Diagnostic criteria for the neuropathologic assessment of Alzheimer's disease

The provisional criteria proposed in 1985 by Khachaturian et al. emphasized numbers of plaques and neglected tangles, as did CERAD (Consortium to Establish a Registry for Alzheimer's Disease). The decision to set an arbitrary number of plaques as "pathologic" assumed that some neuritic plaques are a normal phenomenon in the aging brain. Neuritic plaques and neurofibrillary tangles are age-related lesions, but they are pathologic (i.e., lesions) no matter how many there are. In a clinically demented patient without vascular or other neurodegenerative lesions, a clinico-pathologic diagnosis of AD (a clinico-pathologic entity) can be made with a high level of confidence by demonstrating, and without counting, plaques and tangles. The vast majority of AD cases are straightforward, and diagnostic lesions can be appreciated with a simple silver stain. If patients' histories are unknown or uncertain, the clinical significance of the observed plaques and tangles must remain debatable. This is the essence of the consensus statement with which I wholeheartedly agree. In such cases without a dementia history, one can offer a neuropathologic diagnosis of Senile or Pre-senile Cerebral Disease (not "dementia") of the Alzheimer type. Precise clinico-pathologic correlations and some quantitative measures are needed for elucidating the pathogenesis of AD and for establishing a primary dementing diagnosis when AD is mixed with other dementing diseases. These correlations must be based on periodic and fairly extensive neuropsychological testing followed shortly thereafter by a detailed postmortem neuropathologic evaluation.