Q&A: Use of amyloid surrogate endpoint in Aduhelm approval sets ‘dangerous’ precedent

The FDA approval of Aduhelm, which was based on beta amyloid as a surrogate endpoint, represented a “regulatory failure,” according to members of an advisory committee convened by the FDA in November 2020.

In a correspondence published in The New England Journal of Medicine, advisory committee members highlighted the precedent the approval of Aduhelm (aducanumab; Biogen/Eisai) may set for future regulatory decisions related to Alzheimer’s disease treatments.

“The interim FDA Commissioner’s decision to call for an Office of the Inspector General investigation is appropriate, and we believe such an investigation should be expeditiously performed, so as to learn how this regulatory failure occurred and to ensure that it doesn’t occur again,” the authors wrote. “Though the decision will reverberate for years, early signs of its gravity are already reflected in announcements by two other manufacturers that they will seek accelerated approval for Alzheimer’s treatments on the basis of amyloid as a surrogate.”

Healio Neurology spoke with G. Caleb Alexander, MD, MS, professor of epidemiology and medicine at Johns Hopkins Bloomberg School of Public Health, founding codirector of the Center for Drug Safety and Effectiveness and principal investigator of the Johns Hopkins FDA Center of Excellence in Regulatory Science and Innovation, about the viewpoint that he coauthored to learn more about the advisory committee’s concerns regarding Aduhelm.

Healio: What prompted you and your colleagues to write this paper?

Alexander: The publication provided many of the individuals who were on the November 2020 advisory committee an opportunity to reflect further on the approval process and the suitability of amyloid as a surrogate endpoint for AD treatments. Valid surrogate measures are quite valuable in the clinical research enterprise because they can reduce the costs and duration of clinical trials and speed drug development. There is a long history of using surrogates in clinical research, but what is crucial in this instance is whether amyloid is a valid surrogate.

Valid surrogates have to be able to reliably predict the effects of treatment on a clinical outcome and consistently show benefit in varied clinical settings. We explored reasons that a proposed surrogate measure may fail, which include that it may not be on the causal pathway of the disease process, or an intervention may only target one of many different causal pathways that are ultimately responsible for disease.

The effort here was to provide a brief assessment of the evidence to date regarding the suitability of amyloid as a valid surrogate. There is reason for significant skepticism as to the suitability of whether amyloid truly represents a valid surrogate. Further work is needed to assess the scientific suitability of amyloid or other surrogates. To be clear, there's an overwhelming demand for new treatments for this common and devastating disease. It may well be that someday a valid surrogate is identified, but to date we do not have such a marker. Important work remains to be done to explore the suitability both of amyloid and other potential biomarkers that might serve as surrogate endpoints.

Healio: Why might the FDA have decided to move forward with this approval despite the concerns from the advisory committee?

Alexander: What has been reported is that the lead FDA scientists recognized that there was no path forward for standard approval and at the 11th hour decided to rescue the product using the accelerated pathway. In this instance, the accelerated pathway, with amyloid as a surrogate, was used to rescue Aduhelm from what would otherwise have been a certain death, or at least a rejection at this point. The FDA made a last-minute change and decided to approve it based on amyloid as a surrogate, but as we highlight in our piece, you need not look at other drugs to question whether amyloid is a valid surrogate. The evidence from Aduhelm itself undermines arguments that it is so.

In other words, we don't need to look to other experimental treatments that have failed to have doubts about whether amyloid is a valid surrogate. We can look to the history of Aduhelm’s development, which involved the failed ENGAGE clinical trial and the ever-so-modest improvement in clinical outcome in one group in the EMERGE trial.

Healio: In your opinion, what should be next for Aduhelm?

Alexander: Unfortunately, the horse is out of the barn, and it's now left to health systems, hospitals, payers and the federal government to try to clean up the mess. We've seen hospitals and health systems decline to administer Aduhelm. Recently, the Veterans Affairs Administration declined to routinely cover it. Hospitals, health systems, insurers, clinicians and patients have to navigate this matter and wade through these waters to figure out how to proceed.

The evidence to date suggests that there remains an enormous amount of concern and skepticism regarding this product and its safety and effectiveness. The regulatory management of Aduhelm has established a dangerous and scientifically unwarranted precedent whereby amyloid may be used or argued as a valid surrogate by other manufacturers. Both the FDA and other regulators overseas have an important opportunity to get the science right and to manage future drug applications accordingly.

Healio: Is there anything else you’d like to add?

Alexander: The FDA is a remarkable institution and is widely regarded around the world, for good reason, because of the careful science and regulation it performs day in and day out. It's precisely because of this sterling reputation that this case stands in such stark contrast to the typical careful, scientifically grounded regulation that the FDA exercises.

Reference:

Alexander GC, et al. N Engl J Med. 2021;doi:10.1056/NEJMp2110468.