Neuropathological correlates of volumetric MRI in autopsy-confirmed Lewy body dementia

Abstract

The objective of this study was to determine the neuropathological correlates of regional medial temporal lobe volume measures on magnetic resonance imaging (MRI) in subjects with Lewy body dementia (LBD). Twenty-three autopsy-confirmed LBD cases with an MRI scan close to death (mean 1.5 years) were studied. MRI-based volumetric measures were calculated for total intracranial volume, hippocampus, entorhinal cortex, and amygdala. Quantitative neuropathological analysis of plaques, tangles, and Lewy bodies were carried out in the same regions. Spearman's rho was used to examine correlations between MRI volumes and neuropathology measures and linear regression to assess the relationship between neuropathology and MRI volumes. A significant inverse correlation was observed between normalized amygdala volume and percent area of Lewy bodies in the amygdala (r = −0.461, p = 0.035). There were no other significant correlations between regional MRI volume and measures of neuropathology. Lewy body, but not Alzheimer's disease (AD) pathology was associated with reduced amygdala volume in pathologically-verified LBD cases but neither Lewy body nor Alzheimer's disease pathology was associated with volume loss in the hippocampus or entorhinal cortex, suggesting other neuropathological factors account for atrophy in these structures in LBD.

Introduction

Lewy body dementia (LBD) is a term often used to describe a group of neurodegenerative disorders that share common clinical and pathological features. Dementia with Lewy bodies (DLB) and Parkinson's disease dementia (PDD) are the most common forms and may represent different points on the spectrum of LBD (Lippa et al., 2007). Clinically, DLB and PDD are characterized by fluctuating cognition, visual hallucinations, attentional dysfunctions, and parkinsonism (Lippa et al., 2007, McKeith et al., 2005). Given the similarities, distinction between the two is based purely on the duration of parkinsonian features prior to the development of dementia. Thus, typically duration of parkinsonism 12 months or more preceding the cognitive decline warrants a diagnosis of PDD.

Neuropathologically the relatively widespread presence of Lewy bodies differentiates LBD from other dementia syndromes such as Alzheimer's disease (AD) at postmortem examination. However, it has been widely reported that Alzheimer type pathology (most often amyloid plaques, but also to lesser extent neurofibrillary tangles) often coexists with Lewy body pathology (Armstrong et al., 2000, Mann et al., 1998). An association between diffuse plaques and cognitive impairment in AD has also been described (McKeel et al., 2004). Because diffuse (rather than neuritic) plaques are typically observed in DLB, it remains unclear as to the role they play in its clinical phenotype. Lewy bodies have also been described in many AD cases, where they are (in contrast to DLB) limited to the amygdala and may occur late in the disease process (Hamilton, 2000, Lippa et al., 1998, Sahin et al., 2006).

Neuroimaging studies in DLB have observed gray matter atrophy involving the temporal, parietal, and occipital lobes (Beyer et al., 2007, Burton et al., 2002, Sanchez-Castaneda et al., 2009) and atrophy in the region of the basal forebrain (Whitwell et al., 2007) which is in contrast to the gray matter loss typically seen in the medial temporal and temporo-parietal lobes of AD (Barber et al., 2001, Karas et al., 2003). Although hippocampal atrophy is a sensitive marker for AD (seen in 80%–90% of subjects) (Scheltens et al., 1992) it is not specific as it has been reported in other dementias including vascular dementia (VaD) (Barber et al., 1999) and fronto-temporal dementia (FTD) (Galton et al., 2001).

The entorhinal cortex (EC) which forms part of the medial temporal lobe is affected by tangle pathology at an early stage of AD (Braak and Braak, 1991). EC volume has been shown to be reduced in AD (Bobinski et al., 1999, Killiany et al., 2002), DLB, and PDD compared with controls (Kenny et al., 2008). The amygdala is part of the limbic system and has a major role in memory and learning. Of the few imaging studies that have examined the amygdala in DLB, atrophy has been reported in some (Hashimoto et al., 1998) but not all studies (Barber et al., 2000, Burton et al., 2002) when compared with controls. A postmortem study reported prominent atrophy of the amygdala in cases with Lewy body disease (Cordato et al., 2000), that may be attributable either to the early involvement (Braak et al., 2003) or being 1 of the most neuropathologically affected brain areas in the disease process (Beach et al., 2009). Furthermore, the early pathological changes (Lewy bodies and Lewy neurites) incurred in the amygdala, alongside other limbic structures (e.g., hippocampus, anteromedial temporal mesocortex) are now considered to precede the more global cognitive changes in Lewy body diseases (Braak et al., 2003).

Similarly, studies examining amygdala volume in other dementias have shown conflicting results. Thus, there have been reports of reduced amygdala volume in AD compared with DLB (Barber et al., 2000, Burton et al., 2002) though this is not true of all studies (Hashimoto et al., 1998). There have been reports of no difference in amygdala volume between PDD and DLB (Beyer et al., 2007, Burton et al., 2004, Sanchez-Castaneda et al., 2009, Summerfield et al., 2005). Similarly, the Lewy body pathology appears to be a more constant finding in amygdala in both familial and sporadic AD (Tsuang et al., 2006) as well as in some control subjects (Beach et al., 2009).

Correlations between hippocampal volume reduction and tangle pathology have been shown using postmortem (Gosche et al., 2002, Huesgen et al., 1993) and antemortem magnetic resonance imaging (MRI) (Csernansky et al., 2004, Jack et al., 2002) in AD. It is widely accepted that hippocampal volume in DLB is relatively preserved compared with AD and this may be due to less coexisting AD pathology. However this may not be true of other dementias, for example in fronto-temporal dementia (FTD), severe hippocampal volume loss is often reported in the absence of AD pathology. There is a need to determine the cause of volume loss in LBD and previous studies have not examined volumetric magnetic resonance (MR) in relation to neuropathology in LBD. The aim of this study was to use a unique cohort of autopsy-confirmed LBD cases and investigate the relationship between regional magnetic resonance imaging (MRI) volumes in hippocampi, amygdala, and entorhinal cortex and the burden of neuropathology in these areas.