Last week I discussed my concerns about using biomarkers and brain scans to diagnose Alzheimer’s disease. This week Medscape (www.medscape .com) published an article that suggested that “misdiagnosis of Alzheimer’s disease [in about two thirds of participants] based on positron emission tomography scan readings appears to be a troubling problem that could get worse as more amyloid-specific tracers become available.”

This is the first study to my knowledge to present objective evidence suggesting caution about the use of brain scans to make a diagnosis of Alzheimer’s disease. The unique thing about this study was that the scans were done on community dwelling participants (mean age of about 64 years). Studies used to make the case for the accuracy of scans have so far only used highly selected participants – a sampling bias that distorts results.

The bottom line is that positive scans detect high levels of amyloids but high levels of amyloids don’t mean that the person scanned has Alzheimer’s disease. Indeed, another article (Archives of Neurology, April 23, 2012) presents data that both amyloid (associated with plaques) and tau (associated with tangles) proteins need to be present for cognitive decline. Decline over a period of three years was not associated with high levels of amyloid proteins in cerebrospinal fluid (CSF). Rather decline after was associated with high levels of both tau and amyloid proteins in CSF.

What this means is that scans are still research tools and not diagnostic tools for Alzheimer’s disease. In everyday clinical practice, scans should not be used for diagnosis of Alzheimer’s disease. Instead, diagnosis needs to be made on the basis of careful and detailed history/course, neurological exam, neuropsychological exam, and structural imaging. Furthermore, diagnose often requires serial assessments with the first exam as a baseline. Eliminating any of these steps dramatically increases misdiagnosis and increases unneeded stress.

Donepezil, generic name for Aricept, (and presumably rivastigmine and galantamine) and Namenda were associated with less decline over the course of one year than when both medications were discontinued in patients with moderate to severe Alzheimer’s disease. Adding Namenda was no more effective than treatment with donepezil alone.

An article in the British Medical Journal (March 22, 2012) warns that the 23 mg dose of Aricept is associated with more nausea and vomiting than the 10 mg dose. These effects can lead to serious medical problems. Drs. Schwartz and Woloshin claim that the label for the higher dose is misleading as there is no evidence of clear benefits from the higher dose when compared to the standard dose of 10 mg. They have petitioned the FDA for review.