- Email: [email protected]
Diffusion tensor imaging in Alzheimer's disease and dementia with Lewy bodies
Psychiatry Research: Neuroimaging 194 (2011) 176–183
Contents lists available at ScienceDirect
Psychiatry Research: Neuroimaging j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / p s yc h r e s n s
Diffusion tensor imaging in Alzheimer's disease and dementia with Lewy bodies Michael J. Firbank a,⁎, Andrew M. Blamire b, Andrew Teodorczuk a, Emma Teper a, Dipayan Mitra c, John T. O'Brien a a b c
Institute for Ageing and Health, Newcastle University, Newcastle upon Tyne, NE4 5PL, UK Newcastle Magnetic Resonance Centre, Newcastle University, Newcastle upon Tyne, NE4 5PL, UK Department of Neuroradiology, Newcastle General Hospital, Newcastle upon Tyne, NE4 6BE, UK
a r t i c l e
i n f o
Article history: Received 10 May 2011 Received in revised form 29 July 2011 Accepted 4 August 2011 Keywords: MRI DTI DLB Connectivity Hippocampus
a b s t r a c t White matter changes have been investigated in Alzheimer's disease (AD) in a number of studies using diffusion imaging. Fewer studies have investigated dementia with Lewy bodies (DLB). We used diffusionweighted magnetic resonance imaging (MRI) and high-resolution (0.3 mm in-plane) coronal 3T MRI of the medial temporal lobe in 16 subjects with AD, 16 with DLB and 16 similarly aged healthy subjects. We found increased mean diffusivity in the temporal lobe of AD, and reduced fractional anisotropy (FA) in a small cluster in the right postcentral gyrus region in the DLB group. Mean FA in this cluster correlated with UPDRS (Unified Parkinson's Disease Rating Scale) motor score. We had previously reported reduced visibility in the AD group of a dark appearing layer of the hippocampus in the high-resolution images. In an SPM analysis on all subjects, there were significant clusters of reduced FA in the corpus callosum, fornix and stria terminalis that correlated with the visual rating of the hippocampus. These results suggest that changes to the hippocampus are associated with structural changes to the white matter fibres of the hippocampus output, and that changes in motor function are associated with changes in white matter underlying somatosensory cortex. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Dementia with Lewy bodies (DLB) is the second most common form of neurodegenerative dementia following Alzheimer's disease (AD), accounting for approximately 15% of cases at autopsy (McKeith et al., 2004). DLB shares clinical and pathological features with AD, making it potentially difficult to distinguish in clinical practice. Early and accurate diagnosis of DLB is important for optimum management, including provision of appropriate information to patients and carers, initiation of effective treatments and avoidance of potentially lifethreatening antipsychotic drugs. Diffusion imaging with magnetic resonance has been widely used to investigate the integrity of the white matter microstructure. Mean diffusivity (MD) of water is typically higher where there are fewer barriers to diffusion such as cell walls. Fractional anisotropy (FA) indicates the degree of angular variation in the magnitude of water motion (diffusion), and is highest in directionally coherent fibre bundles such as those found in corpus callosum. Change in MD and FA of the frontal, temporal and parietal lobes has been observed in AD (Kantarci et al., 2001; Bozzali et al., 2002; Head et al., 2004; Naggara et al., 2006; Firbank et al., 2007; Kiuchi et al., 2009) though some ⁎ Corresponding author at: Institute for Ageing and Health, Newcastle University, Wolfson Research Centre, Campus for Ageing and Vitality, Newcastle upon Tyne NE4 5PL, UK. Tel.: + 44 191 248 1319; fax: + 44 191 248 1301. E-mail address: m.j.fi[email protected] (M.J. Firbank). 0925-4927/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.pscychresns.2011.08.002
studies have found few differences between AD and comparable healthy subjects (Bozzao et al., 2001). Studies vary as to which regions are found to be altered, with, for instance, findings of only limited changes in FA of temporal lobe (Damoiseaux et al., 2009), change in temporal lobe MD (Stahl et al., 2007), and changes in frontal lobe and corpus callosum of MD and FA (Chen et al., 2008). However, diffusion changes in the temporal lobe have been consistently reported, with suggestions that the connectivity of the hippocampus is reduced in AD (Kalus et al., 2006; Ringman et al., 2007; Zhou et al., 2008). A recent study (Kiuchi et al., 2011), using tractography, found decreased FA in the uncinate fasciculus of both AD and DLB, with DLB additionally showing more posterior change. The few other studies in DLB also suggest more posterior changes (Firbank et al., 2007; Ota et al., 2009; Kantarci et al., 2010), with the temporal lobe relatively unaffected, in keeping with structural preservation in DLB compared to AD (Whitwell et al., 2007). We hypothesised that in this study we would find decreases in FA and increases in MD principally in the temporal lobe of AD, and more parietal changes in DLB. In an analysis of high resolution coronal magnetic resonance images (MRI) of the hippocampus, we found that a structure probably representing the stratum moleculare, stratum lacunosum and stratum radiatum was less visible in AD subjects than in DLB (Firbank et al., 2010). Neuropathological studies have found that this layer is one of the earliest affected by Alzheimer type pathology (Lace et al., 2009) and that pathology spreads through the hippocampus in stages defined broadly by the internal connectivity — i.e. from input to
M.J. Firbank et al. / Psychiatry Research: Neuroimaging 194 (2011) 176–183
output layers. We hypothesised that these hippocampal changes would be associated with changes to the white matter (WM) connecting the hippocampus to other brain areas, in particular, the fornix, which is the main output from the hippocampus, and the cingulum bundle, which is the main input (Duvernoy, 1998; Mori et al., 2005). Although medial temporal lobe atrophy is less common in DLB, it is still present (Tam et al., 2005), and we hypothesised that any alterations in connecting WM would be present regardless of the origin of the hippocampal changes. In this study, we used diffusion-weighted imaging to investigate differences in WM integrity between AD, DLB and healthy subjects. We also investigated the relationship between the changes seen on the high-resolution hippocampus imaging and FA changes in WM. 2. Methods 2.1. Participants We recruited 16 people with Alzheimer's disease and 16 with dementia with Lewy bodies, from clinical Old Age Psychiatry, Geriatric Medicine and Neurology Services. Sixteen healthy subjects of similar age were also recruited from spouses and friends of participants with dementia, as well as from a register of subjects who have previously indicated willingness to participate in research. Subjects are the same as those in our previous article (Firbank et al., 2010). All subjects were aged over 60 and did not have contra-indications for MRI. Subjects with dementia had mild to moderate severity (MMSE N 10). All Alzheimer's disease subjects fulfilled criteria for probable AD according to NINCDS/ADRDA (McKhann et al., 1984). Cases of dementia with Lewy bodies all met criteria for probable DLB according to the consensus criteria (McKeith et al., 2005). All diagnoses were made by consensus between two experienced clinicians, a method we have previously validated against autopsy diagnosis (McKeith et al., 2000). Routine clinical workup for dementia included detailed physical, neurological and neuropsychiatric examinations, including screening blood tests and CT scan. Additional assessments performed were of cognition (Cambridge Cognitive Examination (CAMCOG)) (Roth et al., 1986), mood (Cornell Depression Scale) (Alexopoulos et al., 1988), neuropsychiatric features (Neuropsychiatric Inventory (NPI)) (Cummings et al., 1994), clinical fluctuation (Clinical Assessment of Fluctuation Scale) (Walker et al., 2000), memory (Rey Auditory Verbal Learning Test (Rey, 1964)) and motor features of parkinsonism (UPDRS (Unified Parkinson's Disease Rating Scale) subsection III, recorded for each side of the body) (Fahn et al., 1987). Nine of the DLB subjects were taking anti-parkinsonian medications. The UPDRS assessment was not timed relative to subjects taking parkinsonian medication. Duration of dementia was determined from a review of the patient's medical case notes. Exclusion criteria included severe concurrent illness (apart from dementia for patients), space-occupying lesions on imaging, history of stroke and contraindications to MRI. In addition, controls had no history of psychiatric illnesses. The study was approved by the local ethics committee, and all subjects gave signed informed consent for participation. 2.2. MRI acquisition Subjects were scanned on a 3T MRI system (Intera Achieva scanner; Philips, Eindhoven, the Netherlands). Images acquired included a T1 weighted volumetric sequence covering the whole brain (MPRAGE, Sagittal acquisition, slice thickness 1.2 mm, voxel size 1.15 × 1.15 mm; TR = 9.6 ms; TE 4.6 ms; flip angle = 8 o; SENSE factor = 2). Diffusion images were acquired with FLAIR weighting to reduce the influence of CSF. Acquisition parameters were TR 7000 ms, TE 68 ms, TI 2200 ms. SENSE factor = 2, slice thickness = 2.5 mm, field of
177
view 260 × 260 mm, acquisition matrix 120 × 93. Two b values were used: 4 acquisitions with b = 0 s/mm 2, and images with diffusion weighting in 30 directions (1 acquisition each) with b = 1000 s/mm 2, with the gradient directions uniformly spaced around a sphere. High resolution coronal imaging was also performed, based on previous work (Mueller et al., 2007) with a sequence optimised locally to provide good hippocampal contrast (turbo spin echo — turbo factor 15; 24 slices; slice thickness 2 mm, field of view 210 × 167; pixel resolution 0.41 × 0.52 mm; TR 2568 ms; TE 19 ms; centric-ordered, flip angle 90 o; 2 acquisitions — acquisition time = 2 × 2:50). After the first 13 subjects (5 Control, 7 AD, 1 DLB) one of the acquisitions was replaced by three acquisitions of a higher resolution sequence for the remaining subjects to improve visualisation of the hippocampus structure (Firbank et al., 2010) with the following parameters altered: 12 slices; pixel resolution 0.27 × 0.35 mm; TR 3852 ms; 3 acquisitions — acquisition time= 3 × 2:07. The number of acquisitions was increased to maintain signal-to-noise ratio (SNR) in the face of smaller voxels. Data were acquired using multiple acquisitions to allow correction of patient motion prior to averaging to increase SNR. This approach was found to maintain highest resolution in pilot studies compared with direct averaging by the scanner. The coronal images were positioned for each subject so that they were at an angle of 25 o to the line from genu to splenium of the corpus callosum. We have found previously that this is a reliable method of angling the slices approximately perpendicular to the main axis of the hippocampus (Firbank et al., 2010). 2.3. Volumetric image analysis The T1 weighted anatomical images were segmented into grey and white matter, and spatially normalised using the segmentation algorithm in SPM5 (http://www.fil.ion.ucl.ac.uk/spm/). The ratio of total brain volume/intracranial volume was derived from the ratio of grey+white/grey+white+CSF volume. 2.4. High resolution image analysis The analysis of the high resolution coronal images is described elsewhere (Firbank et al., 2010). Briefly, we used the FLIRT image registration tool (Jenkinson and Smith, 2001)(part of FSL http://www. fmrib.ox.ac.uk/fsl/) to register all the high resolution images from each subject together, and interpolate to 0.27 × 0.27 mm resolution. A high SNR image was then created by summing together all the registered high resolution images for that subject. Images were reviewed to check for subject motion, and one subject (DLB) was excluded due to excessive motion, and for one other subject, one of the high resolution images was unacceptable. The other subjects did not have excessive motion artefacts on the images. Images were viewed with the freely available itk-snap package (Yushkevich et al., 2006) (http://www.itksnap.org/pmwiki/pmwiki.php). We examined three coronal slices, starting on the slice on which the head of the hippocampus was no longer visible, and the two slices posterior to that. A hypointense line could be seen in the hippocampus which is likely to represent fibres in the hippocampal layers of stratum moleculare, stratum lacunosum and stratum radiatum (Wieshmann et al., 1999; Thomas et al., 2008) (see Fig. 1). We observed considerable variability in how clearly this line could be visualised between subjects, and a single rater (blind to diagnosis) assigned each hippocampus a score (1 to 5) according to how clearly the hippocampus internal structure was depicted throughout the three slices examined. On this scale, 5 = line clearly visualised throughout, 4 = most of the line clearly visualised with good contrast for most of its length, 3 = line semi clearly defined, either with some sections of good contrast and some poor, or partly blurred along all its length, 2 = line mostly not clearly defined, but recognisable, typically with
178
M.J. Firbank et al. / Psychiatry Research: Neuroimaging 194 (2011) 176–183
Fig. 1. Coronal high resolution image, with location of hypointense line indicated by an arrow. On the left two subjects with well depicted hypointense line, on the right, two subjects with very poorly depicted line.
poor contrast or blurred throughout its length, 1 = line not visualised at all. We previously reported reliability of this rating scale (Firbank et al., 2010) — intraclass correlation coefficient = 0.97 for within rater, and 0.8 for between rater.
was avoided (see Fig. 2A). We also drew a ROI in the corticospinal tract, where we did not expect to see correlation with hippocampus rating. The ROI (2× 2 voxels) was drawn just superior to the corpus callosum, (Fig. 2B) using the principal eigenvector image to select voxels with an inferior-superior direction of diffusion.
2.5. Diffusion image analysis 2.6. Statistics To correct the diffusion-weighted images for the distorting effect of eddy currents, we adapted the approach of Shen (Shen et al., 2004) and used an affine registration (in FSL's FLIRT) to register each pair of opposite diffusion-weighted images together. The eddy-corrected diffusion-weighted images were then registered with a rigid body registration to the b = 0 s/mm 2 image. Mean diffusivity (MD) and fractional anisotropy (FA) images were calculated using FSL tensor analysis of the aligned diffusion-weighted images. Further processing of the images was done with SPM5. The b = 0 s/ mm 2 FLAIR-weighted image was coregistered with the T1-weighted anatomical image, and the spatial normalisation parameters derived during the segmentation process above were applied to the FA and MD images. Images were smoothed with a Gaussian FWHM of 5 mm. An average FA image was generated and thresholded at 0.2 to generate a binary mask for use with the voxelwise statistical analysis. To compare diffusion parameters between the three subject groups, we performed a voxelwise analysis of variance (ANOVA) in SPM5 of both FA and MD. In a previous report (Firbank et al., 2010), we describe our finding that the hippocampus viewed on the high resolution coronal images had less well defined structure on the visual rating scale in the Alzheimer's disease group. We hypothesised that this might relate to loss of integrity of white matter fibres within and connecting the hippocampus. To investigate this, we performed a regression of FA images in SPM, including data from all subjects, with covariates of disease group, brain/ICV ratio and the hippocampus visual rating scale described above. To verify that correlations with hippocampus rating seen on the voxelwise analysis were not merely a result of brain atrophy, we additionally measured mean FA within a region of interest (ROI) in the fornix: on four adjacent coronal slices in the body of the fornix, a voxel in the centre of the fornix was selected. Regions were drawn individually on each subject's native space b = 0 s/mm2 image, ensuring that CSF
We used ANOVA to compare demographic factors between groups with Tukey post hoc comparisons. Chi squared test was used to compare sex. All statistical analysis was performed with Minitab 15 (Minitab Inc, Pennsylvania, USA). The following three SPM analyses were performed, including all subjects: 1) An ANOVA model of MD by disease group. 2) An ANOVA model of FA by disease group. 3) A regression model of FA with disease group, brain/ICV ratio and hippocampus rating scale. Results
Fig. 2. A) midline sagittal section showing the fornix ROI. B) coronal and axial section showing the corticospinal tract ROI.
M.J. Firbank et al. / Psychiatry Research: Neuroimaging 194 (2011) 176–183
179
Table 1 Subject demographics. Control
AD
DLB
N = 16
N = 16
N = 16
Sex (female: male) Education in years Education post 16 (yes: no) MMSE Duration dementia (months)
76.3 (8.2) [61–93] 7: 9 11.5 [9–18] 8: 8 29 [26–30] –
UPDRS CAMCOG NPI total Rey total trials 1–5 (max 75) Rey delayed recall (max 15) Fluctuation score Hypertension yes: no Diabetes yes: no Intracranial volume (ml)
2.0 [0–14] 97 (3.5) – 42 [30–61] 8 [5–14] 0 [0] (n = 4) 7: 9 1: 15 1504 (150)
77.3 (8.9) [64–94] 8: 8 10.5 [9–16] 5: 11 21.5 [16–27]a 40.4 (25) [6–72] 5.5 [1–13]a 69 (11.4)a 8.5 (11.8) 21 [5–31]a 0 [0–3]a 0 [0–9] 6: 10 2: 14 1449 (159)
81.0 (5.9) [70–88] 6: 10 9.0 [8–10]b,c 0: 16b,c 18 [15–27]c 43.7 (24) [3–96] 17.5 [9–33]b,c 67.4 (14.0)c 24.1 (11.7)b 18 [4–36]c 1.0 [0–8]b,c 7 [0–16]b,c 6: 10 0: 16 1472 (140)
Age in years
§F = 1.6; p = 0.2 † p = 0.9 ‡ H = 18; p = 0.001 † p = 0.004 ‡ H = 31; p b 0.001 t = 0.16; p = 0.9 ‡ H = 32; p b 0.001 ‡ H = 32 ; p b 0.001 t = 3.6 ; p = 0.001 ‡ H = 31 ; p b 0.001 ‡ H = 35 ; p b 0.001 ‡ H = 13 ; p = 0.002 † p = 0.7 † p = 0.3 §F = 0.56, p = 0.6
Values in the table are mean (S.D.) or median [range]. Post hoc p b 0.05 (a) AD vs. Control; (b) DLB vs. AD; (c) DLB vs. Control. § ANOVA. † Fisher's exact test. ‡ Kruskal–Wallis. MMSE=Mini Mental State Examination; CAMCOG=Cambridge Cognitive Examination; UPDRS=Unified Parkinson's Disease Rating Scale (subsection 3); NPI=Neuropsychiatric Inventory; Rey=Rey Auditory Verbal Learning Test; Fluctuation score=Clinical Assessment of Fluctuation Scale.
from the SPM analysis were thresholded at p = 0.001 uncorrected for multiple comparisons, and we report significant clusters with p b 0.05 corrected for multiple comparisons with familywise error correction. 3. Results Subject demographics are described in Table 1. There were no differences between groups in age or sex distribution. Duration of dementia and CAMCOG score did not differ between AD and DLB. The SPM group ANOVA analysis results are summarised in Table 2 and Fig. 3, which report those clusters that were significant after correcting for multiple comparisons. The comparison of mean diffusivity gave two significant clusters (Fig. 3a) where MD was higher in AD relative to control subjects, focussed in the inferior and anterior medial temporal lobe, including the uncinate fasciculus and inferior longitudinal fasciculus. In the comparison of fractional anisotropy there was one cluster in the white matter beneath the post-central gyrus (Fig. 3b) where FA was lower in DLB relative to control. There were no other significant differences found between groups. The tables also show mean diffusivity within the clusters, with the MD values being in the range of brain tissue, demonstrating that the clusters do not contain large portions of CSF. Fig. 4 and Table 3 show the results of the SPM regression of hippocampus visual rating against FA in all subjects, controlling for brain size and disease group. There were a number of significant clusters where decreased FA correlated with poor hippocampus visual
rating. Significant clusters include those in the corpus callosum, fornix and stria terminalis. Since a number of regions were close to the ventricles, we wished to determine that the correlation was not just due to poor spatial registration resulting in variable amounts of CSF within the cluster. We determined the mean FA within each cluster from the original (unsmoothed) FA images averaging just those pixels with FA N0.25. The correlations were significant (p b =0.001) in each cluster, with the exception of the cerebellum where the significance was only p = 0.02. In addition, we repeated the spatial normalisation by calculating the mean of all the normalised FA images in the study, and (using the SPM normalisation function) renormalised the original FA maps to this average FA template. This should minimise spatial normalisation errors by using a cohort-specific template. The significant correlations with hippocampus clearness were still present in essentially the same locations in this analysis. Because the fornix is a very small structure, potentially difficult to accurately spatially normalise, we also performed a region of interest analysis, and manually positioned an ROI in the fornix, avoiding CSF (see Fig. 2a), and determined the mean FA. The mean FA in this ROI correlated (r = 0.58; p = 0.001) with hippocampus visual rating (see Fig. 5). As expected, there was no significant correlation between the FA in the corticospinal tract ROI and hippocampus rating (r = − 0.15; p = 0.3). Repeating this analysis with a covariate to control for the two high resolution sequences used did not alter the significance of the results.
Table 2 SPM ANOVA comparison of MD and FA between groups. Significant clusters (p b 0.05) corrected for multiple comparisons with familywise error rate (FWE) correction. x,y,z (MNI) AD MD N Control − 22,−6,−32 − 52,−36,−24 − 36,−44,−20 34,−4,−34 DLB FA b Control 28,−28,48
MD within cluster (×10− 3 mm2/s) Mean (S.D.)
Cluster p value (FWE)
Cluster size (mm3)
T statistic (at peak voxel)
b 0.001 Same cluster as above Same cluster as above b 0.001
2216
L temporal lobe
0.82 (0.03)
920
4.82 4.71 4.63 4.52
R temporal lobe
0.82 (0.04)
0.017
704
4.95
White matter deep to right Brodmann area 3
0.76 (0.04)
MNI = Montreal Neurological Institute brain template.
180
M.J. Firbank et al. / Psychiatry Research: Neuroimaging 194 (2011) 176–183
Fig. 3. Voxelwise comparisons of SPM results thresholded at p b 0.001 (uncorrected for multiple comparisons) with a minimum cluster size of 50 voxels. Significant clusters overlaid on an average brain. a) MD AD N Control – axial slices at 4-mm intervals starting at − 38 mm in MNI coordinates. b) FA DLB b Control (slices at MNI coordinates − 26, 28, 48 mm).
To investigate whether the lower FA in the right side post-central gyrus region of DLB was related to parkinsonian motor features in the patients, we performed a post hoc correlation between the mean FA in this cluster and the UPDRS motor score using the contralateral (i.e. left side) assessment. In both the DLB group alone (r = − 0.51; p = 0.04) and the combined AD and DLB group (r = −0.57; p =0.001), there was a correlation between decreased FA and increased UPDRS motor score (see Fig. 6). 4. Discussion The main finding of this study was that visual rating of hippocampus on high resolution MRI in subjects with AD and DLB correlated with fractional anisotropy in the fornix, stria terminalis and corpus callosum. The fornix is the main output from the hippocampus, and the stria terminalis from the amygdala. This suggests that atrophy of the medial temporal lobe is associated with degeneration of the outgoing fibres. Other groups have found decreased FA in the fornix in AD (Mielke et al., 2009; Kantarci et al., 2010), and one report (Stricker et al., 2009) concluded that late myelinating fibres, including the fornix and splenium of the corpus callosum, had lower FA in AD than controls,
Fig. 4. a. SPM results — regions where FA correlates with decreased hippocampus visual rating, controlling for disease group and brain/intracranial volume ratio. SPM thresholded at p b 0.001 uncorrected for multiple comparisons. Number on each slice indicates MNI x coordinate. b. SPM FA-hippocampus correlations overlaid onto coronal mean FA template (MNI y = − 64 mm and at 5-mm intervals).
and suggested retrogenesis. Reduction of fornix FA has also been observed in presymptomatic familial AD (Ringman et al., 2007). Villain (Villain et al., 2008) found decreased white matter density in AD in the corpus callosum, fornix and cingulum, and also correlations between grey matter density in the hippocampus region with white matter density in the cingulum and splenium of corpus callosum. This lower value of FA in the outgoing fibres could be a result of Wallerian degeneration, with primary degeneration of hippocampal neurones due to AD pathology leading to loss of output fibres, and hence downstream changes. fMRI studies of older subjects have
M.J. Firbank et al. / Psychiatry Research: Neuroimaging 194 (2011) 176–183
181
Table 3 SPM correlation of hippocampus visual rating against FA, controlling for brain size and dementia group. Significant clusters (p b 0.05) corrected for multiple comparisons with familywise error rate (FWE) correction. x,y,z (MNI)
Cluster p value (FWE)
Cluster size (mm3)
T statistic (at peak voxel)
− 18,−26,28 − 6 − 14,28 10,−18,30 26,−36,24 0,−6,10 32,44,−8 − 24,−38,2 − 18,−30,12 14,−66,−46
b 0.001 Same cluster as above b 0.001 Same cluster as above 0.040 0.020 0.017 Same cluster as above 0.040
2896
5.20 5.03 4.46 4.07 5.08 5.04 4.30 3.90 4.19
2048 576 672 688 576
MD within cluster (×10− 3 mm2/s) mean (S.D.) L periventricular WM L corpus callosum body R corpus callosum body R periventricular WM fornix R anterior frontal lobe Fornix/stria terminalis Stria terminalis R superior cerebellar peduncle
0.66 " 0.66 " 0.98 0.78 1.03 " 0.65
(0.17) (0.17) (0.18) (0.04) (0.17) (0.07)
MNI = Montreal Neurological Institute brain template.
shown that functional activity of the hippocampus decreases as memory problems get worse (O'Brien et al., 2010). The output pathway along the fornix passes via the anterior thalamic nucleus to the posterior cingulate (Duvernoy, 1998), a region that has marked reduction in cerebral blood flow and metabolism early on in both AD and DLB (Herholz et al., 2002; Firbank et al., 2003). Functional connectivity between the posterior cingulate and the hippocampus is lower in AD (Zhang et al., 2010) and this could be partly mediated through changes in the connecting white matter tracts. The corpus callosum (CC) is not directly connected to the hippocampus, though reductions in posterior CC volume and diffusivity (Di Paola et al., 2010) have been reported early in AD. The relationship with hippocampus volume may be due to shared common neuropathological causes. The periventricular WM cluster (Table 3) is in the vicinity of the tapetum, a fibre bundle which passes from the temporal lobe through the corpus callosum. The anterior frontal lobe cluster is in the region of the uncinate fasciculus, which connects the temporal lobe (including the hippocampus formation) to the frontal lobe. In both these cases, it may be that hippocampal degeneration is associated with wider damage to the temporal lobe, and hence connecting white matter fibre bundles. A number of the regions in which FA correlated with hippocampus rating were adjacent to CSF. We have endeavoured to demonstrate
that the correlations were not due to the presence of CSF. However, there remains the possibility that some of the significant clusters in the SPM analysis are regions where there is an increased proportion of CSF due to inaccurate spatial registration resulting from generalised brain atrophy. Our finding of more differences in the temporal lobe in AD compared to DLB is in keeping with previous diffusion studies (Kantarci et al., 2010) and also with studies of brain atrophy (Whitwell et al., 2007) which show marked temporal lobe changes in AD. Interestingly, the temporal lobe differences in AD were not in the same location as the correlates of hippocampus degeneration, suggesting perhaps that damage to input and output fibres of the hippocampus occurs by differing processes. We found a small cluster in the white matter underlying the postcentral gyrus where FA was lower in the DLB group relative to the control group. The mean FA in this cluster was significantly correlated with UPDRS score in both the DLB and combined dementia groups, suggesting a relationship between white matter change and motor function. Supporting this finding, a study by Lee et al. (2010) found bilateral clusters of lower FA in their Parkinson disease dementia (PDD) group (though not DLB) extending up to approximately 21, −22,40 and −22, −29, 42 in MNI coordinates close to the cluster that we observed. Other studies in DLB have used region of interest approaches, and hence this region has not been otherwise investigated. An fMRI study of fine motor function (Foki et al., 2010) found relatively decreased activation in post-central gyrus in PD, and
Fig. 5. Mean FA in manually drawn ROI in body of fornix vs. hippocampus visual rating. Correlation coefficient r = 0.58; p = 0.001.
Fig. 6. Mean FA in the right postcentral gyrus white matter cluster (MNI coordinate 28, −28,48) against left side UPDRS motor score. Correlation coefficient r = −0.57; p =0.001.
182
M.J. Firbank et al. / Psychiatry Research: Neuroimaging 194 (2011) 176–183
integration of sensory with motor function is impaired in PD (Abbruzzese and Berardelli, 2003). There is also the possibility that this finding might be related to vascular disease, which is associated with both changes to the white matter evident in diffusion imaging, and with motor signs (Sachdev et al., 2005; Della Nave et al., 2007). Our previous study in DLB found reduced FA in the white matter of the precuneus (Firbank et al., 2007). This finding receives some support from the study here — there was a bilateral pair of clusters of voxels which passed the 0.001 significance threshold (but which were not significant on the corrected cluster threshold) with reduced FA in DLB compared to control (MNI 16, −60, 54; 31 voxels, and − 16 −58, 44; 14 voxels) close to the precuneus region where we previously reported reduced FA in DLB. Limitations of the study are that we changed the hippocampus structural imaging sequence part way through the study. However, we do not feel that this will have influenced the results to a great extent — including the sequence type in the analysis did not alter the significance of the findings either in the analysis here, or in our previous report which focussed on the hippocampus imaging. We also relied upon clinical diagnosis, without neuropathological confirmation, although we have previously demonstrated good correlation between clinical and pathological diagnosis (McKeith et al., 2000). The associations between hippocampus visual rating and FA were much stronger than any differences in FA between AD or DLB and control subjects, suggesting that atrophy of the hippocampus is consistently associated with regional brain changes, but that these associated FA changes may not directly translate into functional deficits clinically. Studies have shown that the difference in FA between young and old subjects is of an equal or greater magnitude to that between subjects with dementia and old matched controls. (Head et al., 2004; Damoiseaux et al., 2009) despite there being less cognitive difference between young and old subjects than AD. It may be that DTI is sensitive to factors such as differences in brain water content which do not necessarily directly impact upon cognition. Other factors, i.e. brain pathology in other areas, neurochemical changes, etc., may therefore be important in addition. In conclusion, we found evidence of temporal lobe changes in Alzheimer's disease, with degeneration of the hippocampus associated with breakdown of the white matter in its outgoing WM fibres. We also found lower fractional anisotropy of the white matter in the post-central gyrus area of DLB, which correlated with worse motor function. Longitudinal studies starting with subjects at risk of dementia, or with very early stage disease, are necessary to determine the relative time course of these changes and their relationship to functional changes.
Acknowledgments We are grateful for funding from the Alzheimer's Research Trust. This work was supported by the UK NIHR Biomedical Research Centre for Ageing and Age-related Disease award to the Newcastle upon Tyne Hospitals NHS Foundation Trust. We also thank the North East DeNDRoN (Dementia and Neurodegenerative Diseases Research Network) team for help with subject recruitment.
References Abbruzzese, G., Berardelli, A., 2003. Sensorimotor integration in movement disorders. Movement Disorders 18, 231–240. Alexopoulos, G.S., Abrams, R.C., Young, R.C., Shamoian, C.A., 1988. Cornell scale for depression in dementia. Biological Psychiatry 23, 271–284. Bozzali, M., Falini, A., Franceschi, M., Cercignani, M., Zuffi, M., Scotti, G., Comi, G., Filippi, M., 2002. White matter damage in Alzheimer's disease assessed in vivo using diffusion tensor magnetic resonance imaging. Journal of Neurology, Neurosurgery and Psychiatry 72, 742–746. Bozzao, A., Floris, R., Baviera, M.E., Apruzzese, A., Simonetti, G., 2001. Diffusion and perfusion MR imaging in cases of Alzheimer's disease: correlations with cortical atrophy and lesion load. American Journal of Neuroradiology 22, 1030–1036.
Chen, T.-F., Lin, C.-C., Chen, Y.-F., Liu, H.-M., Hua, M.-S., Huang, Y.-C., Chiu, M.-J., 2008. Diffusion tensor changes in patients with amnesic mild cognitive impairment and various dementias. Psychiatry Research: Neuroimaging 173, 15–21. Cummings, J.L., Mega, M., Gray, K., Rosenberg-Thompson, S., Carusi, D.A., Gornbein, J., 1994. The neuropsychiatric inventory: comprehensive assessment of psychopathology in dementia. Neurology 44, 2308–2314. Damoiseaux, J.S., Smith, S.M., Witter, M.P., Sanz-Arigita, E.J., Barkhof, F., Scheltens, P., Stam, C.J., Zarei, M., Rombouts, S.A.R.B., 2009. White matter tract integrity in aging and Alzheimer's disease. Human Brain Mapping 30, 1051–1059. Della Nave, R., Foresti, S., Pratesi, A., Ginestroni, A., Inzitari, M., Salvadori, E., Giannelli, M., Diciotti, S., Inzitari, D., Mascalchi, M., 2007. Whole-brain histogram and voxelbased analyses of diffusion tensor imaging in patients with leukoaraiosis: correlation with motor and cognitive impairment. American Journal of Neuroradiology 28, 1313–1319. Di Paola, M., Spalletta, G., Caltagirone, C., 2010. In vivo structural neuroanatomy of corpus callosum in Alzheimer's disease and mild cognitive impairment using different MRI techniques: a review. Journal of Alzheimer's disease 20, 67–95. Duvernoy, H.M., 1998. The Human Hippocampus: Functional Anatomy, Vascularization and Serial Sections with MRI. Springer Verlag, Berlin. Fahn, S., Elton, R., Members of the UPDRS development commitee, 1987. Unified Parkinson's disease rating scale. In: Fahn, S., Marsden, C.D., Calne, D.B., Goldstein, M. (Eds.), Recent Developments in Parkinson's Disease. MacMillan Healthcare Information, Florham Park, NJ. Firbank, M.J., Colloby, S.J., Burn, D.J., McKeith, I.G., O'Brien, J.T., 2003. Regional cerebral blood flow in Parkinson's disease with and without dementia. Neuroimage 20, 1309–1319. Firbank, M.J., Blamire, A.M., Krishnan, M.S., Teodorczuk, A., English, P.T., Gholkar, A., Harrison, R.M., O'Brien, J.T., 2007. Diffusion tensor imaging in dementia with Lewy bodies and Alzheimer's disease. Psychiatry Research: Neuroimaging 155, 135–145. Firbank, M.J., Blamire, A.M., Teodorczuk, A., Teper, E., Burton, E.J., Mitra, D., O'Brien, J.T., 2010. High resolution imaging of the medial temporal lobe in Alzheimer's disease and dementia with Lewy bodies. Journal of Alzheimer's disease 21, 1129–1140. Foki, T., Pirker, W., Klinger, N., Geißler, A., Rath, J., Steinkellner, T., Hoellinger, I., Gruber, S., Haubenberger, D., Lehrner, J., Pusswald, G., Trattnig, S., Auff, E., Beisteiner, R., 2010. fMRI correlates of apraxia in Parkinson's disease patients off medication. Experimental Neurology 225, 416–422. Head, D., Buckner, R.L., Shimony, J.S., Williams, L.E., Akbudak, E., Conturo, T.E., McAvoy, M., Morris, J.C., Snyder, A.Z., 2004. Differential vulnerability of anterior white matter in nondemented aging with minimal acceleration in dementia of the Alzheimer type: evidence from diffusion tensor imaging. Cerebral Cortex 14, 410–423. Herholz, K., Salmon, E., Perani, D., Baron, J.C., Holhoff, V., Frölich, L., Schönknecht, P., Ito, K., Mielke, R., Kalbe, E., Zündorf, G., Delbeuck, X., Pelati, O., Anchisi, D., Fazio, F., Kerrouche, N., Desgranges, B., Eustache, F., Beuthien-Baumann, B., Menzel, C., Schröder, J., Kato, T., Arahata, Y., Henze, M., Heiss, W.-D., 2002. Discrimination between Alzheimer dementia and controls by automated analysis of multicentre FDG PET. Neuroimage 17, 302–316. Jenkinson, M., Smith, S.M., 2001. A global optimisation method for robust affine registration of brain images. Medical Image Analysis 5, 143–156. Kalus, P., Slotboom, J., Gallinat, J., Mahlberg, R., Cattapan-Ludewig, K., Wiest, R., Nyffeler, T., Buri, C., Federspiel, A., Kunz, D., Schroth, G., Kiefer, C., 2006. Examining the gateway to the limbic system with diffusion tensor imaging: the perforant pathway in dementia. Neuroimage 30, 713–720. Kantarci, K., Jack, C.R., Xu, Y., Campeau, N.G., O'Brien, P.C., Smith, G.E., Ivnik, R.J., Boeve, B.F., Kokmen, E., Tangalos, E.G., Petersen, R.C., 2001. Mild cognitive impairment and Alzheimer disease: regional diffusivity of water. Radiology 219, 101–107. Kantarci, K., Avula, R., Senjem, M.L., Samikoglu, A.R., Zhang, B., Weigand, S.D., Przybelski, S.A., Edmonson, H.A., Vemuri, P., Knopman, D.S., Ferman, T.J., Boeve, B.F., Petersen, R.C., Jack, C.R., 2010. Dementia with Lewy bodies and Alzheimer disease: neurodegenerative patterns characterized by DTI. Neurology 74, 1814–1821. Kiuchi, K., Morikawa, M., Taoka, T., Nagashima, T., Yamauchi, T., Makinodan, M., Norimoto, K., Hashimoto, K., Kosaka, J., Inoue, Y., Inoue, M., Kichikawa, K., Kishimoto, T., 2009. Abnormalities of the uncinate fasiculus and posterior cingulate fasciculus in mild cognitive impairment and early Alzheimer's disease: a diffusion tensor tractography study. Brain Research 1287, 184–191. Kiuchi, K., Morikawa, M., Taoka, T., Kitamura, S., Nagashima, T., Makinodan, M., Nakagawa, K., Fukusumi, M., Ikeshita, K., Inoue, M., Kichikawa, K., Kishimoto, T., 2011. White matter changes in dementia with Lewy bodies and Alzheimer's disease: a tractography-based study. Journal of Psychiatric Research 45 (8), 1095–1100. Lace, G., Savva, G.M., Forster, G., de Silva, R., Brayne, C., Matthews, F.E., Barclay, J.J., Dakin, L., Ince, P.G., Wharton, S.B., 2009. Hippocampal tau pathology is related to neuroanatomical connections: an ageing population based study. Brain 132, 1324–1334. Lee, J.E., Park, H.-J., Park, B., Song, S.K., Sohn, Y.H., Lee, J.D., Lee, P.H., 2010. A comparative analysis of cognitive profiles and white matter alterations using voxel based diffusion tensor imaging between patients with Parkinson's disease dementia and dementia with Lewy bodies. Journal of Neurology, Neurosurgery & Psychiatry 81, 320–326. McKeith, I.G., Ballard, C.G., Perry, R.H., Ince, P.G., O'Brien, J.T., Neill, D., Lowery, K., Jaros, E., Barber, R., Thompson, P., Swann, A., Fairbairn, A.F., Perry, E.K., 2000. Prospective validation of consensus criteria for the diagnosis of dementia with Lewy bodies. Neurology 54, 1050–1058. McKeith, I., Mintzer, J., Aarsland, D., Burn, D., Chiu, H., Cohen-Mansfield, J., Dickson, D., Dubois, B., Duda, J.E., Feldman, H., Gauthier, S., Halliday, G., Lawlor, B., Lippa, C., Lopez, O.L., Carlos Machado, J., O'Brien, J., Playfer, J., Reid, W., 2004. International
M.J. Firbank et al. / Psychiatry Research: Neuroimaging 194 (2011) 176–183 Psychogeriatric Association Expert Meeting on DLB. 2004. Dementia with Lewy bodies. Lancet Neurology 3, 19–28. McKeith, I.G., Dickson, D.W., Lowe, J., Emre, M., O'Brien, J.T., Feldman, H., Cummings, J., Duda, J.E., Lippa, C., Perry, E.K., Aarsland, D., Arai, H., Ballard, C.G., Boeve, B., Burn, D.J., Costa, D., Del Ser, T., Dubois, B., Galasko, D., Gauthier, S., Goetz, C.G., Gomez-Tortosa, E., Halliday, G., Hansen, L.A., Hardy, J., Iwatsubo, T., Kalaria, R.N., Kaufer, D., Kenny, R.A., Korczyn, A., Kosaka, K., Lee, V.M.-Y., Lees, A., Litvan, I., Londos, E., Lopez, O.L., Minoshima, S., Mizuno, Y., Molina, J.A., Mukaetova-Ladinska, E.B., Pasquier, F., Perry, R.H., Schulz, J.B., Trojanowski, J.Q., Yamada, M., 2005. Diagnosis and management of dementia with Lewy bodies: third report of the DLB consortium. Neurology 65, 1863–1872. McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D., Stadlan, E.M., 1984. Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer's disease. Neurology 34, 939–944. Mielke, M.M., Kozauer, N.A., Chan, K.C.G., George, M., Toroney, J., Zerrate, M., BandeenRoche, K., Wang, M.C., vanZijl, P., Pekar, J.J., Mori, S., Lyketsos, C.G., Albert, M., 2009. Regionally-specific diffusion tensor imaging in mild cognitive impairment and Alzheimer's disease. Neuroimage 46, 47–55. Mori, S., Wakana, S., Nagae-Poetscher, L.M., van Zijl, P.C.M., 2005. MRI Atlas of Human Anatomy, 1 ed. Elsevier, Amsterdam. Mueller, S.G., Stables, L., Du, A.T., Schuff, N., Truran, D., Cashdollar, N., Weiner, M.W., 2007. Measurement of hippocampal subfields and age-related changes with high resolution MRI at 4 T. Neurobiology of Aging 28, 719–726. Naggara, O., Oppenheim, C., Rieu, D., Raoux, N., Rodrigo, S., Dalla, Barba G., Meder, J.F., 2006. Diffusion tensor imaging in early Alzheimer's disease. Psychiatry Research: Neuroimaging section 146, 243–249. O'Brien, J.L., O'Keefe, K.M., La Violette, P.S., DeLuca, A.N., Blacker, D., Dickerson, B.C., Sperling, R.A., 2010. Longitudinal fMRI in elderly reveals loss of hippocampal activation with clinical decline. Neurology 74, 1969–1976. Ota, M., Sato, N., Ogawa, M., Murata, M., Kuno, S., Kida, J., Asada, T., 2009. Degeneration of dementia with Lewy bodies measured by diffusion tensor imaging. NMR in Biomedicine 22, 280–284. Rey, A., 1964. L'examen Clinique en Psychologie. Presses Universitaires de France, Paris. Ringman, J.M., O'Neill, J., Geschwind, D., Medina, L., Apostolova, L.G., Rodriguez, Y., Schaffer, B., Varpetian, A., Tseng, B., Ortiz, F., Fitten, J., Cummings, J.L., Bartzokis, G., 2007. Diffusion tensor imaging in preclinical and presymptomatic carriers of familial Alzheimer's disease mutations. Brain 130, 1767–1776. Roth, M., Tym, E., Mountjoy, C.Q., Huppert, F.A., Hendrie, H., Verma, S., Goddard, R., 1986. CAMDEX — a standardised instrument for the diagnosis of mental disorder in the elderly with special reference to the early detection of dementia. British Journal of Psychiatry 149, 698–709. Sachdev, P.S., Wen, W., Christensen, H., Jorm, A.F., 2005. White matter hyperintensities are related to physical disability and poor motor function. Journal of Neurology, Neurosurgery & Psychiatry 76, 362–367.
183
Shen, Y., Larkman, D.J., Counsell, S., Pu, I.M., Edwards, D., Hajnal, J.V., 2004. Correction of high-order eddy current induced geometric distortion in diffusion-weighted echoplanar images. Magnetic Resonance in Medicine 52, 1184–1189. Stahl, R., Dietrich, O., Teipel, S.J., Hampel, H., Reiser, M.F., Schoenberg, S.O., 2007. White matter damage in Alzheimer disease and mild cognitive impairment: assessment with diffusion-tensor MR imaging and parallel imaging techniques. Radiology 243, 483–492. Stricker, N.H., Schweinsburg, B.C., Delano-Wood, L., Wierenga, C.E., Bangen, K.J., Haaland, K.Y., Frank, L.R., Salmon, D.P., Bondi, M.W., 2009. Decreased white matter integrity in late-myelinating fiber pathways in Alzheimer's disease supports retrogenesis. Neuroimage 45, 10–16. Tam, C.W., Burton, E.J., McKeith, I.G., Burn, D.J., O'Brien, J.T., 2005. Temporal lobe atrophy on MRI in Parkinson disease with dementia: a comparison with Alzheimer disease and dementia with Lewy bodies. Neurology 64, 861–865. Thomas, B.P., Welch, E.B., Blake, D., Niederhauser, B.D., Whetsell, W.O., Anderson, A.W., Gore, J.C., Avison, M.J., Creasy, J.L., 2008. High resolution 7 T MRI of the human hippocampus in vivo. Journal of Magnetic Resonance Imaging 28, 1266–1272. Villain, N., Desgranges, B., Viader, F., de la Sayette, V., Mézenge, F., Landeau, B., Baron, J.-C., Eustache, F., Chételat, G., 2008. Relationship between hippocampal atrophy, white matter disruption, and gray matter hypometabolism in Alzheimer's disease. The Journal of Neuroscience 28, 6174–6181. Walker, M.P., Ayre, G.A., Cummings, J.L., Wesnes, K., McKeith, I.G., O'Brien, J.T., Ballard, C.G., 2000. The clinician assessment of fluctuation and the one day fluctuation assessment scale. British Journal of Psychiatry 177, 252–256. Whitwell, J.L., Weigand, S.D., Shiung, M.M., Boeve, B.F., Ferman, T.J., Smith, G.E., Knopman, D.S., Petersen, R.C., Benarroch, E.E., Josephs, K.A., Jack, C.R., 2007. Focal atrophy in dementia with Lewy bodies on MRI: a distinct pattern from Alzheimer's disease. Brain 130, 708–719. Wieshmann, U.C., Symms, M.R., Mottershead, J.P., MacManus, D.G., Barker, G.J., Tofts, P.S., Revesz, T., Stevens, J.M., Shorvon, S.D., 1999. Hippocampal layers on high resolution magnetic resonance images: real or imaginary? Journal of Anatomy 195, 131–135. Yushkevich, P.A., Piven, J., Hazlett, H.C., Smith, R.G., Ho, S., Gee, J.C., Gerig, G., 2006. Userguided 3D active contour segmentation of anatomical structures: significantly improved efficiency and reliability. Neuroimage 31, 1116–1128. Zhang, H.Y., Wang, S.J., Liu, B., Ma, Z.L., Yang, M., Zhang, Z.J., Teng, G.J., 2010. Resting brain connectivivty: changes during the progress of Alzheimer disease. Radiology 256, 598–606. Zhou, Y., Dougherty, J.H., Hubner, K.F., Bai, B., Cannon, R.L., Hutson, R.K., 2008. Abnormal connectivity in the posterior cingulate and hippocampus in early Alzheimer's disease and mild cognitive impairment. Alzheimer's and Dementia 4, 265–270.
Contents lists available at ScienceDirect
Psychiatry Research: Neuroimaging j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / p s yc h r e s n s
Diffusion tensor imaging in Alzheimer's disease and dementia with Lewy bodies Michael J. Firbank a,⁎, Andrew M. Blamire b, Andrew Teodorczuk a, Emma Teper a, Dipayan Mitra c, John T. O'Brien a a b c
Institute for Ageing and Health, Newcastle University, Newcastle upon Tyne, NE4 5PL, UK Newcastle Magnetic Resonance Centre, Newcastle University, Newcastle upon Tyne, NE4 5PL, UK Department of Neuroradiology, Newcastle General Hospital, Newcastle upon Tyne, NE4 6BE, UK
a r t i c l e
i n f o
Article history: Received 10 May 2011 Received in revised form 29 July 2011 Accepted 4 August 2011 Keywords: MRI DTI DLB Connectivity Hippocampus
a b s t r a c t White matter changes have been investigated in Alzheimer's disease (AD) in a number of studies using diffusion imaging. Fewer studies have investigated dementia with Lewy bodies (DLB). We used diffusionweighted magnetic resonance imaging (MRI) and high-resolution (0.3 mm in-plane) coronal 3T MRI of the medial temporal lobe in 16 subjects with AD, 16 with DLB and 16 similarly aged healthy subjects. We found increased mean diffusivity in the temporal lobe of AD, and reduced fractional anisotropy (FA) in a small cluster in the right postcentral gyrus region in the DLB group. Mean FA in this cluster correlated with UPDRS (Unified Parkinson's Disease Rating Scale) motor score. We had previously reported reduced visibility in the AD group of a dark appearing layer of the hippocampus in the high-resolution images. In an SPM analysis on all subjects, there were significant clusters of reduced FA in the corpus callosum, fornix and stria terminalis that correlated with the visual rating of the hippocampus. These results suggest that changes to the hippocampus are associated with structural changes to the white matter fibres of the hippocampus output, and that changes in motor function are associated with changes in white matter underlying somatosensory cortex. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Dementia with Lewy bodies (DLB) is the second most common form of neurodegenerative dementia following Alzheimer's disease (AD), accounting for approximately 15% of cases at autopsy (McKeith et al., 2004). DLB shares clinical and pathological features with AD, making it potentially difficult to distinguish in clinical practice. Early and accurate diagnosis of DLB is important for optimum management, including provision of appropriate information to patients and carers, initiation of effective treatments and avoidance of potentially lifethreatening antipsychotic drugs. Diffusion imaging with magnetic resonance has been widely used to investigate the integrity of the white matter microstructure. Mean diffusivity (MD) of water is typically higher where there are fewer barriers to diffusion such as cell walls. Fractional anisotropy (FA) indicates the degree of angular variation in the magnitude of water motion (diffusion), and is highest in directionally coherent fibre bundles such as those found in corpus callosum. Change in MD and FA of the frontal, temporal and parietal lobes has been observed in AD (Kantarci et al., 2001; Bozzali et al., 2002; Head et al., 2004; Naggara et al., 2006; Firbank et al., 2007; Kiuchi et al., 2009) though some ⁎ Corresponding author at: Institute for Ageing and Health, Newcastle University, Wolfson Research Centre, Campus for Ageing and Vitality, Newcastle upon Tyne NE4 5PL, UK. Tel.: + 44 191 248 1319; fax: + 44 191 248 1301. E-mail address: m.j.fi[email protected] (M.J. Firbank). 0925-4927/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.pscychresns.2011.08.002
studies have found few differences between AD and comparable healthy subjects (Bozzao et al., 2001). Studies vary as to which regions are found to be altered, with, for instance, findings of only limited changes in FA of temporal lobe (Damoiseaux et al., 2009), change in temporal lobe MD (Stahl et al., 2007), and changes in frontal lobe and corpus callosum of MD and FA (Chen et al., 2008). However, diffusion changes in the temporal lobe have been consistently reported, with suggestions that the connectivity of the hippocampus is reduced in AD (Kalus et al., 2006; Ringman et al., 2007; Zhou et al., 2008). A recent study (Kiuchi et al., 2011), using tractography, found decreased FA in the uncinate fasciculus of both AD and DLB, with DLB additionally showing more posterior change. The few other studies in DLB also suggest more posterior changes (Firbank et al., 2007; Ota et al., 2009; Kantarci et al., 2010), with the temporal lobe relatively unaffected, in keeping with structural preservation in DLB compared to AD (Whitwell et al., 2007). We hypothesised that in this study we would find decreases in FA and increases in MD principally in the temporal lobe of AD, and more parietal changes in DLB. In an analysis of high resolution coronal magnetic resonance images (MRI) of the hippocampus, we found that a structure probably representing the stratum moleculare, stratum lacunosum and stratum radiatum was less visible in AD subjects than in DLB (Firbank et al., 2010). Neuropathological studies have found that this layer is one of the earliest affected by Alzheimer type pathology (Lace et al., 2009) and that pathology spreads through the hippocampus in stages defined broadly by the internal connectivity — i.e. from input to
M.J. Firbank et al. / Psychiatry Research: Neuroimaging 194 (2011) 176–183
output layers. We hypothesised that these hippocampal changes would be associated with changes to the white matter (WM) connecting the hippocampus to other brain areas, in particular, the fornix, which is the main output from the hippocampus, and the cingulum bundle, which is the main input (Duvernoy, 1998; Mori et al., 2005). Although medial temporal lobe atrophy is less common in DLB, it is still present (Tam et al., 2005), and we hypothesised that any alterations in connecting WM would be present regardless of the origin of the hippocampal changes. In this study, we used diffusion-weighted imaging to investigate differences in WM integrity between AD, DLB and healthy subjects. We also investigated the relationship between the changes seen on the high-resolution hippocampus imaging and FA changes in WM. 2. Methods 2.1. Participants We recruited 16 people with Alzheimer's disease and 16 with dementia with Lewy bodies, from clinical Old Age Psychiatry, Geriatric Medicine and Neurology Services. Sixteen healthy subjects of similar age were also recruited from spouses and friends of participants with dementia, as well as from a register of subjects who have previously indicated willingness to participate in research. Subjects are the same as those in our previous article (Firbank et al., 2010). All subjects were aged over 60 and did not have contra-indications for MRI. Subjects with dementia had mild to moderate severity (MMSE N 10). All Alzheimer's disease subjects fulfilled criteria for probable AD according to NINCDS/ADRDA (McKhann et al., 1984). Cases of dementia with Lewy bodies all met criteria for probable DLB according to the consensus criteria (McKeith et al., 2005). All diagnoses were made by consensus between two experienced clinicians, a method we have previously validated against autopsy diagnosis (McKeith et al., 2000). Routine clinical workup for dementia included detailed physical, neurological and neuropsychiatric examinations, including screening blood tests and CT scan. Additional assessments performed were of cognition (Cambridge Cognitive Examination (CAMCOG)) (Roth et al., 1986), mood (Cornell Depression Scale) (Alexopoulos et al., 1988), neuropsychiatric features (Neuropsychiatric Inventory (NPI)) (Cummings et al., 1994), clinical fluctuation (Clinical Assessment of Fluctuation Scale) (Walker et al., 2000), memory (Rey Auditory Verbal Learning Test (Rey, 1964)) and motor features of parkinsonism (UPDRS (Unified Parkinson's Disease Rating Scale) subsection III, recorded for each side of the body) (Fahn et al., 1987). Nine of the DLB subjects were taking anti-parkinsonian medications. The UPDRS assessment was not timed relative to subjects taking parkinsonian medication. Duration of dementia was determined from a review of the patient's medical case notes. Exclusion criteria included severe concurrent illness (apart from dementia for patients), space-occupying lesions on imaging, history of stroke and contraindications to MRI. In addition, controls had no history of psychiatric illnesses. The study was approved by the local ethics committee, and all subjects gave signed informed consent for participation. 2.2. MRI acquisition Subjects were scanned on a 3T MRI system (Intera Achieva scanner; Philips, Eindhoven, the Netherlands). Images acquired included a T1 weighted volumetric sequence covering the whole brain (MPRAGE, Sagittal acquisition, slice thickness 1.2 mm, voxel size 1.15 × 1.15 mm; TR = 9.6 ms; TE 4.6 ms; flip angle = 8 o; SENSE factor = 2). Diffusion images were acquired with FLAIR weighting to reduce the influence of CSF. Acquisition parameters were TR 7000 ms, TE 68 ms, TI 2200 ms. SENSE factor = 2, slice thickness = 2.5 mm, field of
177
view 260 × 260 mm, acquisition matrix 120 × 93. Two b values were used: 4 acquisitions with b = 0 s/mm 2, and images with diffusion weighting in 30 directions (1 acquisition each) with b = 1000 s/mm 2, with the gradient directions uniformly spaced around a sphere. High resolution coronal imaging was also performed, based on previous work (Mueller et al., 2007) with a sequence optimised locally to provide good hippocampal contrast (turbo spin echo — turbo factor 15; 24 slices; slice thickness 2 mm, field of view 210 × 167; pixel resolution 0.41 × 0.52 mm; TR 2568 ms; TE 19 ms; centric-ordered, flip angle 90 o; 2 acquisitions — acquisition time = 2 × 2:50). After the first 13 subjects (5 Control, 7 AD, 1 DLB) one of the acquisitions was replaced by three acquisitions of a higher resolution sequence for the remaining subjects to improve visualisation of the hippocampus structure (Firbank et al., 2010) with the following parameters altered: 12 slices; pixel resolution 0.27 × 0.35 mm; TR 3852 ms; 3 acquisitions — acquisition time= 3 × 2:07. The number of acquisitions was increased to maintain signal-to-noise ratio (SNR) in the face of smaller voxels. Data were acquired using multiple acquisitions to allow correction of patient motion prior to averaging to increase SNR. This approach was found to maintain highest resolution in pilot studies compared with direct averaging by the scanner. The coronal images were positioned for each subject so that they were at an angle of 25 o to the line from genu to splenium of the corpus callosum. We have found previously that this is a reliable method of angling the slices approximately perpendicular to the main axis of the hippocampus (Firbank et al., 2010). 2.3. Volumetric image analysis The T1 weighted anatomical images were segmented into grey and white matter, and spatially normalised using the segmentation algorithm in SPM5 (http://www.fil.ion.ucl.ac.uk/spm/). The ratio of total brain volume/intracranial volume was derived from the ratio of grey+white/grey+white+CSF volume. 2.4. High resolution image analysis The analysis of the high resolution coronal images is described elsewhere (Firbank et al., 2010). Briefly, we used the FLIRT image registration tool (Jenkinson and Smith, 2001)(part of FSL http://www. fmrib.ox.ac.uk/fsl/) to register all the high resolution images from each subject together, and interpolate to 0.27 × 0.27 mm resolution. A high SNR image was then created by summing together all the registered high resolution images for that subject. Images were reviewed to check for subject motion, and one subject (DLB) was excluded due to excessive motion, and for one other subject, one of the high resolution images was unacceptable. The other subjects did not have excessive motion artefacts on the images. Images were viewed with the freely available itk-snap package (Yushkevich et al., 2006) (http://www.itksnap.org/pmwiki/pmwiki.php). We examined three coronal slices, starting on the slice on which the head of the hippocampus was no longer visible, and the two slices posterior to that. A hypointense line could be seen in the hippocampus which is likely to represent fibres in the hippocampal layers of stratum moleculare, stratum lacunosum and stratum radiatum (Wieshmann et al., 1999; Thomas et al., 2008) (see Fig. 1). We observed considerable variability in how clearly this line could be visualised between subjects, and a single rater (blind to diagnosis) assigned each hippocampus a score (1 to 5) according to how clearly the hippocampus internal structure was depicted throughout the three slices examined. On this scale, 5 = line clearly visualised throughout, 4 = most of the line clearly visualised with good contrast for most of its length, 3 = line semi clearly defined, either with some sections of good contrast and some poor, or partly blurred along all its length, 2 = line mostly not clearly defined, but recognisable, typically with
178
M.J. Firbank et al. / Psychiatry Research: Neuroimaging 194 (2011) 176–183
Fig. 1. Coronal high resolution image, with location of hypointense line indicated by an arrow. On the left two subjects with well depicted hypointense line, on the right, two subjects with very poorly depicted line.
poor contrast or blurred throughout its length, 1 = line not visualised at all. We previously reported reliability of this rating scale (Firbank et al., 2010) — intraclass correlation coefficient = 0.97 for within rater, and 0.8 for between rater.
was avoided (see Fig. 2A). We also drew a ROI in the corticospinal tract, where we did not expect to see correlation with hippocampus rating. The ROI (2× 2 voxels) was drawn just superior to the corpus callosum, (Fig. 2B) using the principal eigenvector image to select voxels with an inferior-superior direction of diffusion.
2.5. Diffusion image analysis 2.6. Statistics To correct the diffusion-weighted images for the distorting effect of eddy currents, we adapted the approach of Shen (Shen et al., 2004) and used an affine registration (in FSL's FLIRT) to register each pair of opposite diffusion-weighted images together. The eddy-corrected diffusion-weighted images were then registered with a rigid body registration to the b = 0 s/mm 2 image. Mean diffusivity (MD) and fractional anisotropy (FA) images were calculated using FSL tensor analysis of the aligned diffusion-weighted images. Further processing of the images was done with SPM5. The b = 0 s/ mm 2 FLAIR-weighted image was coregistered with the T1-weighted anatomical image, and the spatial normalisation parameters derived during the segmentation process above were applied to the FA and MD images. Images were smoothed with a Gaussian FWHM of 5 mm. An average FA image was generated and thresholded at 0.2 to generate a binary mask for use with the voxelwise statistical analysis. To compare diffusion parameters between the three subject groups, we performed a voxelwise analysis of variance (ANOVA) in SPM5 of both FA and MD. In a previous report (Firbank et al., 2010), we describe our finding that the hippocampus viewed on the high resolution coronal images had less well defined structure on the visual rating scale in the Alzheimer's disease group. We hypothesised that this might relate to loss of integrity of white matter fibres within and connecting the hippocampus. To investigate this, we performed a regression of FA images in SPM, including data from all subjects, with covariates of disease group, brain/ICV ratio and the hippocampus visual rating scale described above. To verify that correlations with hippocampus rating seen on the voxelwise analysis were not merely a result of brain atrophy, we additionally measured mean FA within a region of interest (ROI) in the fornix: on four adjacent coronal slices in the body of the fornix, a voxel in the centre of the fornix was selected. Regions were drawn individually on each subject's native space b = 0 s/mm2 image, ensuring that CSF
We used ANOVA to compare demographic factors between groups with Tukey post hoc comparisons. Chi squared test was used to compare sex. All statistical analysis was performed with Minitab 15 (Minitab Inc, Pennsylvania, USA). The following three SPM analyses were performed, including all subjects: 1) An ANOVA model of MD by disease group. 2) An ANOVA model of FA by disease group. 3) A regression model of FA with disease group, brain/ICV ratio and hippocampus rating scale. Results
Fig. 2. A) midline sagittal section showing the fornix ROI. B) coronal and axial section showing the corticospinal tract ROI.
M.J. Firbank et al. / Psychiatry Research: Neuroimaging 194 (2011) 176–183
179
Table 1 Subject demographics. Control
AD
DLB
N = 16
N = 16
N = 16
Sex (female: male) Education in years Education post 16 (yes: no) MMSE Duration dementia (months)
76.3 (8.2) [61–93] 7: 9 11.5 [9–18] 8: 8 29 [26–30] –
UPDRS CAMCOG NPI total Rey total trials 1–5 (max 75) Rey delayed recall (max 15) Fluctuation score Hypertension yes: no Diabetes yes: no Intracranial volume (ml)
2.0 [0–14] 97 (3.5) – 42 [30–61] 8 [5–14] 0 [0] (n = 4) 7: 9 1: 15 1504 (150)
77.3 (8.9) [64–94] 8: 8 10.5 [9–16] 5: 11 21.5 [16–27]a 40.4 (25) [6–72] 5.5 [1–13]a 69 (11.4)a 8.5 (11.8) 21 [5–31]a 0 [0–3]a 0 [0–9] 6: 10 2: 14 1449 (159)
81.0 (5.9) [70–88] 6: 10 9.0 [8–10]b,c 0: 16b,c 18 [15–27]c 43.7 (24) [3–96] 17.5 [9–33]b,c 67.4 (14.0)c 24.1 (11.7)b 18 [4–36]c 1.0 [0–8]b,c 7 [0–16]b,c 6: 10 0: 16 1472 (140)
Age in years
§F = 1.6; p = 0.2 † p = 0.9 ‡ H = 18; p = 0.001 † p = 0.004 ‡ H = 31; p b 0.001 t = 0.16; p = 0.9 ‡ H = 32; p b 0.001 ‡ H = 32 ; p b 0.001 t = 3.6 ; p = 0.001 ‡ H = 31 ; p b 0.001 ‡ H = 35 ; p b 0.001 ‡ H = 13 ; p = 0.002 † p = 0.7 † p = 0.3 §F = 0.56, p = 0.6
Values in the table are mean (S.D.) or median [range]. Post hoc p b 0.05 (a) AD vs. Control; (b) DLB vs. AD; (c) DLB vs. Control. § ANOVA. † Fisher's exact test. ‡ Kruskal–Wallis. MMSE=Mini Mental State Examination; CAMCOG=Cambridge Cognitive Examination; UPDRS=Unified Parkinson's Disease Rating Scale (subsection 3); NPI=Neuropsychiatric Inventory; Rey=Rey Auditory Verbal Learning Test; Fluctuation score=Clinical Assessment of Fluctuation Scale.
from the SPM analysis were thresholded at p = 0.001 uncorrected for multiple comparisons, and we report significant clusters with p b 0.05 corrected for multiple comparisons with familywise error correction. 3. Results Subject demographics are described in Table 1. There were no differences between groups in age or sex distribution. Duration of dementia and CAMCOG score did not differ between AD and DLB. The SPM group ANOVA analysis results are summarised in Table 2 and Fig. 3, which report those clusters that were significant after correcting for multiple comparisons. The comparison of mean diffusivity gave two significant clusters (Fig. 3a) where MD was higher in AD relative to control subjects, focussed in the inferior and anterior medial temporal lobe, including the uncinate fasciculus and inferior longitudinal fasciculus. In the comparison of fractional anisotropy there was one cluster in the white matter beneath the post-central gyrus (Fig. 3b) where FA was lower in DLB relative to control. There were no other significant differences found between groups. The tables also show mean diffusivity within the clusters, with the MD values being in the range of brain tissue, demonstrating that the clusters do not contain large portions of CSF. Fig. 4 and Table 3 show the results of the SPM regression of hippocampus visual rating against FA in all subjects, controlling for brain size and disease group. There were a number of significant clusters where decreased FA correlated with poor hippocampus visual
rating. Significant clusters include those in the corpus callosum, fornix and stria terminalis. Since a number of regions were close to the ventricles, we wished to determine that the correlation was not just due to poor spatial registration resulting in variable amounts of CSF within the cluster. We determined the mean FA within each cluster from the original (unsmoothed) FA images averaging just those pixels with FA N0.25. The correlations were significant (p b =0.001) in each cluster, with the exception of the cerebellum where the significance was only p = 0.02. In addition, we repeated the spatial normalisation by calculating the mean of all the normalised FA images in the study, and (using the SPM normalisation function) renormalised the original FA maps to this average FA template. This should minimise spatial normalisation errors by using a cohort-specific template. The significant correlations with hippocampus clearness were still present in essentially the same locations in this analysis. Because the fornix is a very small structure, potentially difficult to accurately spatially normalise, we also performed a region of interest analysis, and manually positioned an ROI in the fornix, avoiding CSF (see Fig. 2a), and determined the mean FA. The mean FA in this ROI correlated (r = 0.58; p = 0.001) with hippocampus visual rating (see Fig. 5). As expected, there was no significant correlation between the FA in the corticospinal tract ROI and hippocampus rating (r = − 0.15; p = 0.3). Repeating this analysis with a covariate to control for the two high resolution sequences used did not alter the significance of the results.
Table 2 SPM ANOVA comparison of MD and FA between groups. Significant clusters (p b 0.05) corrected for multiple comparisons with familywise error rate (FWE) correction. x,y,z (MNI) AD MD N Control − 22,−6,−32 − 52,−36,−24 − 36,−44,−20 34,−4,−34 DLB FA b Control 28,−28,48
MD within cluster (×10− 3 mm2/s) Mean (S.D.)
Cluster p value (FWE)
Cluster size (mm3)
T statistic (at peak voxel)
b 0.001 Same cluster as above Same cluster as above b 0.001
2216
L temporal lobe
0.82 (0.03)
920
4.82 4.71 4.63 4.52
R temporal lobe
0.82 (0.04)
0.017
704
4.95
White matter deep to right Brodmann area 3
0.76 (0.04)
MNI = Montreal Neurological Institute brain template.
180
M.J. Firbank et al. / Psychiatry Research: Neuroimaging 194 (2011) 176–183
Fig. 3. Voxelwise comparisons of SPM results thresholded at p b 0.001 (uncorrected for multiple comparisons) with a minimum cluster size of 50 voxels. Significant clusters overlaid on an average brain. a) MD AD N Control – axial slices at 4-mm intervals starting at − 38 mm in MNI coordinates. b) FA DLB b Control (slices at MNI coordinates − 26, 28, 48 mm).
To investigate whether the lower FA in the right side post-central gyrus region of DLB was related to parkinsonian motor features in the patients, we performed a post hoc correlation between the mean FA in this cluster and the UPDRS motor score using the contralateral (i.e. left side) assessment. In both the DLB group alone (r = − 0.51; p = 0.04) and the combined AD and DLB group (r = −0.57; p =0.001), there was a correlation between decreased FA and increased UPDRS motor score (see Fig. 6). 4. Discussion The main finding of this study was that visual rating of hippocampus on high resolution MRI in subjects with AD and DLB correlated with fractional anisotropy in the fornix, stria terminalis and corpus callosum. The fornix is the main output from the hippocampus, and the stria terminalis from the amygdala. This suggests that atrophy of the medial temporal lobe is associated with degeneration of the outgoing fibres. Other groups have found decreased FA in the fornix in AD (Mielke et al., 2009; Kantarci et al., 2010), and one report (Stricker et al., 2009) concluded that late myelinating fibres, including the fornix and splenium of the corpus callosum, had lower FA in AD than controls,
Fig. 4. a. SPM results — regions where FA correlates with decreased hippocampus visual rating, controlling for disease group and brain/intracranial volume ratio. SPM thresholded at p b 0.001 uncorrected for multiple comparisons. Number on each slice indicates MNI x coordinate. b. SPM FA-hippocampus correlations overlaid onto coronal mean FA template (MNI y = − 64 mm and at 5-mm intervals).
and suggested retrogenesis. Reduction of fornix FA has also been observed in presymptomatic familial AD (Ringman et al., 2007). Villain (Villain et al., 2008) found decreased white matter density in AD in the corpus callosum, fornix and cingulum, and also correlations between grey matter density in the hippocampus region with white matter density in the cingulum and splenium of corpus callosum. This lower value of FA in the outgoing fibres could be a result of Wallerian degeneration, with primary degeneration of hippocampal neurones due to AD pathology leading to loss of output fibres, and hence downstream changes. fMRI studies of older subjects have
M.J. Firbank et al. / Psychiatry Research: Neuroimaging 194 (2011) 176–183
181
Table 3 SPM correlation of hippocampus visual rating against FA, controlling for brain size and dementia group. Significant clusters (p b 0.05) corrected for multiple comparisons with familywise error rate (FWE) correction. x,y,z (MNI)
Cluster p value (FWE)
Cluster size (mm3)
T statistic (at peak voxel)
− 18,−26,28 − 6 − 14,28 10,−18,30 26,−36,24 0,−6,10 32,44,−8 − 24,−38,2 − 18,−30,12 14,−66,−46
b 0.001 Same cluster as above b 0.001 Same cluster as above 0.040 0.020 0.017 Same cluster as above 0.040
2896
5.20 5.03 4.46 4.07 5.08 5.04 4.30 3.90 4.19
2048 576 672 688 576
MD within cluster (×10− 3 mm2/s) mean (S.D.) L periventricular WM L corpus callosum body R corpus callosum body R periventricular WM fornix R anterior frontal lobe Fornix/stria terminalis Stria terminalis R superior cerebellar peduncle
0.66 " 0.66 " 0.98 0.78 1.03 " 0.65
(0.17) (0.17) (0.18) (0.04) (0.17) (0.07)
MNI = Montreal Neurological Institute brain template.
shown that functional activity of the hippocampus decreases as memory problems get worse (O'Brien et al., 2010). The output pathway along the fornix passes via the anterior thalamic nucleus to the posterior cingulate (Duvernoy, 1998), a region that has marked reduction in cerebral blood flow and metabolism early on in both AD and DLB (Herholz et al., 2002; Firbank et al., 2003). Functional connectivity between the posterior cingulate and the hippocampus is lower in AD (Zhang et al., 2010) and this could be partly mediated through changes in the connecting white matter tracts. The corpus callosum (CC) is not directly connected to the hippocampus, though reductions in posterior CC volume and diffusivity (Di Paola et al., 2010) have been reported early in AD. The relationship with hippocampus volume may be due to shared common neuropathological causes. The periventricular WM cluster (Table 3) is in the vicinity of the tapetum, a fibre bundle which passes from the temporal lobe through the corpus callosum. The anterior frontal lobe cluster is in the region of the uncinate fasciculus, which connects the temporal lobe (including the hippocampus formation) to the frontal lobe. In both these cases, it may be that hippocampal degeneration is associated with wider damage to the temporal lobe, and hence connecting white matter fibre bundles. A number of the regions in which FA correlated with hippocampus rating were adjacent to CSF. We have endeavoured to demonstrate
that the correlations were not due to the presence of CSF. However, there remains the possibility that some of the significant clusters in the SPM analysis are regions where there is an increased proportion of CSF due to inaccurate spatial registration resulting from generalised brain atrophy. Our finding of more differences in the temporal lobe in AD compared to DLB is in keeping with previous diffusion studies (Kantarci et al., 2010) and also with studies of brain atrophy (Whitwell et al., 2007) which show marked temporal lobe changes in AD. Interestingly, the temporal lobe differences in AD were not in the same location as the correlates of hippocampus degeneration, suggesting perhaps that damage to input and output fibres of the hippocampus occurs by differing processes. We found a small cluster in the white matter underlying the postcentral gyrus where FA was lower in the DLB group relative to the control group. The mean FA in this cluster was significantly correlated with UPDRS score in both the DLB and combined dementia groups, suggesting a relationship between white matter change and motor function. Supporting this finding, a study by Lee et al. (2010) found bilateral clusters of lower FA in their Parkinson disease dementia (PDD) group (though not DLB) extending up to approximately 21, −22,40 and −22, −29, 42 in MNI coordinates close to the cluster that we observed. Other studies in DLB have used region of interest approaches, and hence this region has not been otherwise investigated. An fMRI study of fine motor function (Foki et al., 2010) found relatively decreased activation in post-central gyrus in PD, and
Fig. 5. Mean FA in manually drawn ROI in body of fornix vs. hippocampus visual rating. Correlation coefficient r = 0.58; p = 0.001.
Fig. 6. Mean FA in the right postcentral gyrus white matter cluster (MNI coordinate 28, −28,48) against left side UPDRS motor score. Correlation coefficient r = −0.57; p =0.001.
182
M.J. Firbank et al. / Psychiatry Research: Neuroimaging 194 (2011) 176–183
integration of sensory with motor function is impaired in PD (Abbruzzese and Berardelli, 2003). There is also the possibility that this finding might be related to vascular disease, which is associated with both changes to the white matter evident in diffusion imaging, and with motor signs (Sachdev et al., 2005; Della Nave et al., 2007). Our previous study in DLB found reduced FA in the white matter of the precuneus (Firbank et al., 2007). This finding receives some support from the study here — there was a bilateral pair of clusters of voxels which passed the 0.001 significance threshold (but which were not significant on the corrected cluster threshold) with reduced FA in DLB compared to control (MNI 16, −60, 54; 31 voxels, and − 16 −58, 44; 14 voxels) close to the precuneus region where we previously reported reduced FA in DLB. Limitations of the study are that we changed the hippocampus structural imaging sequence part way through the study. However, we do not feel that this will have influenced the results to a great extent — including the sequence type in the analysis did not alter the significance of the findings either in the analysis here, or in our previous report which focussed on the hippocampus imaging. We also relied upon clinical diagnosis, without neuropathological confirmation, although we have previously demonstrated good correlation between clinical and pathological diagnosis (McKeith et al., 2000). The associations between hippocampus visual rating and FA were much stronger than any differences in FA between AD or DLB and control subjects, suggesting that atrophy of the hippocampus is consistently associated with regional brain changes, but that these associated FA changes may not directly translate into functional deficits clinically. Studies have shown that the difference in FA between young and old subjects is of an equal or greater magnitude to that between subjects with dementia and old matched controls. (Head et al., 2004; Damoiseaux et al., 2009) despite there being less cognitive difference between young and old subjects than AD. It may be that DTI is sensitive to factors such as differences in brain water content which do not necessarily directly impact upon cognition. Other factors, i.e. brain pathology in other areas, neurochemical changes, etc., may therefore be important in addition. In conclusion, we found evidence of temporal lobe changes in Alzheimer's disease, with degeneration of the hippocampus associated with breakdown of the white matter in its outgoing WM fibres. We also found lower fractional anisotropy of the white matter in the post-central gyrus area of DLB, which correlated with worse motor function. Longitudinal studies starting with subjects at risk of dementia, or with very early stage disease, are necessary to determine the relative time course of these changes and their relationship to functional changes.
Acknowledgments We are grateful for funding from the Alzheimer's Research Trust. This work was supported by the UK NIHR Biomedical Research Centre for Ageing and Age-related Disease award to the Newcastle upon Tyne Hospitals NHS Foundation Trust. We also thank the North East DeNDRoN (Dementia and Neurodegenerative Diseases Research Network) team for help with subject recruitment.
References Abbruzzese, G., Berardelli, A., 2003. Sensorimotor integration in movement disorders. Movement Disorders 18, 231–240. Alexopoulos, G.S., Abrams, R.C., Young, R.C., Shamoian, C.A., 1988. Cornell scale for depression in dementia. Biological Psychiatry 23, 271–284. Bozzali, M., Falini, A., Franceschi, M., Cercignani, M., Zuffi, M., Scotti, G., Comi, G., Filippi, M., 2002. White matter damage in Alzheimer's disease assessed in vivo using diffusion tensor magnetic resonance imaging. Journal of Neurology, Neurosurgery and Psychiatry 72, 742–746. Bozzao, A., Floris, R., Baviera, M.E., Apruzzese, A., Simonetti, G., 2001. Diffusion and perfusion MR imaging in cases of Alzheimer's disease: correlations with cortical atrophy and lesion load. American Journal of Neuroradiology 22, 1030–1036.
Chen, T.-F., Lin, C.-C., Chen, Y.-F., Liu, H.-M., Hua, M.-S., Huang, Y.-C., Chiu, M.-J., 2008. Diffusion tensor changes in patients with amnesic mild cognitive impairment and various dementias. Psychiatry Research: Neuroimaging 173, 15–21. Cummings, J.L., Mega, M., Gray, K., Rosenberg-Thompson, S., Carusi, D.A., Gornbein, J., 1994. The neuropsychiatric inventory: comprehensive assessment of psychopathology in dementia. Neurology 44, 2308–2314. Damoiseaux, J.S., Smith, S.M., Witter, M.P., Sanz-Arigita, E.J., Barkhof, F., Scheltens, P., Stam, C.J., Zarei, M., Rombouts, S.A.R.B., 2009. White matter tract integrity in aging and Alzheimer's disease. Human Brain Mapping 30, 1051–1059. Della Nave, R., Foresti, S., Pratesi, A., Ginestroni, A., Inzitari, M., Salvadori, E., Giannelli, M., Diciotti, S., Inzitari, D., Mascalchi, M., 2007. Whole-brain histogram and voxelbased analyses of diffusion tensor imaging in patients with leukoaraiosis: correlation with motor and cognitive impairment. American Journal of Neuroradiology 28, 1313–1319. Di Paola, M., Spalletta, G., Caltagirone, C., 2010. In vivo structural neuroanatomy of corpus callosum in Alzheimer's disease and mild cognitive impairment using different MRI techniques: a review. Journal of Alzheimer's disease 20, 67–95. Duvernoy, H.M., 1998. The Human Hippocampus: Functional Anatomy, Vascularization and Serial Sections with MRI. Springer Verlag, Berlin. Fahn, S., Elton, R., Members of the UPDRS development commitee, 1987. Unified Parkinson's disease rating scale. In: Fahn, S., Marsden, C.D., Calne, D.B., Goldstein, M. (Eds.), Recent Developments in Parkinson's Disease. MacMillan Healthcare Information, Florham Park, NJ. Firbank, M.J., Colloby, S.J., Burn, D.J., McKeith, I.G., O'Brien, J.T., 2003. Regional cerebral blood flow in Parkinson's disease with and without dementia. Neuroimage 20, 1309–1319. Firbank, M.J., Blamire, A.M., Krishnan, M.S., Teodorczuk, A., English, P.T., Gholkar, A., Harrison, R.M., O'Brien, J.T., 2007. Diffusion tensor imaging in dementia with Lewy bodies and Alzheimer's disease. Psychiatry Research: Neuroimaging 155, 135–145. Firbank, M.J., Blamire, A.M., Teodorczuk, A., Teper, E., Burton, E.J., Mitra, D., O'Brien, J.T., 2010. High resolution imaging of the medial temporal lobe in Alzheimer's disease and dementia with Lewy bodies. Journal of Alzheimer's disease 21, 1129–1140. Foki, T., Pirker, W., Klinger, N., Geißler, A., Rath, J., Steinkellner, T., Hoellinger, I., Gruber, S., Haubenberger, D., Lehrner, J., Pusswald, G., Trattnig, S., Auff, E., Beisteiner, R., 2010. fMRI correlates of apraxia in Parkinson's disease patients off medication. Experimental Neurology 225, 416–422. Head, D., Buckner, R.L., Shimony, J.S., Williams, L.E., Akbudak, E., Conturo, T.E., McAvoy, M., Morris, J.C., Snyder, A.Z., 2004. Differential vulnerability of anterior white matter in nondemented aging with minimal acceleration in dementia of the Alzheimer type: evidence from diffusion tensor imaging. Cerebral Cortex 14, 410–423. Herholz, K., Salmon, E., Perani, D., Baron, J.C., Holhoff, V., Frölich, L., Schönknecht, P., Ito, K., Mielke, R., Kalbe, E., Zündorf, G., Delbeuck, X., Pelati, O., Anchisi, D., Fazio, F., Kerrouche, N., Desgranges, B., Eustache, F., Beuthien-Baumann, B., Menzel, C., Schröder, J., Kato, T., Arahata, Y., Henze, M., Heiss, W.-D., 2002. Discrimination between Alzheimer dementia and controls by automated analysis of multicentre FDG PET. Neuroimage 17, 302–316. Jenkinson, M., Smith, S.M., 2001. A global optimisation method for robust affine registration of brain images. Medical Image Analysis 5, 143–156. Kalus, P., Slotboom, J., Gallinat, J., Mahlberg, R., Cattapan-Ludewig, K., Wiest, R., Nyffeler, T., Buri, C., Federspiel, A., Kunz, D., Schroth, G., Kiefer, C., 2006. Examining the gateway to the limbic system with diffusion tensor imaging: the perforant pathway in dementia. Neuroimage 30, 713–720. Kantarci, K., Jack, C.R., Xu, Y., Campeau, N.G., O'Brien, P.C., Smith, G.E., Ivnik, R.J., Boeve, B.F., Kokmen, E., Tangalos, E.G., Petersen, R.C., 2001. Mild cognitive impairment and Alzheimer disease: regional diffusivity of water. Radiology 219, 101–107. Kantarci, K., Avula, R., Senjem, M.L., Samikoglu, A.R., Zhang, B., Weigand, S.D., Przybelski, S.A., Edmonson, H.A., Vemuri, P., Knopman, D.S., Ferman, T.J., Boeve, B.F., Petersen, R.C., Jack, C.R., 2010. Dementia with Lewy bodies and Alzheimer disease: neurodegenerative patterns characterized by DTI. Neurology 74, 1814–1821. Kiuchi, K., Morikawa, M., Taoka, T., Nagashima, T., Yamauchi, T., Makinodan, M., Norimoto, K., Hashimoto, K., Kosaka, J., Inoue, Y., Inoue, M., Kichikawa, K., Kishimoto, T., 2009. Abnormalities of the uncinate fasiculus and posterior cingulate fasciculus in mild cognitive impairment and early Alzheimer's disease: a diffusion tensor tractography study. Brain Research 1287, 184–191. Kiuchi, K., Morikawa, M., Taoka, T., Kitamura, S., Nagashima, T., Makinodan, M., Nakagawa, K., Fukusumi, M., Ikeshita, K., Inoue, M., Kichikawa, K., Kishimoto, T., 2011. White matter changes in dementia with Lewy bodies and Alzheimer's disease: a tractography-based study. Journal of Psychiatric Research 45 (8), 1095–1100. Lace, G., Savva, G.M., Forster, G., de Silva, R., Brayne, C., Matthews, F.E., Barclay, J.J., Dakin, L., Ince, P.G., Wharton, S.B., 2009. Hippocampal tau pathology is related to neuroanatomical connections: an ageing population based study. Brain 132, 1324–1334. Lee, J.E., Park, H.-J., Park, B., Song, S.K., Sohn, Y.H., Lee, J.D., Lee, P.H., 2010. A comparative analysis of cognitive profiles and white matter alterations using voxel based diffusion tensor imaging between patients with Parkinson's disease dementia and dementia with Lewy bodies. Journal of Neurology, Neurosurgery & Psychiatry 81, 320–326. McKeith, I.G., Ballard, C.G., Perry, R.H., Ince, P.G., O'Brien, J.T., Neill, D., Lowery, K., Jaros, E., Barber, R., Thompson, P., Swann, A., Fairbairn, A.F., Perry, E.K., 2000. Prospective validation of consensus criteria for the diagnosis of dementia with Lewy bodies. Neurology 54, 1050–1058. McKeith, I., Mintzer, J., Aarsland, D., Burn, D., Chiu, H., Cohen-Mansfield, J., Dickson, D., Dubois, B., Duda, J.E., Feldman, H., Gauthier, S., Halliday, G., Lawlor, B., Lippa, C., Lopez, O.L., Carlos Machado, J., O'Brien, J., Playfer, J., Reid, W., 2004. International
M.J. Firbank et al. / Psychiatry Research: Neuroimaging 194 (2011) 176–183 Psychogeriatric Association Expert Meeting on DLB. 2004. Dementia with Lewy bodies. Lancet Neurology 3, 19–28. McKeith, I.G., Dickson, D.W., Lowe, J., Emre, M., O'Brien, J.T., Feldman, H., Cummings, J., Duda, J.E., Lippa, C., Perry, E.K., Aarsland, D., Arai, H., Ballard, C.G., Boeve, B., Burn, D.J., Costa, D., Del Ser, T., Dubois, B., Galasko, D., Gauthier, S., Goetz, C.G., Gomez-Tortosa, E., Halliday, G., Hansen, L.A., Hardy, J., Iwatsubo, T., Kalaria, R.N., Kaufer, D., Kenny, R.A., Korczyn, A., Kosaka, K., Lee, V.M.-Y., Lees, A., Litvan, I., Londos, E., Lopez, O.L., Minoshima, S., Mizuno, Y., Molina, J.A., Mukaetova-Ladinska, E.B., Pasquier, F., Perry, R.H., Schulz, J.B., Trojanowski, J.Q., Yamada, M., 2005. Diagnosis and management of dementia with Lewy bodies: third report of the DLB consortium. Neurology 65, 1863–1872. McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D., Stadlan, E.M., 1984. Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer's disease. Neurology 34, 939–944. Mielke, M.M., Kozauer, N.A., Chan, K.C.G., George, M., Toroney, J., Zerrate, M., BandeenRoche, K., Wang, M.C., vanZijl, P., Pekar, J.J., Mori, S., Lyketsos, C.G., Albert, M., 2009. Regionally-specific diffusion tensor imaging in mild cognitive impairment and Alzheimer's disease. Neuroimage 46, 47–55. Mori, S., Wakana, S., Nagae-Poetscher, L.M., van Zijl, P.C.M., 2005. MRI Atlas of Human Anatomy, 1 ed. Elsevier, Amsterdam. Mueller, S.G., Stables, L., Du, A.T., Schuff, N., Truran, D., Cashdollar, N., Weiner, M.W., 2007. Measurement of hippocampal subfields and age-related changes with high resolution MRI at 4 T. Neurobiology of Aging 28, 719–726. Naggara, O., Oppenheim, C., Rieu, D., Raoux, N., Rodrigo, S., Dalla, Barba G., Meder, J.F., 2006. Diffusion tensor imaging in early Alzheimer's disease. Psychiatry Research: Neuroimaging section 146, 243–249. O'Brien, J.L., O'Keefe, K.M., La Violette, P.S., DeLuca, A.N., Blacker, D., Dickerson, B.C., Sperling, R.A., 2010. Longitudinal fMRI in elderly reveals loss of hippocampal activation with clinical decline. Neurology 74, 1969–1976. Ota, M., Sato, N., Ogawa, M., Murata, M., Kuno, S., Kida, J., Asada, T., 2009. Degeneration of dementia with Lewy bodies measured by diffusion tensor imaging. NMR in Biomedicine 22, 280–284. Rey, A., 1964. L'examen Clinique en Psychologie. Presses Universitaires de France, Paris. Ringman, J.M., O'Neill, J., Geschwind, D., Medina, L., Apostolova, L.G., Rodriguez, Y., Schaffer, B., Varpetian, A., Tseng, B., Ortiz, F., Fitten, J., Cummings, J.L., Bartzokis, G., 2007. Diffusion tensor imaging in preclinical and presymptomatic carriers of familial Alzheimer's disease mutations. Brain 130, 1767–1776. Roth, M., Tym, E., Mountjoy, C.Q., Huppert, F.A., Hendrie, H., Verma, S., Goddard, R., 1986. CAMDEX — a standardised instrument for the diagnosis of mental disorder in the elderly with special reference to the early detection of dementia. British Journal of Psychiatry 149, 698–709. Sachdev, P.S., Wen, W., Christensen, H., Jorm, A.F., 2005. White matter hyperintensities are related to physical disability and poor motor function. Journal of Neurology, Neurosurgery & Psychiatry 76, 362–367.
183
Shen, Y., Larkman, D.J., Counsell, S., Pu, I.M., Edwards, D., Hajnal, J.V., 2004. Correction of high-order eddy current induced geometric distortion in diffusion-weighted echoplanar images. Magnetic Resonance in Medicine 52, 1184–1189. Stahl, R., Dietrich, O., Teipel, S.J., Hampel, H., Reiser, M.F., Schoenberg, S.O., 2007. White matter damage in Alzheimer disease and mild cognitive impairment: assessment with diffusion-tensor MR imaging and parallel imaging techniques. Radiology 243, 483–492. Stricker, N.H., Schweinsburg, B.C., Delano-Wood, L., Wierenga, C.E., Bangen, K.J., Haaland, K.Y., Frank, L.R., Salmon, D.P., Bondi, M.W., 2009. Decreased white matter integrity in late-myelinating fiber pathways in Alzheimer's disease supports retrogenesis. Neuroimage 45, 10–16. Tam, C.W., Burton, E.J., McKeith, I.G., Burn, D.J., O'Brien, J.T., 2005. Temporal lobe atrophy on MRI in Parkinson disease with dementia: a comparison with Alzheimer disease and dementia with Lewy bodies. Neurology 64, 861–865. Thomas, B.P., Welch, E.B., Blake, D., Niederhauser, B.D., Whetsell, W.O., Anderson, A.W., Gore, J.C., Avison, M.J., Creasy, J.L., 2008. High resolution 7 T MRI of the human hippocampus in vivo. Journal of Magnetic Resonance Imaging 28, 1266–1272. Villain, N., Desgranges, B., Viader, F., de la Sayette, V., Mézenge, F., Landeau, B., Baron, J.-C., Eustache, F., Chételat, G., 2008. Relationship between hippocampal atrophy, white matter disruption, and gray matter hypometabolism in Alzheimer's disease. The Journal of Neuroscience 28, 6174–6181. Walker, M.P., Ayre, G.A., Cummings, J.L., Wesnes, K., McKeith, I.G., O'Brien, J.T., Ballard, C.G., 2000. The clinician assessment of fluctuation and the one day fluctuation assessment scale. British Journal of Psychiatry 177, 252–256. Whitwell, J.L., Weigand, S.D., Shiung, M.M., Boeve, B.F., Ferman, T.J., Smith, G.E., Knopman, D.S., Petersen, R.C., Benarroch, E.E., Josephs, K.A., Jack, C.R., 2007. Focal atrophy in dementia with Lewy bodies on MRI: a distinct pattern from Alzheimer's disease. Brain 130, 708–719. Wieshmann, U.C., Symms, M.R., Mottershead, J.P., MacManus, D.G., Barker, G.J., Tofts, P.S., Revesz, T., Stevens, J.M., Shorvon, S.D., 1999. Hippocampal layers on high resolution magnetic resonance images: real or imaginary? Journal of Anatomy 195, 131–135. Yushkevich, P.A., Piven, J., Hazlett, H.C., Smith, R.G., Ho, S., Gee, J.C., Gerig, G., 2006. Userguided 3D active contour segmentation of anatomical structures: significantly improved efficiency and reliability. Neuroimage 31, 1116–1128. Zhang, H.Y., Wang, S.J., Liu, B., Ma, Z.L., Yang, M., Zhang, Z.J., Teng, G.J., 2010. Resting brain connectivivty: changes during the progress of Alzheimer disease. Radiology 256, 598–606. Zhou, Y., Dougherty, J.H., Hubner, K.F., Bai, B., Cannon, R.L., Hutson, R.K., 2008. Abnormal connectivity in the posterior cingulate and hippocampus in early Alzheimer's disease and mild cognitive impairment. Alzheimer's and Dementia 4, 265–270.