval or any other type of approval.”

Mr. Felberbaum, the F.D.A. spokesman, said the agency “concluded that the reduction in amyloid beta plaques was reasonably likely to predict clinical benefit, meeting the requirements for an accelerated approval.”

The 2018 guidance “no longer reflects the current state of the science,” he said, citing data from aducanumab and earlier-stage trials of other anti-amyloid drugs, evidence that many Alzheimer’s experts say is not strong enough to link reduction in amyloid to the likelihood of slowing cognitive decline.

In its written responses to The Times, the agency added that its decision took into account that patients expressed “their willingness to accept some uncertainty about clinical benefit to get earlier access to a potentially clinically valuable drug.”

On April 28, Biogen was told that aducanumab was now being considered for accelerated approval instead of standard approval, according to documents obtained by The Times.

The documents also show that Biogen submitted draft language for a label stating which patients should be eligible for the treatment, a common step in the final stages of a drug review for possible approval. Up until about a week before approval was announced, the label listed “contraindications” — medical conditions that should preclude patients from getting the drug. The final approval label, however, has only one word under contraindications: “None.”

The F.D.A. said it does not comment on labeling negotiations. Biogen said that after “thoughtful consideration,” the conclusion was that it should be up to “the treating physician in the real-world” to decide which patients receive the drug.

Another change was made to the proposed label. Before approval, it said the drug’s purpose was “to delay clinical decline in patients with Alzheimer’s disease.”

But after a comment was put in the margin, the language was changed to “for the treatment of Alzheimer’s disease.” The note said the F.D.A.’s office of prescription drug promotion “is concerned with the promotional implication of the phrase ‘to delay clinical decline’ considering that this product is being approved under accelerated approval.”

“In a promotional context,” the note cautioned, “the phrase suggests a guarantee of efficacy.”