Don’t Give Up on Modifying Alzheimer’s Disease


by Guy McKhann, M.D.

February 12, 2018

This is a column from Dana's print publication, Brain in the News.

In the last decade or so, we have made progress in the modification of acute neurological diseases. This is probably most relevant in how we treat acute strokes, either medically or by intravascular techniques. However, there are also areas in which our progress has been minimal.

Nowhere is this more pronounced than in developing treatments for progressive neurodegenerative diseases such Alzheimer’s disease (AD). Approval of the last drug by the Food & Drug Administration was 14 years ago, and that drug, Aricept, is only marginally beneficial. Why is progress so hard?

In a search for the answer, let’s delve a little further. The progression of AD is slow. So much so that it can be divided into three phases:

The first phase occurs in people who are still cognitively normal, or only minimally impaired. In the second phase, the cognitive impairment progresses so that people experience difficulties with memory and learning new information, but are still functioning relatively well. In the third phase, cognitive impairment starts to set in and progresses on to dementia. This is a state in which a subject cannot learn new information and, in severe forms, cannot recall previously known details. They can even have difficulty recognizing members of their own families.

In 1984, I led a research team that set out to define this third group. Attempts at treatment were aimed at this group, and they all failed. They failed because at this stage of disease, nerve cells in crucial areas required for memory function—such as the hippocampus or entorrhinal cortex—were gone. Imaging studies demonstrated that those areas of the brain were atrophic, and we fell short of finding a way to get replacement neurons into those areas. Treatments were doomed to failure, and millions of dollars were spent on what are now considered fruitless studies.

In recent years, the emphasis has shifted to studying patients in the earlier phases of the disease. The earlier one goes, the harder it is to design such a study. If you are going to evaluate some drug or procedure, you need a control group that is identical to your treatment group—except they are not getting the treatment. The assumption is that, over time, the control group will progress in their AD. However, we now know that this progression from phase one to phase three can take years. It is not steady, and subjects may have periods with no apparent development. Thus, if the endpoint of your study is the stability of your treatment group as compared to the progression of the control group, you are committed to a study of long duration. One possible way around this dilemma would involve choosing subjects who are known to have a more rapid progression of Alzheimer’s.

Are there parameters of the disease, other than clinical progression, that might be used as outcomes? It has been known since the disease was first described in 1906 that there is a distinctive pathology: the accumulation of plaques and tangles. The plaques, which are extracellular, contain a protein called amyloid and the tangles, which are within neurons, contain a protein called Tau. Most studies in recent years were aimed at either blocking the formation of amyloid or removing amyloid from the brain. Drugs were developed that did both of these things, but so far none have modified the disease. Drugs that decrease or prevent Tau are just being tried, with or without anti-amyloid drugs.

So what seemed to be straightforward approaches have not worked. Advocates of these approaches have not given up and are agitating for earlier studies of longer duration and with higher doses. Such trials are frightfully expensive, and pharmaceutical companies—having been burned by unsuccessful studies in the past—are reluctant to fund additional studies. Pfizer, a leading pharmaceutical company, recently announced that it would no longer be involved in such studies, having already spent millions in previous research.

Right now, what’s needed is an approach that modifies the disease in any form. To my way of thinking, that means much smaller studies, including non-clinical end-points. The subject population should be as homogeneous as possible. Whether any positive findings can be expanded to a larger population is a question we can address down the line. Right now, some positive evidence that an approach works needs to be our main concern.

Guy McKhann, M.D., is professor of neurology and neuroscience at the Zanvyl Krieger Mind/Brain Institute, Johns Hopkins University, Baltimore. He serves as scientific consultant for the Dana Foundation and scientific advisor for Brain in the News.