NCHR Comment on FDA’s Draft Guidance on Developing Drugs for Early Alzheimer’s Disease

National Center for Health Research: May 16, 2018

National Center for Health Research’s Public Comment on
Early Alzheimer’s Disease: Developing Drugs for Treatment;
Draft Guidance for Industry; Availability

The National Center for Health Research is a nonprofit research center staffed by scientists, medical professionals, and health experts who analyze and review research on a range of health issues.

We support the encouragement of research to develop drugs that would prevent or slow the progression of early Alzheimer’s disease (AD). Even delaying the onset or progression of AD by a few years would be an enormous benefit to patients, caregivers, and society. Further, we support the FDA’s recognition of the importance of efficacy endpoints that demonstrate a clinically meaningful effect on daily function and/or cognition for drug development. It is not appropriate to depend entirely on the use of biomarkers and of limited, narrow measures of neuropsychological performance. It will be important that when these newly developed drugs are evaluated for approval that the FDA maintains high standards for evidence that are based on sound science.

The use of biomarkers as surrogate endpoints to measure efficacy as the basis of accelerated approval is a major concern. In  2015, our Center published an article in BMJ showing how early studies based on biomarkers raised high hopes for several Alzheimer’s drugs – but studies of memory indicated the drugs were not effective.1  In fact, one of the drugs was found to decrease memory function and to increase the risk of skin cancer!

This is especially concerning for prevention of AD because changes in the brain that are thought to eventually contribute to AD ‒ the presumed targets of these drugs ‒ can begin decades before symptoms occur. However, because there are numerous risk factors for AD, it is not yet possible to accurately determine if or when a person might be likely to develop AD, especially at this early stage. This means that it is likely that a large number of people who would not have developed symptoms would take a drug for potentially decades, exposed to the risks without receiving any benefit.

We agree with the FDA’s statement in the guidance that there are not any biomarkers with sufficient evidence to justify their use for approval for this indication. Several biomarkers are correlated with disease progression, but it is still unclear whether prevention or reversal of their accumulation is necessary or sufficient to ultimately prevent disease occurrence/progression or to reverse it. We found this in our previous study, which was based on studies submitted to the FDA for drugs that seemed extremely promising, and there has been limited progress in developing clinically meaningful biomarkers to evaluate efficacy since then. Biomarkers used to evaluate a drug’s efficacy will need to be biologically and clinically well characterized to demonstrate that they accurately reflect a clinical effect on AD onset and/or progression. From a practical standpoint, the biomarker must be able to be robustly tested in patients using validated biochemical assays. The diagnostic test must be performed through minimally-invasive and minimally harmful means.

In addition, the magnitude of effect should be large. An effect on a biomarker or on only one or a few narrow neuropsychiatric or functional measures may not be clinically meaningful for patients. Furthermore, small reductions in population-level risk in the research study population may not be generalizable to more diverse patients and situations throughout the U.S. For comparison, it is important to consider that lifestyle changes reduce the risk of developing AD and also provide additional benefits on health and well-being without the risk that many medications pose. Whichever enpoints are chosen, the effect will need to be large enough to have an observable, meaningful improvement on the lives of patients.

We urge you to keep in mind that, even if a drug is indicated for a narrow, high-risk population, it will likely be advertised and prescribed for a much wider, lower-risk population. And as shown in our previous research on Alzheimer’s drug development, even drugs that look promising can do much more harm than good.

In conclusion, drug development and subsequent approval for early Alzheimer’s disease has the potential to provide a great benefit to patients, caregivers, and society. But to do so, the drugs must provide large clinically meaning benefit not just hope.

NCHR can be reached through Stephanie Fox-Rawlings, PhD, at


  1. Zuckerman DM, Jury NJ, Silcox CE. 21st Century Cures Act and similar policy efforts: at what cost? BMJ 2015; 351:h6122.