Skip to Main Content

One of the many things I’ve learned during four decades of doing research on Alzheimer’s disease is that the work always brings surprises. Biogen’s announcement on Tuesday about its experimental Alzheimer’s drug, aducanumab, was a big one.

Back in March, the company stopped two large Phase 3 clinical trials of the drug after futility analyses showed that aducanumab was unlikely to provide a benefit compared to placebo.


That decision was a blow to the hypothesis that a protein called beta-amyloid is a cause of Alzheimer’s. It cast a cloud over years of research, and many of us in the field began feeling like we were swimming upstream against pronouncements that anti-amyloid therapies won’t work. Several biopharmaceutical companies started rethinking their research in this area.

Biogen’s latest announcement is likely to give some in the field whiplash.

Why the reversal?


In the futility analyses, Biogen didn’t include some patients on the highest doses of aducanumab and some who had taken the drug for longer times. Since then, the company has included all trial participants. What emerged from the updated analyses is that the dose of aducanumab that gets across the blood-brain barrier and the duration of treatment matter.

This about-face highlights a problem. The all-or-none decisions represented by a futility analysis are sensible from a commercial perspective regarding how a company will fund further work. But they don’t make as much sense scientifically. In retrospect, it would have been ideal if Biogen had completed the analysis of all available data and then made its best judgment. That’s now happened.

Not long after Biogen’s decision to stop the trials, I wrote in Nature Reviews Neurology that when Biogen had a chance to look more closely at the data, it would likely observe some participants who benefited from aducanumab. At the time, I didn’t think the later analyses would be as good as what I heard Tuesday morning, which is that one of the two trials fully achieved its primary and secondary endpoints.

When Biogen pulled the plug on its aducanumab trials in March, some Alzheimer’s researchers, journalists, and members of the public called for stopping, or at least curtailing, work based on the amyloid hypothesis. I didn’t agree. A massive amount of neuropathological, biochemical, biomarker, and genetic evidence in humans accrued over the last 35 years, along with extensive animal modeling, all suggested that beta-amyloid is very likely implicated in the basic biology of Alzheimer’s disease.

Instead of stopping the work, we needed to move ahead with better trials, even though aducanumab had apparently failed in Phase 3. By better I mean trials conducted among individuals at earlier stages of the disease, or among those with positive beta-amyloid scans or those who have elevated levels of another key protein, tau, in their cerebrospinal fluid. Another target group could be individuals with the “purest” form of Alzheimer’s — those without significant vascular lesions or any evidence of so-called mixed dementia.

In this regard, the development of biomarkers measured in blood samples, including certain forms of the beta-amyloid and tau proteins, is coming along quickly. This work is timely if earlier identification of the disease means greater chances that therapy with aducanumab or other agents can help prevent, or at least modify, memory loss and other symptoms such as trouble doing everyday activities.

Biogen’s positive re-analyses of the aducanumab data represent an important proof of concept: a drug can meaningfully affect clinical symptoms in Alzheimer’s. For the industry, it’s a signal that should enable it to go forward with other anti-amyloid trials as well as with trials targeting tau, microglial inflammation, and more.

The news is also important for people with early Alzheimer’s, or at risk for it. Some patients of mine who seemed anecdotally to be doing well as they participated in the aducanumab trials were crushed when Biogen announced that the drug didn’t work and the company wasn’t going to provide it anymore. Biogen has announced that patients who had volunteered to take part in the trials, at some risk and trouble to themselves, should be able to take it again.

Scientists like me have been waiting for something to work, to give us confidence that we were at least on the right track. I have always believed that once a drug looked like it worked and got across the finish line with the FDA, it would open the floodgates to make better versions. I’m not talking about me-too drugs, but ones that actually work better than the first ones.

Take, for example, lovastatin (Mevacor), the first statin approved by the FDA in 1987. It was later clinically eclipsed by atorvastatin (Lipitor) and other “more modern” statins. A similar thing has occurred with blood pressure drugs.

I believe that aducanumab is the breakthrough we have long been waiting for, one that I believe will stand the test of time. But even if the FDA decides not to approve the drug at this point, it can’t take away the evidence that anti-amyloid approaches, even in people who already have symptoms, can benefit both clinical features and pathological changes in the brain.

This roller coaster has been frustrating for patients, their families, researchers, and those in the pharmaceutical industry. It indicates that neurodegenerative disease is very complex and we need to continuously second-guess our assumptions and our findings.

Dennis J. Selkoe, M.D, is co-director of the Ann Romney Center for Neurologic Diseases at Brigham and Women’s Hospital and professor of neurologic diseases at Harvard Medical School in Boston.

Hear Selkoe talk about the 2021 aducanumab approval in an episode of the “First Opinion Podcast.”

  • Are you currently looking for patients for clinical trials? Or is this drug available to the public yet? My father is 73 and is suffering from hereditary mild cognitive impairment and we were told there was nothing that could be done for him

  • I will become a believer when I see the adequate data. For the time being this is absolutely not the case. We might be facing a miracle, or just dramatic marketing. It is urgent for a good publication to appear in a respected medical journal

    • “A good publication in a respected medical journal” is not the only benchmark. A great deal of important industrial research is initially unpublished, and some of it may be proprietary or otherwise privileged. Here’s my concern. Several years ago, a number of negative studies for endovascular therapy of stroke appeared in “respected medical journals” such as the New England Journal of Medicine. As a result, endovascular procedures were written off as ineffective and there were calls to stop using them in the Neurology and general medical community. However, practitioners were seeing astounding results and lives were being saved. They kept up what they were doing. Subsequently, a number of overwhelmingly positive studies with opposite conclusions were published in the same journal (NEJM). Now, endovascular therapy is considered standard of care, and the entire equipoise of the field has flipped in a matter of months. Published data are generally a few years behind where the field is truly at, and persistence and circumspection are needed to get to the truth. But accepting pat answers from prior work, or simply relying on peer review to vet cutting edge approaches, will sometimes yield false negative results. Everyone is paranoid about Type I errors in Alzheimer’s disease, but we should be equally concerned about Type II errors. You can “believe” or “not believe” as much as you want, but the truth always comes out with time.

    • And that Sir is why we have come up empty 35 years, billions of man hours and countless dollars…with absolutely nothing to show. AB is a symptom not a cause. The people most vested in the AB hypothesis are those whose entire careers have been built on the hypothesis, handing the baton down to their postgrads etc. Time to stop the rotten cycle.

    • Dr. Robert Shore thinks that some investigators are “vested” in the amyloid hypothesis, and that is the reason for a lack of therapies. He is wrong. The reason for a lack of therapies is simply a lack of massive transformational support for early stage basic research by the federal government, and a lack of visionary M.D. physician-scientists to take these insights from the “bench to bedside.” NIH needs to more heavily fund and invest in M.D. researchers with cutting edge ideas. Ph.D. researchers are not leading the way to clinical applications. Has nothing to do with the validity of the amyloid hypothesis. A drug either works or not, and more interventions need to be tested.

Comments are closed.