Objective: Appropriate use criteria (AUC; Johnson et al, 2013) provide guidelines for selecting patients for whom amyloid PET could be useful. This study evaluated the impact of amyloid PET on diagnosis/management in patients likely to meet AUC.

Material and methods: We examined 229 cases from a completed study of florbetapir amyloid PET (FBP-PET) in patients with a cognitive decline evaluation in whom Alzheimer’s Disease (AD) was suspected, but with < 85% confidence in the diagnosis. All cases received a provisional diagnosis and management prior to FBP-PET. Information for 172 cases after 3-months’ follow-up was also available on actual diagnosis/management post-FBP-PET. Cases were classified as likely meeting AUC (AUC-like) or not.

Result: 125/229 (55%) subjects were AUC-like. NonAUC cases included typical AD, Mild Cognitive Impairment (MCI) due to AD, Cognitive Decline without objective evidence of impairment (CD) and dementia or cognitive impairment with specific nonAD diagnosis. 59/125 (47%) AUC-like cases were amyloid positive (Aβ+). Among nonAUC cases, 29% (CD), 49% (MCI due to AD), 53% (non-AD) and 73% (typical AD) were Aβ+. Of 172 cases with follow-up information, diagnosis/management changed after FBP-PET in 58%/88% and 45%/77% of AUC-like and nonAUC, respectively.

Conclusions: FBP-PET altered diagnosis/management in patients selected according to AUC. Additionally, AUC generally excluded patients with a relatively high (typical AD) or low (CD) probability of Aβ+ scan.