Objective: We aimed to assess sensitivity and specificity of the updated criteria for behavioral variant frontotemporal dementia (bvFTD) based on a large autopsy-confirmed cohort of patients with dementia.
Methods: Two hundred thirty-nine consecutive pathologically confirmed dementia patients, clinically assessed in a specialist cognitive unit were identified. Patients with predominant aphasia, motor disorders, or insufficient clinical information were excluded. Frontotemporal Dementia Consensus criteria were applied to anonymized clinical data taken from patients' initial assessment by raters who were blinded to clinical and pathologic diagnosis.
Results: The final study cohort comprised 156 patients with predominantly early-onset dementia. The updated criteria for possible bvFTD had a sensitivity of 95% and specificity of 82%. Probable bvFTD criteria had a sensitivity of 85% and specificity of 95%. False positives were predominantly patients with presenile Alzheimer disease.
Conclusion: Revised diagnostic criteria show encouragingly high sensitivity and specificity when applied to patients with early-onset dementia. They therefore provide a useful tool both for specialist researchers and general clinicians. There is a need for further prospective studies of sensitivity and specificity involving a broader spectrum of patients with dementia.