Hitting the sweet spot in the spectrum of Alzheimer’s disease for future trials of anti-amyloid medications

As I have discussed in previous posts, evidence of beta-amyloid plaques can be found in the brains of people up to 20 years prior to the onset of cognitive impairment due to Alzheimer’s disease. Abnormal tau protein starts to appear in neurofibrillary tangles in the brain later but still at least several years before onset of cognitive impairment. Information from amyloid and tau PET scans as well as newer blood tests for beta-amyloid and tau have confirmed this time line.

We have also talked about the disappointing failures so far of anti-amyloid monoclonal antibodies in slowing cognitive decline in Alzheimer’s disease despite effectively removing beta-amyloid from the brain. Several recent trials are studying the effects of these drugs earlier in the disease, before there has been any cognitive impairment. All subjects in these trials must have biomarker evidence of brain amyloid as well as normal cognition.  Most studies require a positive amyloid PET scan, but at least one is using a sensitive and specific blood test for beta-amyloid to qualify participants for the study.  Although I am very excited about these “preventative” trials and I really hope they succeed, there is an intrinsic problem. If the asymptomatic amyloid-positive phase of Alzheimer’s can last up to 20 years, how will we know if removing the amyloid will prevent or slow Alzheimer’s disease? It is impractical to wait 20 years to find out. Ideally, we should be studying subjects who are amyloid-positive, don’t yet have tau in their brains, but will become tau-positive within a few years.

A recent paper in Neurology shows how it might be possible to identify research subjects who are in this narrow window, those who currently have amyloid but no tau, but who have characteristics suggesting that they will become tau-positive and cognitively impaired within five years. These investigators at the Mayo Clinic collected data from the Alzheimer’s Disease Neuroimaging Initiative, a large, international database containing amyloid and tau PET scans and MRIs obtained from hundreds of subjects representing all stages of the Alzheimer’s spectrum. While all subjects had detectable brain amyloid at the outset, and no tau, those with higher levels of amyloid, atrophy of the hippocampus seen on MRI, APOE-4 positivity, and less preserved cognition were significantly more likely to have a positive tau PET scan and/or dementia in five years. Restricting participants in trials of anti-amyloid medications to those with higher amyloid burden, smaller hippocampi, and APOE-4 positivity could be important for proof of concept: can anti-amyloid medications slow or even stop the progression of Alzheimer’s disease? If such a trial were to yield negative results, it would probably be the final nail in the coffin for anti-amyloid therapies.  But if these trials of anti-amyloid medications restricted to those on the verge of converting to tau-positivity and cognitive impairment are positive, it will be a great leap forward, and further studies with broader inclusion criteria could begin.