Clinical-neuropathological correlations in Alzheimer's disease and related dementias

Abstract

Objective: To compare neurologists' initial clinical diagnoses made according to National Institute of Neurological and Communicative Disorders and Stroke-Alzheimer's Disease and Related Disorders Association (NINCDS-ADRDA) and Diagnostic and Statistical Manual of Mental Disorders, Revised Third Edition guidelines with neuropathological diagnoses of Alzheimer's disease (AD) and related dementias.

Design: Consecutive autopsies in a prospective cohort study.

Setting: Community-dwelling patients with dementia referred to neurologists at an Alzheimer's Disease Research Center and satellite clinics (n = 151) and patients initially evaluated when institutionalized (n = 19).

Patients: Of 204 elderly patients who had an autopsy performed, 170 had received a complete dementia evaluation according to NINCDS-ADRDA guidelines.

Main outcome measures: Percentage agreement between neurologists' initial clinical diagnoses and pathological findings.

Results: Of 137 patients clinically diagnosed as having probable or possible AD, 123 (90%) had AD neuropathological findings; this included 29 with AD accompanied by Lewy bodies, and 14 with AD and one or more infarcts. Cases of vascular and mixed dementia (AD and infarct[s]) had lower rates of agreement with pathological findings. Possible AD cases were more likely than probable AD cases to show pathological features other than AD. Clinicians predicted the presence or absence of AD pathological findings significantly better than chance. In patients with AD pathological lesions, older age of onset and male gender were significantly associated with shorter duration from disease onset to death.

Conclusions: Clinicians accurately predicted AD pathological findings or their absence in most cases. Attributing other degenerative dementias to AD, misdiagnosing patients with combined AD and Lewy bodies and misjudging the vascular contribution to dementia were the major areas of inaccuracy. Formal criteria for dementia associated with non-AD lesions, Lewy bodies, and infarcts need to be developed and tested.