PET imaging of amyloid with Florbetapir F 18 and PET imaging of dopamine degeneration with 18F-AV-133 (florbenazine) in patients with Alzheimer's disease and Lewy body disorders

Abstract

Background: Biomarkers based on the underlying pathology of Alzheimer's disease (AD) and Dementia with Lewy Bodies (DLB) have the potential to improve diagnosis and understanding of the substrate for cognitive impairment in these disorders. The objective of this study was to compare the patterns of amyloid and dopamine PET imaging in patients with AD, DLB and Parkinson's disease (PD) using the amyloid imaging agent florbetapir F 18 and 18F-AV-133 (florbenazine), a marker for vesicular monamine type 2 transporters (VMAT2).

Methods: Patients with DLB and AD, Parkinson's disease (PD) and healthy controls (HC) were recruited for this study. On separate days, subjects received intravenous injections of florbetapir, and florbenazine. Amyloid burden and VMAT2 density were assessed quantitatively and by binary clinical interpretation. Imaging results for both tracers were compared across the four individual diagnostic groups and for combined groups based on underlying pathology (AD/DLB vs. PD/HC for amyloid burden and PD/DLB vs. AD/HC for VMAT binding) and correlated with measures of cognition and parkinsonism.

Results: 11 DLB, 10 AD, 5 PD, and 5 controls participated in the study. Amyloid binding was significantly higher in the combined AD/DLB patient group (n = 21) compared to the PD/HC groups (n = 10, mean SUVr: 1.42 vs. 1.07; p = 0.0006). VMAT2 density was significantly lower in the PD/DLB group (n = 16) compared to the AD/ HC group (n = 15; 1.83 vs. 2.97; p < 0.0001). Within the DLB group, there was a significant correlation between cognitive performance and striatal florbenazine binding (r = 0.73; p = 0.011).

Conclusions: The results of this study show significant differences in both florbetapir and florbenazine imaging that are consistent with expected pathology. In addition, VMAT density correlated significantly with cognitive impairment in DLB patients (ClinicalTrials.gov identifier: NCT00857506, registered March 5, 2009).

Figure 1

Scatterplots showing a) average cortical florbetapir SUVr values for each group and b) florbenazine binding for the lowest posterior putamen region for each group. See text for description of discrimination between groups.

Figure 2

Pattern of scan results for florbetapir and florbenazine for subjects in the 4 diagnostic groups. The vertical line shows the florbenazine SUVr cut-off (2.12) that agrees best with expert visual interpretation. The horizontal line (SUVr = 1.1) marks the published quantitative cut-off for florbetapir. The distribution of imaging patterns (proportion of subjects in each quadrant) differed significantly between groups (chi2 = 41.7, p <0.001) In the AD, PD and HC groups, there was one dominant pattern (for example, all of the AD cases had positive florbetapir scans and negative florbenazine scans). While VMAT2 binding was low in the DLB group on average, three subjects had SUVr values above the cut-off, and no single imaging pattern was observed in a majority of DLB patient.