Foster N, King R, Wang A, Landau S, Jagust W, Chen K, Reiman E.
Diagnostic Classification with Amyloid PET and FDG-PET among Clinically Diagnosed Alzheimer’s Disease Patients in the Alzheimer’s Disease Neuroimaging Initiative.
Human Amyloid Imaging Abstract. 2012 Jan 1;
Objective: To determine the degree of diagnostic concordance between amyloid PET and FDG-PET scans in subjects diagnosed with probable Alzheimer’s disease (AD) based on clinical history and examination in the Alzheimer’s Disease Neuroimaging Initiative (ADNI).
Methods: Initial amyloid PET was performed between/07-10/11 in2 subjects with mild probable AD at the time of their scans (38 men,4 women) using either1C-PIB (19 subjects) or8F-florbetapir (43 subjects). Preliminary analysis was performed on2 of these subjects. Amyloid binding in cortical gray matter relative to cerebellar gray matter was calculated and scans were categorized as either AD-like or not AD-like based upon this ratio using threshold values derived from our previously published studies of1C-PIB. Concurrent FDG-PET scans on these subjects were analyzed using-dimensional stereotactic surface projection (3D-SSP) mapping. Two experienced raters independently reviewed the metabolic and statisticalD-SSP maps and classified them based upon visual interpretation using our previously published criteria. We then evaluated the degree of concordance of amyloid PET and FDG-PET scan classifications in each subject.
Results: Among the2 clinically probable AD subjects evaluated in this preliminary analysis, (23%) had amyloid scans classified as negative and (14%) had FDG-PET scans with a pattern of hypometabolism that was not AD-like. In only of these cases was there correspondence between amyloid PET and FDG-PET classification.
Conclusions: The results of both FDG-PET and amyloid PET scans raise concerns about the accuracy of clinical diagnosis in ADNI AD subjects. Additional analyses are needed to understand the source of this unexpected discrepancy, which could include analytic techniques, variability in image acquisition and choice of diagnostic thresholds. This will be critical for implementing these techniques as diagnostic biomarkers. FDG-PET and amyloid PET contribute different and complementary information that can be exploited in clinical trials.