Characterizing regional correlation, laterality and symmetry of amyloid deposition in mild cognitive impairment and Alzheimer's disease with Pittsburgh Compound B

Abstract

We evaluated the region-to-region correlation, laterality and asymmetry of amyloid deposition in subjects with mild cognitive impairment (MCI) or Alzheimer's disease (AD) using the amyloid tracer, Pittsburgh Compound B (PiB). Seventeen subjects, including 7 with MCI (MMSE 26.7 ± 2.4) and 10 with AD (MMSE of 24.8 ± 2.7) underwent PiB imaging. Measures of laterality (i.e., group-wise predilection for right or left) and asymmetry (i.e., group-wise predilection for unequal PiB retention between the two hemispheres) were calculated for 17 Regions of Interest (ROIs). Regional correlations were calculated along with within-group and between-groups statistical analyses of laterality and asymmetry metrics. The median correlation between PiB retention across all pairs of ROIs was 0.65, with highest correlations found in areas of highest PiB retention (r = 0.74). Overall, PiB retention was symmetric bilaterally, but there was PiB laterality in MCI in dorsal frontal cortex [(t(6) = 3.05, p = 0.02, L > R] and sensory-motor area [t(6) = 3.10, p = 0.02, L > R] and in AD in the occipital pole (t(9) = −2.63, p = 0.03, R > L). The most significant asymmetries in PiB retention were found in sub-cortical white matter (t(6) = 3.99, p = 0.01) and middle precuneus [(t(6) = 3.57, p = 0.01] in MCI, and in lateral temporal cortex (t(9) = 3.02, p = 0.01) and anterior ventral striatum [t(9) = 2.37, p = 0.04] in AD. No group differences (AD versus MCI) were detected in laterality [F (1, 15) = 0.15, p = 0.7] or asymmetry [F (1, 15) = 0.7, p = 0.42].

Introduction

Alzheimer's disease (AD) is the most common cause of clinical dementia of the elderly, diagnosed definitively after autopsy based upon the presence of two pathologic hallmarks: neurofibrillary tangles and amyloid plaques (Khachaturian, 1985, Mirra et al., 1991, NIA/Reagan Workgroup, 1997). Structural neuroimaging can be used to identify characteristic patterns of atrophy and to exclude other clinical conditions in order to support the diagnosis of AD (Husain and Garrett, 2005). However, molecular imaging of individuals at risk for AD has the potential to allow for definitive, even pre-symptomatic, diagnosis, evaluation of the effect of disease modifying drugs, and provide a better understanding of the underlying pathophysiology of the disease (Jagust, 2004). Such work also may result in a better understanding of mild cognitive impairment (MCI), which is thought to represent a transition phase between healthy aging and dementia (Petersen, 2004).

One such molecular imaging tool is Pittsburgh Compound B (PiB), a thioflavin-T derivative that crosses the blood brain barrier and binds with high affinity to amyloid allowing for in vivo visualization using positron emission tomography (PET) (Klunk et al., 2004). The purpose of this study was to better understand patterns of amyloid deposition in MCI and AD, as measured by PiB PET, with respect to correlation of PiB retention across brain areas, the laterality of this retention, and the symmetry across hemispheres. To our knowledge, there have been no neuropathological or neuroimaging studies addressing the issue of region-to-region correlation, laterality and symmetry of amyloid deposition. In the case of neuropathology research studies, this may be because only one side of the brain is typically examined histopathologically at autopsy (Braak et al., 2006). PiB PET is an ideal tool to measure symmetry and laterality of amyloid deposition, since the whole brain can be assessed simultaneously in vivo. The presence or absence of symmetrical amyloid deposition and region-to-region correlation of amyloid deposition have implications for the natural history of AD. Is amyloid deposition a global, symmetric phenomenon, that correlates across most of the brain, or does it appear lateralized on one side of the brain or in certain brain regions? For this study, we focused only on persons with AD or MCI as they consistently have measurable PiB deposition. Cognitively normal subjects were not included in this study as 75–80% lack detectable amyloid burden as measured by PiB (Klunk et al., 2004, Mintun et al., 2006).