UCSF

Objective

To evaluate the effect of amyloid imaging on clinical decision making.

Methods

We conducted a retrospective analysis of 140 cognitively impaired patients (mean age 65.0 years, 46% primary β-amyloid (Aβ) diagnosis, mean Mini-Mental State Examination 22.3) who underwent amyloid (Pittsburgh compound B [PiB]) PET as part of observational research studies and were evaluated clinically before and after the scan. One hundred thirty-four concurrently underwent fluorodeoxyglucose (FDG)-PET. We assessed for changes between the pre- and post-PET clinical diagnosis (from Aβ to non-Aβ diagnosis or vice versa) and Alzheimer disease treatment plan. The association between PiB/FDG results and changes in management was evaluated using χ(2) and multivariate logistic regression. Postmortem diagnosis was available for 24 patients (17%).

Results

Concordance between scan results and baseline diagnosis was high (PiB 84%, FDG 82%). The primary diagnosis changed after PET in 13/140 patients (9%) overall but in 5/13 (38%) patients considered pre-PET diagnostic dilemmas. When examined independently, discordant PiB and discordant FDG were both associated with diagnostic change (unadjusted p < 0.0001). However, when examined together in a multivariate logistic regression, only discordant PiB remained significant (adjusted p = 0.00013). Changes in treatment were associated with discordant PiB in patients with non-Aβ diagnoses (adjusted p = 0.028), while FDG had no effect on therapy. Both PiB (96%) and FDG (91%) showed high agreement with autopsy diagnosis.

Conclusions

PET had a moderate effect on clinical outcomes. Discordant PiB had a greater effect than discordant FDG, and influence on diagnosis was greater than on treatment. Prospective studies are needed to better characterize the clinical role of amyloid PET.