Background: Few studies have compared the accuracy of [(18)F]fluorodeoxyglucose (FDG) PET to the accuracy of clinical and pathologic diagnosis in dementia patients.
Methods: Forty-four individuals with dementia, cognitive impairment, or normal cognitive function underwent clinical initial evaluation (IE) and PET scanning and were followed up for approximately 4 years until a final evaluation (FE) and 5 years until death and autopsy. Clinical, pathologic, and imaging diagnoses were categorized as Alzheimer disease (AD) or not AD.
Results: Sensitivity of the IE for the pathologic diagnosis of AD was 0.76, and specificity was 0.58; PET had values of 0.84 and 0.74, and FE had values of 0.88 and 0.63. Positive predictive values for IE, PET, and FE were 0.70, 0.81, and 0.76. Negative predictive values were 0.65, 0.78, and 0.80. The diagnosis of AD was associated with a 70% probability of detecting AD pathology; with a positive PET scan this increased to 84%, and with a negative PET scan this decreased to 31%. A diagnosis of not AD at IE was associated with a 35% probability of AD pathology, increasing to 70% with a positive PET scan.
Conclusions: As a diagnostic tool, PET is superior to a baseline clinical evaluation and similar to an evaluation performed 4 years later. Although the addition of [(18)F]fluorodeoxyglucose PET to a clinical diagnosis provides useful information that can affect the likelihood of detecting Alzheimer disease pathology, the value of this technique in the current clinical environment with limited therapeutic options is likely to be modest.