Biomarkers for Alzheimer’s disease

Early in my career as a general neurologist, there was no way to accurately make a diagnosis of Alzheimer’s disease in a living patient. The only way then to know for sure that a patient had Alzheimer’s was to see the classic amyloid plaques and tau-containing neurofibrillary tangles in the brain at autopsy. We could exclude some potentially treatable mimics of Alzheimer’s like certain brain tumors and so-called normal pressure hydrocephalus with CT or MRI scans, but we really had little help in distinguishing the several types of progressive dementia due either to Alzheimer’s disease, vascular dementia, Lewy body dementia, Parkinson’s disease or frontotemporal dementia (FTD). In the late 1980s and early 1990s, it really didn’t matter much if we couldn’t tell them apart because we had no treatment for any of these disorders. Then in 1996, donepezil (Aricept) was approved for treatment of Alzheimer’s disease. Suddenly it became more important to be able to be as accurate as possible in our diagnosis of dementia. Donepezil provides some cognitive benefit in many patients with Alzheimer’s as well in some with Lewy-body dementia, but it sometimes makes patients with FTD worse. We neurologists had to hone our diagnostic skills.  Alzheimer’s disease often, but not always, presents first with impairment of verbal memory.  Lewy body dementia classically is associated with early visual hallucinations. FTD usually begins much earlier than Alzheimer’s, and it often causes marked personality changes that precede cognitive decline. However not every patient who came to autopsy turned out to have the cause of dementia that we had diagnosed during life. We needed better biomarkers, tests that directly or indirectly determine what type of dementia was present while the patient was still alive.

Brain imaging tests such as CT and MRI can be helpful biomarkers for FTD because of the typical atrophy of the frontal and temporal lobes sparing more posterior parts of the brain.  For other types of dementia, these imaging tests are not as helpful, although they may be important to rule out tumors and hydrocephalus. Over the last ten years or so, PET scans that can image beta-amyloid plaques and tau-containing tangles have been developed and are now clinically available.  These scans are very useful in the staging of Alzheimer’s disease.  They are very expensive and generally not covered by insurance, but they still important in research settings.

Measurements of beta-amyloid and tau in spinal fluid have been a useful tool, but these require a spinal tap that may be uncomfortable. There has been an extensive search for blood tests for forms of beta-amyloid and tau that correlate with Alzheimer’s disease progression. It seems that a new test is described every few months that is better than the others. Currently, the leader is an assay for p-tau217 (tau protein phosphorylated at position 217). P-tau217 can be detected in plasma well before the onset of cognitive impairment, and it continues to climb as Alzheimer’s progresses while also correlating with plasma beta-amyloid levels. Thus, it seems to be specific for Alzheimer’s disease. Most of these blood tests are not yet commercially available, but they are already being used to screen subjects for clinical trials of potential therapeutic drugs. This is becoming increasingly important as it seems likely that the effective treatment of Alzheimer’s disease may require administering drugs before symptoms of cognitive impairment have begun.  In the near future, the ideal biomarkers will help us define the best time to start treatment.  If we wait for the onset of cognitive impairment, it will probably be too late.