Abstract
1789
Objectives Recent guidelines integrate biomarkers into diagnostic algorithms for dementia. Since the impact of these recommendations has not been fully evaluated with autopsy data, we evaluated the utility of PiB and FDG PET in a group of subjects with brain autopsy.
Methods Brain autopsies from 28 subjects in the Mayo Clinic Study of Aging and Mayo Clinic Alzheimer’s Disease Research Center were analyzed. Autopsy diagnoses were compared to PiB and FDG imaging results (scans were performed 0.19 to 4.8 years before death).
Results Autopsy diagnoses included AD (N=12), vascular disease (VaD) (N=4), PSP (N=4), LBD (N=3), normal (N=3), and metastatic cancer (MC) (N=1). PiB categorized all subjects correctly except for the MC subject (PiB of 1.6) and 1 normal subject (PiB of 1.9). Of the 12 AD subjects, 2 had a clinical diagnosis of LBD prior to death and one of these also had mixed AD/DLBD pathology; FDG and PiB data helped categorized both of these subjects in concordance with autopsy findings. Three of four PSP, VaD and the remaining normal subjects all had normal PiB scans. Of the 4 PSP subjects, 2 had a consensus diagnosis of AD prior to death and 2 of MCI; PIB was negative in 3 (1.19-1.23) and positive in one with mixed LBD pathology. Of the VaD subjects, 2 had a consensus diagnosis of MCI; both were negative on PiB (1.17-1.25).
Conclusions PiB and FDG biomarkers were concordant with autopsy diagnoses in 8 subjects (4 PSP, 2 AD, and 2 VaD) who had alternate or MCI consensus diagnoses antemortem. Therefore, in 29% (8/28) of subjects, PiB and FDG biomarkers improved disease characterization. These data support the integration of FDG and PiB PET in dementia diagnostic guidelines.