OBJECTIVE:

The authors explored whether positron emission tomography (PET) with 2-(1-{6-[(2-[fluorine-18]fluoroethyl)(methyl) amino]-2-naphthyl} ethylidene)malononitrile (FDDNP), a molecule that binds to plaques and tangles in vitro, might identify homogeneous subgroups of persons in middle-aged and older persons with mild cognitive impairment (MCI) or normal cognition.

PARTICIPANTS:

Fifty-six subjects (MCI, N = 29; normal cognition, N = 27).

MEASUREMENTS:

FDDNP-PET scans were performed. Logan parametric images were produced using cerebellum as a reference region, and relative distribution volumes were obtained for regions of interest (ROIs) known to accumulate plaques and tangles in Alzheimer disease (AD). Cluster analysis was used to identify subgroups of subjects according to FDDNP signal distribution. Once the FDDNP clusters were identified, the authors then characterized the clusters also with respect to diagnosis and cognitive test performances and conducted analyses on cluster differences in these variables.

RESULTS:

The authors identified three FDDNP clusters: high signal in lateral temporal and posterior cingulate ROIs (high temporal-posterior cingulate HT/PC); low signal in all ROIs (low global [LG] cluster); high frontal and parietal signal with intermediate temporal and posterior cingulate signal (HF/PA). Most MCI subjects belonged to the HT/PC and HF/PA clusters, whereas most cognitively normal subjects were in the LG cluster. On cognitive tests, the HT/PC and the HF/PA clusters performed significantly worse than LG but did not significantly differ from each other.

CONCLUSIONS:

This approach may be useful in identifying potential high-risk imaging cluster patterns. Longitudinal follow-up would be performed to determine the association of these subgroups with diagnostic and functional outcome.