A JOURNAL OF NEUROLOGY
Magnetic resonance imaging evidence for
presymptomatic change in thalamus and caudate in
familial Alzheimer’s disease
Natalie S. Ryan,1Shiva Keihaninejad,1,2Timothy J. Shakespeare,1Manja Lehmann,1
Sebastian J. Crutch,1Ian B. Malone,1John S. Thornton,3,4Laura Mancini,3,4Harpreet Hyare,5
Tarek Yousry,3,4Gerard R. Ridgway,1,6Hui Zhang,2Marc Modat,1,2Daniel C. Alexander,2
Martin N. Rossor,1Sebastien Ourselin1,2and Nick C. Fox1
1 Dementia Research Centre, Department of Neurodegenerative Disease, University College London (UCL) Institute of Neurology, Queen Square,
London, WC1N 3BG, UK
2 Centre for Medical Image Computing, UCL, Malet Place, London, WC1E 6BT, UK
3 Lysholm Department of Neuroradiology, National Hospital for Neurology and Neurosurgery, UCLH NHS Foundation Trust, Queen Square, London,
WC1N 3BG, UK
4 Neuroradiological Academic Unit, UCL Institute of Neurology, Queen Square, London, WC1N 3BG, UK
5 MRC Prion Unit, Department of Neurodegenerative Disease, UCL Institute of Neurology, Queen Square, London, WC1N 3BG, UK
6 Wellcome Trust Centre for Neuroimaging, UCL Institute of Neurology, Queen Square, London, WC1N 3BG, UK
Correspondence to: Natalie S Ryan,
Dementia Research Centre,
Box 16 National Hospital for Neurology and Neurosugery,
London WC1N 3BG, UK
Amyloid imaging studies of presymptomatic familial Alzheimer’s disease have revealed the striatum and thalamus to be the
earliest sites of amyloid deposition. This study aimed to investigate whether there are associated volume and diffusivity changes
in these subcortical structures during the presymptomatic and symptomatic stages of familial Alzheimer’s disease. As the
thalamus and striatum are involved in neural networks subserving complex cognitive and behavioural functions, we also
examined the diffusion characteristics in connecting white matter tracts. A cohort of 20 presenilin 1 mutation carriers underwent
volumetric and diffusion tensor magnetic resonance imaging, neuropsychological and clinical assessments; 10 were symptom-
atic, 10 were presymptomatic and on average 5.6 years younger than their expected age at onset; 20 healthy control subjects
were also studied. We conducted region of interest analyses of volume and diffusivity changes in the thalamus, caudate,
putamen and hippocampus and examined diffusion behaviour in the white matter tracts of interest (fornix, cingulum and
corpus callosum). Voxel-based morphometry and tract-based spatial statistics were also used to provide unbiased whole-brain
analyses of group differences in volume and diffusion indices, respectively. We found that reduced volumes of the left thalamus
and bilateral caudate were evident at a presymptomatic stage, together with increased fractional anisotropy of bilateral thalamus
and left caudate. Although no significant hippocampal volume loss was evident presymptomatically, reduced mean diffusivity
was observed in the right hippocampus and reduced mean and axial diffusivity in the right cingulum. In contrast, symptomatic
mutation carriers showed increased mean, axial and in particular radial diffusivity, with reduced fractional anisotropy, in all of
the white matter tracts of interest. The symptomatic group also showed atrophy and increased mean diffusivity in all of the
subcortical grey matter regions of interest, with increased fractional anisotropy in bilateral putamen. We propose that axonal
doi:10.1093/brain/awt065 Brain 2013: Page 1 of 16 |
Received October 22, 2012. Revised January 29, 2013. Accepted January 31, 2013.
? The Author (2013). Published by Oxford University Press on behalf of the Guarantors of Brain.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0/), which
permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Brain Advance Access published March 28, 2013
at University College London on April 16, 2013
injury may be an early event in presymptomatic Alzheimer’s disease, causing an initial fall in axial and mean diffusivity, which
then increases with loss of axonal density. The selective degeneration of long-coursing white matter tracts, with relative
preservation of short interneurons, may account for the increase in fractional anisotropy that is seen in the thalamus and
caudate presymptomatically. It may be owing to their dense connectivity that imaging changes are seen first in the thalamus
and striatum, which then progress to involve other regions in a vulnerable neuronal network.
Keywords: familial Alzheimer’s disease; presenilin 1 (PSEN1, PS1); presymptomatic; diffusion tensor imaging; subcortical atrophy
Abbreviations: DTI = diffusion tensor imaging; PMC = presymptomatic mutation carrier; SMC = symptomatic mutation carrier;
TBSS = tract-based spatial statistics
and, unless more effective treatments are developed, this figure is
predicted to double every two decades (Ferri et al., 2005). A small
minority of cases are caused by autosomal dominantly inherited mu-
tations in the presenilin 1 (PSEN1), presenilin 2 (PSEN2) or amyloid
precursor protein (APP) genes, or by APP duplications. Although
rare, these mutations show virtually complete penetrance and mu-
tationcarriersthereforeprovide aunique opportunitytogaininsights
into the earliest stages of the disease. A wealth of evidence now
indicates that the symptoms of Alzheimer’s disease are preceded
by a long period of gradual accrual of pathological change (Price
and Morris, 1999; Bateman et al., 2012). The disappointing results
of trials of putative disease-modifying therapies in established
Alzheimer’s disease have shifted attention towards starting treat-
ment earlier in the disease course. Familial Alzheimer’s disease mu-
tation carriers represent a cohort in whom treatment could be
initiated at a presymptomatic stage and a number of such trials are
As the individuals participating in these trials will have little or no
cognitive impairment, there is a need to understand the trajectory
of biomarker changes in presymptomatic familial Alzheimer’s
disease, in order to both assess disease progression and look for
The cross-sectional results of the large international Dominantly
Inherited Alzheimer Network (DIAN) study have recently been
published, indicating that CSF biomarker changes are evident
over two decades before an individual’s expected age at symptom
onset, as determined by their parental age at onset (Bateman
et al., 2012). Reduced concentrations of CSF amyloid-b42 and
increased concentrations of CSF tau were detected at 25 and 15
years from expected symptom onset, respectively. Amyloid-b de-
position on Pittsburgh compound B-PET scans (Klunk et al., 2004)
and hippocampal volume loss were apparent at 15 years and cere-
bral hypometabolism at 10 years before expected symptom onset.
In the PSEN1 E280A Colombian kindred, altered hippocampal ac-
tivation on functional MRI during memory encoding has been
observed approximately two decades before expected age at
onset (Reiman et al., 2012). Presymptomatic hippocampal atrophy
has also been detected in longitudinal studies, which have fol-
lowed smaller cohorts through the presymptomatic stage to age
at symptom onset. At a group level, mutation carriers showed
hippocampal and whole brain volume loss compared with control
subjects at 3 and 1 year before symptom onset (Ridha et al.,
2006). However, longitudinal measures were able to detect
changes earlier, with differences in hippocampal and whole brain
atrophy rates observed at 5.5 and 3.5 years before symptom onset
in mutation carriers and cortical thinning at 4.1 and 1.8 years
presymptomatically for precuneus and posterior cingulate cortex,
respectively (Ridha et al., 2006; Knight et al., 2011a).
Previous volumetric imaging studies have mainly focused on the
these structures play in higher cognitive functions and memory, and
their apparent vulnerability to Alzheimer’s disease pathology.
However, Pittsburgh compound B-PET studies have indicated that
prominent amyloid deposition first appears in the striatum and thal-
amus in presymptomatic familial Alzheimer’s disease (Klunk et al.,
2007; Knight et al., 2011b). These subcortical structures have
received relatively little attention from volumetric imaging studies
in Alzheimer’s disease. They are known to be significantly atrophied
in sporadic Alzheimer’s disease but it is not known how early this
atrophy occurs (de Jong et al., 2008). A recent familial Alzheimer’s
disease study reported decreased volumes of thalamus, caudate and
putamen in presymptomatic mutation carriers who were on average
15 years younger than their family’s median age at dementia diag-
nosis, with a trend towards increasing caudate size in those with
memory deficits who were 10 years younger (Lee et al., 2013).
Increased caudate volumes were reported in another cohort of pre-
symptomatic mutation carriers who were approximately a decade
younger than their expected age at symptom onset (Fortea et al.,
2010). This study also included diffusion imaging and found asso-
ciated changes in mean diffusivity in the caudate of presymptomatic
Although the thalamus and striatum are structures that are not
conventionally associated with Alzheimer’s disease, it has long
been known that they are affected by Alzheimer’s disease path-
ology (Braak and Braak, 1990, 1991). It is also becoming increas-
ingly recognized that, given their dense connections with the
cortex, they are involved in a variety of neural networks subser-
ving complex cognitive and behavioural functions (Middleton and
Strick, 2000). The caudate, for example, appears to play a key role
in supporting the planning and execution of behaviour needed to
achieve complex goals (Grahn et al., 2008), whilst thalamic nuclei
are involved in declarative memory function at a number of dif-
ferent levels (Van der Werf et al., 2003). There is a growing ap-
preciation of the idea that Alzheimer’s disease, and indeed
neurodegenerative diseases in general, should be considered as
Brain 2013: Page 2 of 16N. S. Ryan et al.
at University College London on April 16, 2013
network disorders in which there is selective degeneration of vul-
nerable neuronal circuits (Warren et al., 2012). It therefore seems
appropriate to investigate the subcortical grey matter regions
involved in such networks, as well as the cortical areas and their
connecting white matter.
Diffusion tensor imaging (DTI) allows assessment of alterations
in tissue microstructure in specific brain regions and can be used to
examine the integrity of connecting white matter pathways
(Basser et al., 1994). DTI exploits the fact that water molecules
show preferential diffusion along the major axis of a white matter
fibre bundle (anisotropy), owing to barriers hindering perpendicu-
lar diffusion including axonal membranes and the myelin sheath.
The effects of a pathological process on white matter tracts can be
explored by investigating whether there are associated changes in
the overall magnitude of diffusion (mean diffusivity) or the degree
of its directionality (fractional anisotropy). Examining the compo-
nent eigenvalues of the diffusion tensor, from which fractional
anisotropy is derived, provides further information on how diffu-
sion is altered in both the principal direction of the tract (axial
diffusivity) and in the plane perpendicular to this (radial diffusiv-
ity). If there are changes in axial diffusivity and radial diffusivity
which are proportional, there will be no corresponding change in
fractional anisotropy, making it important to examine these com-
ponents separately (Acosta-Cabronero et al., 2010). DTI can be
used to characterize grey matter microstructure too (Bozzali et al.,
2002; McKinstry et al., 2002), with mean diffusivity a commonly
used metric (Benedetti et al., 2006). In fibre-rich structures such as
the thalamus and striatum, fractional anisotropy may also be a
useful indicator of pathological change (Ciccarelli et al., 2001;
Tovar-Moll et al., 2009; Bohanna et al., 2011). Many DTI studies
of sporadic Alzheimer’s disease have now been published, report-
ing decreased fractional anisotropy and increased mean diffusivity
in various tracts, most notably the cingulum, corpus callosum and
white matter of the parahippocampal gyrus (Chua et al., 2008;
Stebbins and Murphy, 2009). There has been just one previous
DTI study of presymptomatic familial Alzheimer’s disease, which
examined fractional anisotropy only and found it to be globally
reduced in the white matter of presymptomatic mutation carriers,
most strikingly in the fornix (Ringman et al., 2007).
We investigated a cohort of PSEN1 presymptomatic mutation car-
riers who were relatively close to their expected age at symptom
onset, together with groups of symptomatic mutation carriers and
control subjects. We used region of interest analyses to examine
volumetric and diffusivity differences in the subcortical structures
that are the main focus of this study, and altered diffusivity in
white matter tracts of interest. We also used automated whole
brain analysis techniques; voxel-based morphometry and tract-
based spatial statistics (TBSS), to provide whole-brain assessments
of group differences in volume and DTI indices, respectively.
Materials and methods
Twenty PSEN1 muation carriers were included in this study, all of
whom had pathogenic mutations that have previously been reported
elsewhere. Ten of these were presymptomatic mutation carriers; three
M146I, two E184D and one each of M139V, Y115C, intron 4, L171P,
L262F. Ten were symptomatic mutation carriers; three intron 4, two
E280G and one each of Y115H, E120K, P264L, R269H and R278I. A
group of 20 healthy control subjects (including two gene-negative
siblings) was also studied. All subjects in the study underwent assess-
ment with the Mini-Mental State Examination. The PSEN1 mutation
carriers also underwent neurological examination and detailed neuro-
psychological assessment. The neuropsychological test battery com-
prised measuresof general intellectual
Abbreviated Scale of Intelligence), estimated premorbid IQ (National
AdultReadingTest); verbal and
(Recognition Memory Test for words and faces); naming (Graded
Naming Test); calculation (Graded Difficulty Arithmetic Test); visuoper-
ceptual skills (object decision test from the Visual Object and Space
Perception battery); speed and executive function (Stroop Test). All
mutation carriers identified a close informant who was interviewed
separately to gain a collateral history; specific enquiry was made re-
garding cognitive symptoms in each domain, anxiety and depression.
Mutation carriers were defined as symptomatic if consistent symptoms
of cognitive decline were reported by the subject and/or their close
informant. If there was a discrepancy between the opinion of the
subject and their informant, the perspective of the informant was fa-
voured as insight may not always be preserved in familial Alzheimer’s
disease. Estimated time to onset was calculated for the presympto-
matic mutation carriers by subtracting the participant’s current age
from the age at which their parent first developed symptoms of pro-
gressive cognitive decline. The participants were recruited from an
ongoing longitudinal study of familial Alzheimer’s disease at the
Dementia ResearchCentre, University
Institute of Neurology, which receives research referrals from across
the UK. Some of the subjects had participated in our previous imaging
studies of familial Alzheimer’s disease (Ridha et al., 2006; Knight et al.,
2011a, b) but the data used for this study have not been reported
elsewhere. All of the subjects in this study were aware of their muta-
tion status, having undergone clinical diagnostic or predictive genetic
testing at approved clinical centres separately to their involvement in
research. All subjects gave written informed consent according to the
Declaration of Helsinki and approval was received from the local ethics
committee. Consent was taken by a clinician experienced in the as-
sessment of patients with cognitive impairment and all subjects were
considered to have capacity to consent according to the Mental
Capacity Act of 2005. Subject demographics are detailed in Table 1.
College London (UCL)
Magnetic resonance image acquisition
All subjects were scanned on the same 3 T Siemens TIM Trio scanner
using a 32-channel phased array head-coil. A sagittal 3D MP-RAGE T1-
time = 2.9/2200/900ms, dimensions of 256 ? 256 ? 208, voxel size
of 1.1 ? 1.1 ? 1.1mm) and a coronal T2FLAIR sequence (echo time/
repetitiontime/inversion time = 87/9000/2500ms,
0.9375 ? 0.9375 ? 5mm) were acquired. Two 64-direction DTI se-
quences were acquired with a single shot, spin-echo echo planar
imaging (EPI) sequence (field of view 240 mm; matrix 96 ? 96; yielding
an isotropic voxel size of 2.5 ? 2.5 ? 2.5mm; 55 contiguous axial slices;
repetition time: 6800ms; echo time: 91ms; b-value: 1000s/mm2), aug-
mented with parallel imaging acceleration (GRAPPA). Nine acquisitions
without diffusion weighting were acquired (b = 0s/mm2). For all sub-
jects, volumetric T1, DTI and T2FLAIR images were assessed visually in
all planes toensure adequatecoverage andtoexclude significant motion,
artefacts or unexpected pathology.
Thalamus and caudate in familial ADBrain 2013: Page 3 of 16 |
at University College London on April 16, 2013
Grey matter region of interest
Region of interest segmentation in T1-weighted space was used to
investigate the volumes of the subcortical structures of interest in
this study; namely the thalamus, caudate, putamen and hippocampus.
The Multi-Atlas Propagation and Segmentation technique was used
(Keihaninejad et al., 2011), which was previously developed for hip-
pocampal segmentation and has been used in brain extraction (Leung
et al., 2010, 2011). In Multi-Atlas Propagation and Segmentation, all
of the atlases in a template library are first compared with the target
image. Our template library consisted of 30 subjects (median age of
31 ? 8 years, 15 females) whose MRI scans had been manually seg-
mented into 83 anatomical structures (Hammers et al., 2003). Multiple
best-matched atlases were then selected, and the labels in the selected
atlases were propagated to the target image after non-linear image
registration. Label fusion was then applied to combine the labels from
different atlases to create a consensus segmentation in the target
image. We applied Multi-Atlas Propagation and Segmentation to
30 atlas images in a leave-one-out approach in order to determine
the number of best-matched atlases (7-9) and the optimal label
fusion technique [simultaneous truth and performance level estimation
(STAPLE)] (Warfield et al., 2004) required to produce accurate indi-
vidual regions of interest by comparing them to the manually deli-
neated regions. We used the optimized Multi-Atlas Propagation and
Segmentation technique to generate individual regions of interest for
each subject in our data set (Fig. 1). The anatomical validity of the
segmentations was confirmed by two consultant neuroradiologists.
Grey matter region of interest diffusion
tensor imaging analysis
All DTI were registered to the first b = 0 image using 12 degrees of free-
dom FLIRT (FMRIB Software Library) (Jenkinson et al., 2002) for motion
and eddy current correction and realigning diffusion-weighting direc-
tions appropriately. Tensor fitting was performed using Camino (Cook
et al., 2006). For each subject, the T1-weighted image was registered to
the first (b = 0) DTI image using a 12-parameter affine registration
Table 1 Subject demographics, neuropsychological and clinical data
(4 males, 6 females)
(6 males, 4 females)
Age, years 37.8 (4.7)49 (9.4) 44.3 (12.7) 0.27a(SMC versus control)
0.04a(PMC versus control)
50.01a(PMC versus SMC)
Years from estimated age at onset
Disease duration, years
MMSE/ 30 (range)
50.01b(SMC versus control)
0.02b(PMC versus control)
50.01b(PMC versus SMC)
Years of education 13 (2.5)12 (1.7)N/A 0.20a
NART estimated IQ
WASI Verbal IQ
WASI Performance IQ
RMT words (/50)
RMT faces (/50)
Graded naming test (/30)
Graded difficulty arithmetic (/24)
Object decision (/20)
Stroop colour (s)
Stroop word (s)
Stroop ink colour (s)
Values are mean (SD); MMSE = Mini-Mental State Examination; RMT = Recognition Memory Test.
*n = 9 for all neuropsychological measures except Wechsler Abbreviated Scale of Intelligence Performance IQ (WASI PIQ; n = 10), National Adult Reading Test (NART;
n = 7) and Stroop colour ink interference (n = 5), with all reductions in n reflecting participants being untestable on specific tasks. Data for the complete cohort were
available for all other statistical tests.
**Percentage in group manifesting the clinical feature.
aUnpaired t-test, two-tailed test.
bWilcoxon ranksum test.
cLinear regression with group and age as independent variables.
Brain 2013: Page 4 of 16N. S. Ryan et al.
at University College London on April 16, 2013
algorithm (Ourselin et al., 2001). The regions of interest defined using
Multi-Atlas Propagation and Segmentation were transferred to the DTI
space using nearest-neighbour interpolation. The overall average frac-
tional anisotropy and mean diffusivity values were calculated by aver-
aging values for the entire individual region of interest.
White matter tract of interest diffusion
tensor imaging analysis
White matter tract of interest analysis was performed using DTI-TK to
create a study-specific template using iterative tensor-based registra-
tion, which takes into account local fibre orientation (Zhang et al.,
2006; Keihaninejad et al., 2012a). In particular, all linear and
non-linear registrations on the tensor images were performed using
DTI-TK, which has recently been shown to outperform the more con-
ventional tools (Wang et al., 2011). Fractional anisotropy, mean, axial
and radial diffusivity maps were created for the study-specific template
and registered images. The ICBM-DTI-81 white matter labels and
tracts atlas, developed by Johns Hopkins University was used to
locate the white matter tracts of interest (Mori et al., 2005). The
fractional anisotropy map of the Johns Hopkins University atlas was
linearly and non-linearly registered to the study-specific template frac-
tional anisotropy map (Ourselin et al., 2001; Modat et al., 2010). This
transformation was used to warp the labels from the white matter
atlas to the template fractional anisotropy image through nearest
neighbour interpolation. In this study, we focused on six white
matter tracts of interest: the genu, body, and splenium of the
corpus callosum, fornix and left and right cingulum (Fig. 1).
Statistical tests for non-voxelwise
Baseline characteristics and neuropsychological test results (which were
converted to z scores, based on published normative data) were com-
pared using t-tests for normally distributed variables and Wilcoxon
rank sum testing for the Mini-Mental State Examination. Linear regres-
sion was used to factor out age effects on neuropsychological tests
where no detailed age corrections were available (Recognition
Memory Test, Graded Naming Test, Graded Difficulty Arithmetic and
Object Decision). For the region of interest analyses, groups were
compared using separate linear regressions for each outcome measure,
in each region of interest. The region of interest volumetric analyses
were adjusted for age, gender and total intracranial volume. The
region of interest DTI analyses for grey and white matter were ad-
justed for age and gender. Total intracranial volume correction is not
required for DTI as it assesses microstructural properties that are not
expected to be associated with macroscopic head size. All results are
reported as statistically significant if P50.05.
Voxel-based morphometry analysis of
Voxel-based morphometry processing was carried out using SPM8
(Statistical Parametric Mapping, version 8; Wellcome Trust Centre
for Neuroimaging, London, UK). The T1-weighted scans were seg-
mented into grey and white matter using the new segment toolbox
Figure 1 Regions of interest. Grey matter regions of interest (top): segmentations of thalamus (blue), caudate (green), putamen (pink)
and hippocampus (purple) overlaid on the T1image of a single subject. White matter tracts of interest (bottom): segmentations of the
fornix (pale blue), right cingulum (purple), left cingulum (green), genu (yellow), body (red) and splenium (dark blue) of the corpus
callosum, overlaid on the study-specific template.
Thalamus and caudate in familial ADBrain 2013: Page 5 of 16 |
at University College London on April 16, 2013
with default settings (Weiskopf et al., 2011). Segmentations were
produced withrigid alignment
Neurological Institute (MNI) space] and resampled to 1.5mm isotropic
voxels for use with DARTEL (Ashburner, 2007). DARTEL then itera-
tively registered the grey and white matter segments to an evolving
estimate of their group-wise average (Ashburner and Friston, 2009).
The native space tissue segments were then normalized to MNI space
using the DARTEL transformations, modulated to account for volume
changes. A smoothing kernel of 6mm full-width at half-maximum was
applied. Total intracranial volume was calculated for each subject using
Jacobian integration of deformation fields created by the new segment
toolbox (Ridgway et al., 2011). Voxel-wise statistical analysis was per-
formed with non-parametric permutation-testing using the FMRIB
Software Library randomize program (as used for the TBSS statistical
analysis, described below). Grey and white matter images were mod-
elled separately in terms of group (control, presymptomatic, symptom-
atic), age, gender and total intracranial volume (all mean centred).
Statistical significance of differences between groups was assessed
using threshold-free cluster enhancement (Smith and Nichols, 2009),
controlling family-wise error (FWE). Results for symptomatic patients
are shown thresholded at FWE P50.05; presymptomatic differences
are shown at a less stringent level of FWE P50.1, with effect-maps
(difference between adjusted group means as a percentage of the
control mean) provided for a fuller characterization of the atrophy
Voxel-wise diffusion tensor imaging
analysis using tract-based spatial
The DTI data were also analysed with TBSS to investigate whole-brain
white matter microstructural abnormalities (Smith et al., 2006). In this
study the modified TBSS protocol was used by incorporating a
group-wiseatlasas the registration
(Keihaninejad et al., 2012b). In the group-wise TBSS pipeline, a
group-wise atlas is first created based on the subset subjects’ fractional
anisotropy images using the method described in Keihaninejad et al.
(2012b). The voxel-wise statistical analysis was performed by the FSL
randomize program (5000 random permutations) with correction for
multiple comparisons to control FWE (P50.05) using threshold-free
target; group-wise TBSS
cluster enhancement. Age and gender were entered into the analysis
The subjects’ demographic, neuropsychological and clinical data
are summarized in Table 1. The presymptomatic mutation carrier
(PMC) group was, as expected, younger than the symptomatic
mutation carrier (SMC) group and was on average 5.6 years
below the expected age at symptom onset. The mean age of
the control group (44.3 years) fell between that of the PMC
(37.8 years) and SMC groups (49.0 years) and age was used as
a covariate in all analyses, as was gender. There was no clinically
relevant difference in Mini-Mental State Examination score be-
tween the control subjects and presymptomatic mutation carriers;
symptomatic mutation carriers had significantly lower Mini-Mental
State Examination scores than both groups as expected.
The PMC and SMC groups were well matched for educational
level and estimated premorbid IQ. Figure 2 shows mean and
standard error z-scores for the PMC and SMC groups on neuro-
psychological tests. The PMC group showed no significant deficit
on any neuropsychological measure. The SMC group had signifi-
cantly poorer performance than the PMC group on all tests except
naming (Graded Naming Test) and object perception.
The clinical data showed that neuropsychiatric symptoms were
common in the SMC group; 40% had anxiety and 60% had de-
pression. In the PMC group, 10% had anxiety and 20% had de-
pression.The two presymptomatic
symptoms of anxiety and/or depression were both within 2
years of their expected age at onset; both had a long history of
recurrent psychiatric symptoms since their teenage years and
required psychiatric review and medication changes at the time
of their current assessment. None of the subjects had seizures
but 60% of the symptomatic mutation carriers and 40% of the
presymptomatic mutation carriers had myoclonus on neurological
examination. Myoclonus was the only abnormal sign present on
examination of the presymptomatic mutation carriers but 20% of
the symptomatic mutation carriers had extrapyramidal signs and
Figure 2 Mean and standard error z-scores for presymptomatic and symptomatic mutation carriers on standard neuropsychological tests.
GDA = Graded Difficulty Arithmetic; GNT = Graded Naming Test; PIQ = performance IQ; RMT = Recognition Memory Test; VIQ = verbal IQ.
Brain 2013: Page 6 of 16 N. S. Ryan et al.
at University College London on April 16, 2013
60% had pyramidal signs, typically a symmetrical spastic increase
in lower limb tone with hyper-reflexia.
The grey matter region of interest volumetric analysis demon-
strated significant atrophy of the left thalamus in presymptomatic
mutation carriers compared with control subjects (P = 0.023), to-
gether with significant atrophy of both left caudate (P = 0.025)
and right caudate (P = 0.001). Refer to Table 2 for all the grey
matter region of interest volumetric and DTI results. No areas of
significant grey or white matter volume loss were seen in the PMC
group versus control subjects in the less sensitive voxel-based
morphometry analysis after correction for multiple comparisons
at the conventional level of FWE P50.05. However, atrophy in
the bilateral thalamic region was near significant [FWE P = 0.061,
at MNI coordinates of (4.5, ?13.5, 12) mm]. Results thresholded
at FWE P50.1, together with an unthresholded effect map are
shown in Fig. 3. All of the grey matter regions of interest were
significantly atrophied in the SMC group compared with control
subjects. The voxel-based morphometry results for the SMC versus
control group comparison demonstrated grey matter volume loss
in bilateral hippocampus and posterior cortical areas but the most
significant volume loss was seen in the region of bilateral striatum
and thalamus. Extensive loss of white matter volume was also
observed in the symptomatic mutation carriers compared with
control subjects (Fig. 4).
In the presymptomatic group, no significant hippocampal atro-
phy was evident, although the right hippocampus did show sig-
nificantly decreased mean diffusivity (P = 0.046) in the grey matter
region of interest DTI analysis relative to control subjects.
Although not reaching significance, there was also a trend towards
decreased mean diffusivity in all the other subcortical structures in
presymptomatic mutation carriers. This contrasts with the findings
in the symptomatic mutation carriers, who demonstrated increased
mean diffusivity in all the subcortical grey matter structures as-
sessed; the pattern of change in diffusivity that is more typically
observed in neurodegenerative disease. In the presymptomatic
mutation carriers, fractional anisotropy was increased in the left
thalamus (P = 0.016), right thalamus (P = 0.036) and left caudate
(P = 0.004). Fractional anisotropy was also increased in symptom-
atic mutation carriers in left putamen (P = 0.032) and right puta-
men (P = 0.020), but did not differ significantly from control
subjects in the other grey matter structures assessed.
In the white matter tract of interest analysis, presymptomatic
mutation carriers demonstrated
(P = 0.043) and axial diffusivity (P = 0.011) in the right cingulum
only (Table 3). Significantly decreased fractional anisotropy with
increased mean, axial and radial diffusivity was seen in all the
white matter tracts of interest in the symptomatic mutation car-
riers. The TBSS analysis demonstrated widespread decreased frac-
tional anisotropy and increased mean, axial and radial diffusivity in
the white matter of symptomatic mutation carriers. As demon-
strated in Fig. 5, the fall in fractional anisotropy in the fornix
and cingulum of the SMC group appeared to be driven by an
increase in radial diffusivity; very little difference in axial diffusivity
was seen in these structures between the symptomatic mutation
carriers and control subjects. No significant differences in any of
the DTI indices were seen in the TBSS analysis for presymptomatic
mutation carriers versus control subjects.
As a supplementary analysis, region of interest measures for the
PMC versus SMC groups were compared. Significantly decreased
fractional anisotropy with increased mean, axial and radial diffu-
sivity was found in all white matter regions of interest in symp-
tomatic mutation carriers. All grey matter regions of interest
except from the caudates had significantly smaller volumes in
symptomatic mutation carriers compared with presymptomatic
mutation carriers. Whereas fractional anisotropy did not differ,
mean diffusivity was significantly increased in all grey matter re-
gions of interest of symptomatic mutation carriers except right
thalamus and left caudate, where the increase was of borderline
significance (P = 0.061 and P = 0.075, respectively).
It has been established from Pittsburgh compound B-PET imaging
studies that the striatum and thalamus are affected early in the
presymptomatic stage of familial Alzheimer’s disease by amyloid
deposition (Klunk et al., 2007; Knight et al., 2011b). We demon-
strated atrophy of the caudate and thalamus in a cohort of pre-
symptomatic mutation carriers who were on average 5.6 years
prior to their expected age at onset. Half of these presymptomatic
mutation carriers had participated in a previous Pittsburgh com-
pound B-PET study, which found increased thalamic and striatal
amyloid deposition compared with control subjects (Knight et al.,
2011b). The findings of presymptomatic familial Alzheimer’s dis-
ease studies will clearly depend on how far from symptom onset
the individuals in the study are. A previous study examining the
caudate in presymptomatic familial Alzheimer’s disease interest-
ingly demonstrated an unexpected increase in the volume of the
caudate, precuneus and parietotemporal areas in presymptomatic
mutation carriers who were on average 9.9 years from expected
age at onset (Fortea et al., 2010). The authors of this paper
speculated that reactive neuronal hypertrophy and/or inflamma-
tory processes early in the presymptomatic stage may account for
an initial increase in the volume of these structures, which then
undergo progressive atrophy as the disease progresses. Although
their findings, like ours, need to be replicated in larger longitudinal
studies, it is quite possible that pathological processes early in the
presymptomatic stage cause dynamic changes in the volume of
affected structures, which alter in terms of the direction of change
as symptom onset is approached. Their DTI results, like ours,
showed a widespread increase in mean diffusivity in symptomatic
mutation carriers and a contrasting decrease in mean diffusivity in
the presymptomatic mutation carriers.
Increased mean diffusivity and decreased fractional anisotropy
are thought to reflect loss of integrity of cellular structures and are
commonly observed in neurodegenerative diseases, as was the
case with the symptomatic mutation carriers in our study. The
presymptomatic mutation carriers, on the other hand, showed sig-
nificantly decreased mean diffusivity in the right hippocampus and
a trend towards decreased mean diffusivity in the other subcortical
structures studied. Decreased mean diffusivity may reflect earlier
pathological changes in the pathway towards neurodegeneration,
such as microglial activation/accumulation and swelling of neurons
and glia. In a triple transgenic mouse model of Alzheimer’s disease
Thalamus and caudate in familial ADBrain 2013: Page 7 of 16 |
at University College London on April 16, 2013
the appearance of neuritic plaques is associated with hippocampal
astrogliosis and microglial activation and is preceded by an in-
crease in the density of resting microglia (Olabarria et al., 2010;
Rodriguez et al., 2010). A DTI study of the APPsw transgenic
mouse found that the time of significant amyloid plaque accumu-
lation coincided with a fall in the magnitude of diffusion in
hippocampus and cerebral cortex (Sun et al., 2005). Microglial
activation PET studies have demonstrated increased thalamic bind-
ing following traumatic brain injury (Ramlackhansingh et al.,
2011) and middle cerebral artery infarction (Pappata et al.,
2000), with decreased thalamic mean diffusivity seen initially in
acute stroke (Herve et al., 2005), perhaps supporting the
Table 2 Mean ? SD volumes and diffusivity indices in the grey matter regions of interest
Control subjects Presymptomatic mutation carriers Symptomatic mutation
Right thalamus7792 ? 5507676 ? 747
275 (?152, 703)
6962 ? 604
428 (64, 792)*
4273 ? 553
605 (293, 917)*
4376 ? 610
466 (63, 868)*
5234 ? 695
144 (?250, 537)
4989 ? 775
194 (?232, 620)
3153 ? 483
?1.76 (?233, 230)
3118 ? 423
?116 (?345, 112)
6103 ? 543
1620 (1294, 1947)*
5520 ? 404
1533 (1216, 1849)*
3832 ? 525
746 (438, 1053)*
3748 ? 588
940 (537, 1343)*
4228 ? 473
920 (555, 1285)*
3847 ? 655
1142 (736, 1548)*
2295 ? 477
868 (632, 1103)*
2302 ? 459
769 (546, 991)*
Left thalamus 7177 ? 652
Right caudate4689 ? 512
Left caudate4743 ? 506
Right putamen5210 ? 656
Left putamen 5035 ? 680
Right hippocampus3131 ? 304
Left hippocampus3014 ? 278
Right thalamus0.352 ? 0.0170.367 ? 0.019
?0.017 (?0.032, ?0.001)*
0.343 ? 0.008
?0.013 (?0.023, ?0.003)*
0.239 ? 0.030
?0.017 (?0.038, 0.004)
0.248 ? 0.026
?0.025 (?0.041, ?0.009)*
0.227 ? 0.013
?0.009 (?0.021, 0.003)
0.214 ? 0.010
?0.005 (?0.013, 0.003)
0.193 ? 0.022
?0.006 (?0.0214, 0.009)
0.192 ? 0.020
?0.012 (?0.029, 0.005)
0.346 ? 0.031
0.006 (?0.014, 0.025)
0.327 ? 0.031
0.007 (?0.011, 0.025)
0.243 ? 0.034
?0.013 (?0.037, 0.010)
0.236 ? 0.027
?0.011 (?0.029, 0.008)
0.242 ? 0.020
?0.017 (?0.031, ?0.003)*
0.228 ? 0.021
?0.014 (?0.027, ?0.001)*
0.176 ? 0.032
0.004 (?0.014, 0.023)
0.167 ? 0.033
0.007 (?0.013, 0.028)
Left thalamus 0.332 ? 0.013
Right caudate0.229 ? 0.024
Left caudate0.226 ? 0.017
Right putamen 0.222 ? 0.016
Left putamen0.212 ? 0.011
Right hippocampus0.184 ? 0.016
Left hippocampus0.178 ? 0.019
Right thalamus0.824 ? 0.030 0.794 ? 0.043
0.022 (?0.006, 0.049)
0.815 ? 0.037
0.016 (?0.010, 0.042)
0.949 ? 0.068
0.005 (?0.057, 0.066)
0.955 ? 0.066
0.013 (?0.045, 0.071)
0.740 ? 0.028
0.752 ? 0.021
0.012 (?0.016, 0.039)
0.907 ? 0.073
0.044 (0.002, 0.087)*
0.921 ? 0.053
0.027 (?0.017, 0.072)
0.912 ? 0.089
?0.075 (?0.119, ?0.032)*
0.915 ? 0.069
?0.070 (?0.108, ?0.032)*
1.075 ? 0.111
?0.113 (?0.185, ?0.040)*
1.058 ? 0.072
?0.052 (?0.112, 0.008)
0.840 ? 0.060
0.838 ? 0.061
1.094 ? 0.073
1.113 ? 0.096
Left thalamus0.838 ? 0.029
Right caudate0.956 ? 0.071
Left caudate 0.993 ? 0.081
Right putamen 0.761 ? 0.026
Left putamen 0.773 ? 0.040
Right hippocampus 0.958 ? 0.033
0.952 ? 0.052
The differences in adjusted means between the mutation carriers and control group (95% confidence interval) are shown in italics. Significant results
at P50.05 are indicated by an asterisk.
Brain 2013: Page 8 of 16N. S. Ryan et al.
at University College London on April 16, 2013
hypothesis that gliosis accounts for the low mean diffusivity
observed in presymptomatic mutation carriers. Microglia undergo
persistent activation at sites of white matter damage (Wilson
et al., 2004), which may explain why they are observed in densely
connected subcortical structures. In axonal injury, APP is upregu-
lated in white matter at the site of trauma but is later seen in the
dorsal thalamus (Spain et al., 2010), where neuronal atrophy, but
not cell death, has been observed (Lifshitz et al., 2007).
The PMC group in our study showed increased fractional an-
isotropy in bilateral thalamus and left caudate and, in the SMC
group, increased putamen fractional anisotropy was seen bilat-
erally. Fractional anisotropy reflects the degree of directionality
of diffusion of water molecules within a structure; it is therefore
highest in white matter tracts, lower in grey matter and lowest in
CSF. The anisotropy of the thalamus and striatum is probably due
to their compartmentalized cytoarchitectural organization, to-
gether with the large number of white matter fibres passing
through them (Wiegell et al., 2000). Similar to our findings, a
pattern of increased fractional anisotropy in the thalamus and
basal ganglia has been observed in patients with multiple sclerosis
(Ciccarelli et al., 2001; Tovar-Moll et al., 2009). It has been sug-
gested that this may be due to the selective degeneration of white
matter fibres connecting these subcortical structures with cortical
areas combined with the relative preservation of anisotropic intrin-
sic connections between the structures themselves. This may result
in an overall increase in the fractional anisotropy of the subcortical
structures. By the same presumed mechanism, increased fractional
anisotropy was found in a group of patients with mild cognitive
impairment compared with control subjects in the centrum semi-
ovale, due to degeneration of association fibres of the superior
longitudinal fasciculus where they cross with intact motor projec-
tion fibres (Douaud et al., 2011).
Decreased mean diffusivity and axial diffusivity were observed
in the right cingulum of the PMC group, whereas increased mean,
axial and radial diffusivity occurred in all white matter regions
of interest of the symptomatic mutation carriers. This pattern of
events also conforms with the findings of pathological studies of
axonal damage. Axonal injury initially causes swelling, beading and
fragmentation of axons with an associated fall in axial diffusivity,
probably because intracellular diffusion is hindered by the
impedes diffusion in both the intracellular and extracellular space
(Budde and Frank, 2010). Following subsequent clearance of
membrane fragments, the decreased axonal density permits
increased parallel diffusion in the extra-axonal space and therefore
increased axial diffusivity, together with increased radial diffusivity
Figure 3 Voxel-based morphometry results showing (top left) areas of grey matter reduction in presymptomatic mutation carriers
compared with control subjects after correction for multiple comparisons at FWE50.1 and the effect-map for the PMC versus control
group comparison. In the effect-map, regions showing reduced grey matter in the PMC group are shown in red and in control subjects in
blue. Differences between the two adjusted group-means are overlaid on a mean study-specific template. Images shown in radiological
convention (right on left).
Thalamus and caudate in familial ADBrain 2013: Page 9 of 16 |
at University College London on April 16, 2013
secondary to myelin degradation (Song et al., 2003; Concha et al.,
2006). In the APPsw transgenic mouse DTI study, decreased axial
diffusivity was seen diffusely in the white matter from the time of
amyloid plaque accumulation, with increased radial diffusivity
occurring in the corpus callosum at a later stage (Sun et al.,
2005). Although each variable was examined separately, the
TBSS results in our study suggest a more prominent increase in
radial than axial diffusivity in symptomatic mutation carriers com-
pared with control subjects, particularly in the fornix and cingu-
lum, which may be because of this gradual transition of axial
diffusivity from decreased to increased with advancing axonal de-
generation. Other studies of early Alzheimer’s disease have also
found more prominent increases in radial than axial diffusivity,
with axial diffusivity decreased in some tracts (Huang et al.,
Several pathogenic pathways are likely to contribute to the
axonal degeneration witnessed in Alzheimer’s disease, including
defectsin fast axonal transport resultingfrom enhanced
production of amyloid-b (Kanaan et al., 2013). Neurons, having
longer dendrites and axons than any other cell type, are particu-
larly vulnerable to impaired fast axonal transport, which may be
why familial Alzheimer’s disease-causing mutations selectively
affect neurons despite being ubiquitously expressed in cells
throughout the body. It may also explain why short interneurons
tend to be relatively spared in Alzheimer’s disease compared with
long projection neurons (Mattson and Magnus, 2006), an obser-
vation that could well underlie our finding of increased fractional
anisotropy in the thalamus and caudate in presymptomatic familial
Alzheimer’s disease. The cingulum bundle, where we observed
presymptomatic diffusivity changes, is composed of particularly
long-coursing white matter tracts, arising from the thalamus, cin-
gulate gyrus and cortical association areas. It is the major route for
projections from the anterior thalamus to cingulate cortex, hippo-
campus and retrohippocampal regions; a network that is thought
to play a key role in memory and spatial orientation (Neave et al.,
1997). These cognitive functions are often impaired early in
Figure 4 Voxel-based morphometry results showing (top) grey matter reduction and (bottom) white matter reduction in symptomatic
mutation carriers compared with control subjects. Results are shown overlaid on a mean study-specific template and are FWE-corrected
for multiple comparisons. Images shown in radiological convention (right on left).
Brain 2013: Page 10 of 16 N. S. Ryan et al.
at University College London on April 16, 2013
Alzheimer’s disease, perhaps because neurons lose their functional
connections with increasing axonal damage. Eventually, once the
damage has reached a stage where trophic signalling is no longer
available, overt neurodegeneration may occur.
A number of strands of evidence suggest that a network of
limbic neurons undergoes selective
Alzheimer’s disease, including PET findings of hypometabolism in
the thalamus, mamillary bodies, hippocampus and posterior
Table 3 Mean ? SD diffusivity indices in the white matter tracts of interest
Genu of corpus callosum 0.610 ? 0.026 0.621 ? 0.040
0.010 (?0.017, 0.037)
0.613 ? 0.045
?0.000 (?0.029, 0.028)
0.691 ? 0.028
0.005 (?0.015, 0.025)
0.458 ? 0.094
0.017 (?0.036, 0.069)
0.503 ? 0.059
0.018 (?0.018, 0.054)
0.467 ? 0.050
0.011 (?0.023, 0.044)
0.570 ? 0.010
?0.040 (?0.059, ?0.021)*
0.559 ? 0.012
0.639 ? 0.020
0.332 ? 0.046
0.402 ? 0.019
?0.078 (?0.102, ?0.054)*
0.385 ? 0.018
Body of corpus callosum0.612 ? 0.026
Splenium of corpus callosum0.684 ? 0.020
Fornix 0.426 ? 0.044
Right cingulum 0.481 ? 0.033
Left cingulum 0.457 ? 0.035
Genu of corpus callosum0.849 ? 0.0430.835 ? 0.057
0.010 (?0.031, 0.052)
0.855 ? 0.052
0.847 ? 0.068
0.002 (?0.042, 0.046)
1.642 ? 0.388
0.007 (?0.214, 0.229)
0.776 ? 0.047
0.040 (0.001, 0.078)*
0.752 ? 0.062
0.019 (?0.028, 0.066)
0.942 ? 0.026
0.090 (0.058, 0.122)*
0.966 ? 0.041
0.105 (0.071, 0.139)*
0.974 ? 0.068
0.117 (0.073, 0.160)*
2.107 ? 0.147
.369 (0.215, 0.523)*
0.980 ? 0.049
0.157 (0.119, 0.196)*
0.911 ? 0.040
0.132 (0.092, 0.172)*
Body of corpus callosum0.858 ? 0.039
Splenium of corpus callosum0.855 ? 0.040
Fornix1.718 ? 0.199
Right cingulum0.821 ? 0.043
Left cingulum0.776 ? 0.050
Genu of corpus callosum1.543 ? 0.0381.528 ? 0.041
0.010 (?0.023, 0.044)
1.562 ? 0.021
0.003 (?0.027, 0.033)
1.653 ? 0.075
0.004 (?0.044, 0.052)
2.454 ? 0.335
0.021 (?0.177, 0.219)
1.248 ? 0.020
0.042 (0.010, 0.074)*
1.161 ? 0.039
0.018 (?0.021, 0.057)
1.636 ? 0.031
0.088 (0.059, 0.118)*
1.661 ? 0.041
0.087 (0.054, 0.120)*
1.795 ? 0.078
0.130 (0.082, 0.178)*
2.832 ? 0.136
0.279 (0.134, 0.424*)
1.394 ? 0.054
0.102 (0.062, 0.142)*
1.281 ? 0.043
0.088 (0.049, 0.128)*
Body of corpus callosum 1.569 ? 0.040
Splenium of corpus callosum1.662 ? 0.043
Fornix2.536 ? 0.188
Right cingulum 1.291 ? 0.046
Left cingulum 1.188 ? 0.049
Genu of corpus callosum0.503 ? 0.048 0.488 ? 0.068
0.010 (?0.037, 0.057)
0.502 ? 0.069
?0.002 (?0.047, 0.043)
0.444 ? 0.066
0.001 (?0.042, 0.044)
1.236 ? 0.416
0.001 (?0.234, 0.235)
0.540 ? 0.073
0.039 (?0.012, 0.089)
0.547 ? 0.079
0.019 (?0.036, 0.075)
0.595 ? 0.026
0.090 (0.055, 0.125)*
0.618 ? 0.042
0.114 (0.078, 0.151)*
0.563 ? 0.064
0.110 (0.068, 0.152)*
1.744 ? 0.155
0.415 (0.255, 0.575)*
0.772 ? 0.048
0.185 (0.143, 0.227)*
0.725 ? 0.043
0.154 (0.109, 0.198)*
Body of corpus callosum 0.503 ? 0.043
Splenium of corpus callosum 0.451 ? 0.041
Fornix 1.310 ? 0.206
Right cingulum 0.585 ? 0.050
Left cingulum 0.569 ? 0.057
The differences in adjusted means between the mutation carriers and control group (95% confidence interval) are shown in italics. Significant results at P50.05 are
indicated by an asterisk.
Thalamus and caudate in familial AD Brain 2013: Page 11 of 16 |
at University College London on April 16, 2013
cingulate (Nestor et al., 2003). Along with the cingulum findings,
our observation of presymptomatic thalamic atrophy and increased
thalamic fractional anisotropy in familial Alzheimer’s disease adds
further support to this notion. The few pathological studies of
thalamic involvement in Alzheimer’s disease have demonstrated
extracellular amyloid deposits in almost all thalamic nuclei (Braak
and Braak, 1991) and significant thalamic atrophy, thought largely
to be due to loss of axons, dendrites and synaptic structures or
glial cell changes (Xuereb et al., 1991).
The striatumof patients
post-mortem has also been found to contain abundant amyloid
deposits with loss of large cholinergic neurons (Oyanagi et al.,
1987; Braak and Braak, 1990), which potentially underlies our
findings of striatal atrophy in familial Alzheimer’s disease. The ma-
jority of striatal nerve cells appear to resist the development of
neuritic change (Braak and Braak, 1990) despite the presence of
abundant amyloid, which is intriguing given that striatal involve-
ment on Pittsburgh compound B-PET scans is evident over a
withAlzheimer’s disease at
decade before symptom onset in familial Alzheimer’s disease.
One possible hypothesis relates to differences in mechanisms of
calcium homeostasis in different neuronal populations. It has been
found that presenilins act as calcium leak channels in the endo-
plasmic reticulum of hippocampal neurons and many PSEN1 mu-
tations impair this leak function, resulting in endoplasmic reticulum
calcium overload and supranormal calcium release. In medium
spiny striatal neurons (the major neuronal cell type in the stri-
atum), the endoplasmic reticulum is much less leaky for calcium
and presenilins show less involvement in calcium homeostasis
(Nelson et al., 2010).
The idea that widespread white matter degeneration occurs in
familial Alzheimer’s disease is supported by the imaging findings in
our SMC cohort. Volume loss of the white matter appeared to be
more widespread than that of the grey matter in the voxel-based
morphometry analysis, and the TBSS results showed very diffuse
microstructural white matter abnormalities in the SMC cohort. A
variety of PSEN1 mutations are known to be associated with
Figure 5 TBSS results demonstrating areas of significantly decreased fractional anisotropy and significantly increased axial, radial and
mean diffusivity in symptomatic mutation carriers compared with control subjects. Results are FWE-corrected for multiple comparisons
using threshold-free cluster enhancement.
Brain 2013: Page 12 of 16N. S. Ryan et al.
at University College London on April 16, 2013
spastic paraparesis and some individuals with the phenotype have
white matter hyperintensities visible on brain MRI (O’Riordan
et al., 2002; Ryan and Rossor, 2010). It would not, therefore,
be surprising if microstructural damage to the white matter gave
rise to more subtle pyramidal signs and our observation that 60%
of the symptomatic mutation carriers had lower limb spasticity and
hyper-reflexia may well relate to this white matter pathology. It is
not clear whether there are clinical correlates of the presympto-
matic thalamic and caudate changes we observed. Although extra-
pyramidal signs, often associated with striatal pathology, were
seen in 20% of the symptomatic mutation carriers they were
not seen presymptomatically. Presymptomatic myoclonus was
observed, as has been detected in previous studies (Godbolt
et al., 2004). The origin of this myoclonus is unknown but it is
possible that thalamocortical networks are implicated, as they are
in otherconditions suchas
(Gerschlager and Brown, 2009).
Neuropsychologically, the SMC cohort showed deficits in wide-
spread domains at the time they were studied, with verbal recog-
preservation of naming and object perception, as has been
observed previously (Warrington et al., 2001; Godbolt et al.,
2004). No significant neuropsychological deficits were observed
in the PMC group, although 20% of these subjects and 60% of
the symptomatic mutation carriers had symptoms of depression
and/or anxiety. Anxiety and depression are relatively common in
both sporadic and familial Alzheimer’s disease and it is under-
standable that these symptoms may occur in presymptomatic mu-
tation carriers who are aware of their mutation status and
approaching the age at which they saw their parent develop
symptoms. However, it is worth noting that limbic circuits are
well known to be involved in emotional expression and lesions
to anterior thalamic nuclei have been found to cause a variety
of neuropsychiatric symptoms including sadness, agitation and
flattened affect, in addition to amnesia (Schmahmann, 2003).
Disentangling whether presymptomatic psychiatric symptoms are
related more to psychological factors or disease effects will require
longitudinal study of large cohorts of ‘at risk’ individuals, compris-
ing both mutation-positive and -negative individuals, who are un-
aware of their mutation status but exposed to the same anxieties
about the possibility of developing symptoms. Collaborative stu-
dies such as DIAN will be ideally placed to address these issues
and ascertain whether our findings are reproducible in larger
familial Alzheimer’s disease cohorts.
Our study had several limitations. Subject numbers were rela-
tively small, which limits power and may have accounted for the
fact that reduced thalamic volume and increased caudate frac-
tional anisotropy only reached statistical significance on the left
in the PMC region of interest analysis. Larger cohorts will be
required to investigate whether the onset of subcortical imaging
changes is indeed asymmetrical. The optimization of methods for
processing and analysing DTI data is an area of active research.
One potential issue with DTI is its sensitivity to susceptibility arte-
facts arising from neighbouring tissues having very different para-
magnetic properties, which can lead to distortions of echo planar
imaging and misregistration. The lack of echo planar imaging dis-
tortion correction in our study should be considered a minor
limitation, however the deep grey matter structures of interest in
this study do not tend to be severely affected by such artefacts
and a gold standard solution to the problem has not yet been
established (Gholipour et al., 2011). We used mean fractional an-
isotropy to investigate the regions of interest, which is common
practice in the literature, but is nevertheless an aspect that could
be improved in future work. Regional analysis (e.g. of thalamus)
cannot easily address subregional variation in the structure of
interest. Inspection of the grey matter region of interest histo-
grams indicates a degree of non-normality (distributions are
slightly skewed and heavier tailed to the right), which might mo-
tivate consideration of robust statistics, such as trimmed-mean or
an m-estimator (Hampel et al., 1986). Secondly, and more funda-
mentally, the nature of fractional anisotropy as a ratio suggests
that the use of an arithmetic mean (or robust version thereof)
might not be optimal as a measure of central tendency. For ex-
ample, a geometric mean (corresponding to a logarithmic distance
metric) is more appropriate for strictly positive but otherwise un-
bounded quantities. However, fractional anisotropy is bounded
between 0 and 1, and the optimal metric and mean are unclear.
More radically, it has been suggested that fractional anisotropy
should be replaced with a different measure of anisotropy
known as the geodesic anisotropy (Batchelor et al., 2005),
which would have a stronger theoretical basis than the combin-
ation of fractional anisotropy with robust non-arithmetic averages,
but requires further empirical evaluation.
In summary, we demonstrated presymptomatic thalamic and
caudate volume loss and increased fractional anisotropy, with
altered diffusivity in the right hippocampus and cingulum, in
PSEN1 mutation carriers who were ?5.6 years from expected
age at onset. We propose that axonal degeneration is an early
event in familial Alzheimer’s disease pathogenesis and that amyl-
oid accumulation, atrophy and diffusivity changes are seen early
on in the thalamus and striatum due to the large number of white
matter tractsthatthese structures
Involvement of a subcortical structure like the thalamus, which
forms a key node in a vulnerable neuronal network, could potenti-
ate the propagation of pathology through that network. The most
striking degeneration may then be observed in connected neuronal
populations, such as those of the hippocampus. How generalizable
our findings are to sporadic Alzheimer’s disease is an important
question. There is some evidence of early white matter tract de-
generation in sporadic Alzheimer’s disease from DTI studies of
mild cognitive impairment (Chua et al., 2008; Stebbins and
Murphy, 2009) and thalamic atrophy has also been demonstrated
in mild cognitive impairment (Pedro et al., 2012). In healthy con-
trol subjects at increased risk of sporadic Alzheimer’s disease due
to possession of an apolipoprotein E4 allele, decreased fractional
anisotropy has been found in several white matter tracts including
the cingulum (Smith et al., 2010) and volume loss of the hippo-
campus, but not thalamus or striatum, has been observed
(O’Dwyer et al., 2012). The same neuronal network appears to
disintegrate in both sporadic and familial Alzheimer’s disease but it
may be that different nodes of this circuit are particularly vulner-
able to different risk factors for the disease. The development of
more sensitive diffusion imaging techniques such as high-angular-
resolution diffusion imaging (HARDI) (Haroon et al., 2011) and
Thalamus and caudate in familial ADBrain 2013: Page 13 of 16 |
at University College London on April 16, 2013
neurite orientation dispersion and density imaging (NODDI)
(Zhang et al., 2012), and their application to the longitudinal
study of individuals at risk of both familial and sporadic
Alzheimer’s disease, may allow us to further investigate just how
early microstructural abnormalities become evident and when is
the optimal time for therapeutic intervention.
We thank the participants and their families for their generous
support of this study, our clinical colleagues across the UK for
referring patients and the MRC Prion Unit for conducting much
of the genetic analysis. We would also like to acknowledge the
anonymous reviewers who made recommendations that have
substantially improved the paper.
This work was supported by the Medical Research Council (Clinical
Research Training Fellowship G0900421 to N.S.R., Special Training
Fellowship in Biomedical Informatics MR/J014267/1 to G.R.R.,
Senior Clinical Fellowship G116/143 to NCF) and Alzheimer’s
Research UK (Senior Research Fellowship to SJC, Travelling
Research Fellowship to M.L.). The study was undertaken at
UCLH/UCL who received a proportion of funding from the
Department of Health’s National Institute for Health Research
(NIHR) Biomedical Research Centres funding scheme and was sup-
ported by the NIHR Queen Square Dementia Biomedical Research
Unit. N.C.F. and M.N.R. are NIHR senior investigators. The
Dementia Research Centre is an Alzheimer’s Research UK
Co-ordinating Centre and has also received equipment funded
by Alzheimer’s Research UK and Brain Research Trust. The
Wellcome Trust Centre for Neuroimaging is supported by core
funding from the Wellcome Trust, 091593/Z/10/Z. M.M. is sup-
ported by a Comprehensive Biomedical Research Centre Strategic
Investment Award (Ref. 168). The Engineering and Physical
Sciences Research Council support D.C.A. with grant EP/E007748.
Acosta-Cabronero J, Williams GB, Pengas G, Nestor PJ. Absolute diffu-
sivities define the landscape of white matter degeneration in
Alzheimer’s disease. Brain 2010; 133: 529–39.
Ashburner J.A fastdiffeomorphic
Neuroimage 2007; 38: 95–113.
Ashburner J, Friston KJ. Computing average shaped tissue probability
templates. Neuroimage 2009; 45: 333–41.
Basser PJ, Mattiello J, LeBihan D. MR diffusion tensor spectroscopy and
imaging. Biophys J 1994; 66: 259–67.
Batchelor PG, Moakher M, Atkinson D, Calamante F, Connelly A.
A rigorous framework for diffusion tensor calculus. Magn Reson
Med 2005; 53: 221–5.
Bateman RJ, Aisen PS, De Strooper B, Fox NC, Lemere CA, Ringman JM,
et al. Autosomal-dominant Alzheimer’s disease: a review and proposal
for the prevention of Alzheimer’s disease. Alzheimers Res Ther 2011;
Bateman RJ, Xiong C, Benzinger TL, Fagan AM, Goate A, Fox NC, et al.
Clinical and biomarker changes in dominantly Inherited Alzheimer’s
disease. N Engl J Med 2012; 367: 795–804.
Benedetti B, Charil A, Rovaris M, Judica E, Valsasina P, Sormani MP,
et al. Influence of aging on brain gray and white matter changes
assessed by conventional, MT, and DT MRI. Neurology 2006; 66:
Bohanna I, Georgiou-Karastianis N, Egan GF. Connectivity-based seg-
mentation of the striatum in Huntinton’s disease: vulnerability of
motor pathways. Neurobiol Dis 2011; 42: 475–81.
Bozzali M, Cercignani M, Sormani MP, Comi G, Filippi M. Quantification
of brain gray matter damage in different MS phenotypes by use of
diffusion tensor MR imaging. AJNR Am J Neuroradiol 2002; 23:
Braak H, Braak E. Alzheimer’s disease: striatal amyloid deposits and
neurofibrillary changes. J Neuropathol Exp Neurol 1990; 49: 215–24.
Braak H, Braak E. Alzheimer’s disease affects limbic nuclei of the thal-
amus. Acta Neuropathol 1991; 81: 261–8.
Budde MD, Frank JA. Neurite beading is sufficient to decrease the ap-
parent diffusion coefficient after ischemic stroke. Proc Natl Acad Sci
USA 2010; 107: 14472–7.
Chua TC, Wen W, Slavin MJ, Sachdev PS. Diffusion tensor imaging in
mild cognitive impairment and Alzheimer’s disease: a review. Curr
Opin Neurol 2008; 21: 83–92.
Ciccarelli O, Werring DJ, Wheeler-Kingshott CA, Barker GJ, Parker GJ,
Thompson AJ, et al. Investigation of MS normal-appearing brain using
diffusion tensor MRI with clinical correlations. Neurology 2001; 56:
Cook PA, Bai Y, Nedjati-Gilani S, Seunarine KK, Hall MG, Parker GJ,
et al. Camino: open-source diffusion-MRI reconstruction and process-
ing [abstract]. Proc Int Soc Mag Reson Med 2006; 14.
Concha L, Gross DW, Wheatley BM, Beaulieu C. Diffusion tensor ima-
ging of time-dependent axonal and myelin degradation after corpus
callosotomy in epilepsy patients. Neuroimage 2006; 32: 1090–9.
de Jong LW, van der Hiele K, Veer IM, Houwing JJ, Westendorp RG,
Bollen EL, et al. Strongly reduced volumes of putamen and thalamus in
Alzheimer’s disease: an MRI study. Brain 2008; 131: 3277–85.
Douaud G, Jbabdi S, Behrens TE, Menke RA, Gass A, Monsch A, et al.
DTI measures in crossing-fibre areas: increased diffusion anisotropy
reveals early white matter alteration in MCI and mild Alzheimer’s dis-
ease. Neuroimage 2011; 55: 880–90.
Ferri CP, Prince M, Brayne C, Brodaty H, Fratiglioni L, Ganguli M, et al.
Global prevalence of dementia: a Delphi consensus study. Lancet
2005; 366: 2112–17.
Fortea J, Sala-Llonch R, Bartres-Faz D, Bosch B, Llado ´ A, Bargallo ´ N,
et al. Increased cortical thickness and caudate volume precede atrophy
in PSEN1 mutation carriers. J Alzheimers Dis 2010; 22: 909–22.
Gerschlager W, Brown P. Myoclonus. Curr Opin Neurol 2009; 22:
Gholipour A, Kehtarnavaz N, Scherrer B, Warfield S. On the accuracy of
unwarping techniques for the correction of susceptibility-induced geo-
metric distortion in magnetic resonance echo-planar images [abstract].
Conf Proc IEEE Eng Med Biol Soc 2011; 2011: 6997–7000.
Godbolt AK, Cipolotti L, Watt H, Fox NC, Janssen JC, Rossor MN. The
natural history of Alzheimer disease: a longitudinal presymptomatic
and symptomatic study of a familial cohort. Arch Neurol 2004; 61:
Grahn JA, Parkinson JA, Owen AM. The cognitive functions of the
caudate nucleus. Prog Neurobiol 2008; 86: 141–55.
Hammers A, Allom R, Koepp MJ, Free SL, Myers R, Lemieux L, et al.
Three-dimensional maximum probability atlas of the human brain,
with particular reference to the temporal lobe. Hum Brain Mapp
2003; 19: 224–47.
Hampel FR, Ronchetti EM, Rousseeuw PJ, Stahel WA. Robust statistics:
The approach based on influence functions. New York: Wiley; 1986.
Haroon HA, Reynolds H, Carter SF, Embleton KV, Herholz KG, Parker GJ.
HARDI-based microstructural complexity mapping reveals distinct
subcortical and cortical grey matter changes in mild cognitive
Brain 2013: Page 14 of 16N. S. Ryan et al.
at University College London on April 16, 2013
impairment and Alzheimer’s disease [abstract]. Proc Int Soc Mag
Reson Med 2011; 682.
Herve D, Molko N, Pappata S, Buffon F, LeBihan D, Bousser M, et al.
Longitudinal thalamic diffusion changes after middle cerebral artery
infarcts. J Neurol Neurosurg Psychiatry 2005; 76: 200–5.
Huang J, Friedland RP, Auchus AP. Diffusion tensor imaging of
normal-appearing white matter in mild cognitive impairment and
early Alzheimer disease: preliminary evidence of axonal degeneration
in the temporal lobe. AJNR Am J Neuroradiol 2007; 28: 1943–8.
Jenkinson M, Bannister P, Brady M, Smith S. Improved optimization for
the robust and accurate linear registration and motion correction of
brain images. Neuroimage 2002; 17: 825–41.
Kanaan NM, Pigino GF, Brady ST, Lazarov O, Binder LI, Morfini GA.
Axonal degeneration in Alzheimer’s disease: when signaling abnorm-
alities meet the axonal transport system. Exp Neurol 2012; [Epub
ahead of print].
Keihaninejad S, Ryan N, Malone I, Leung K, Hyare H, Ourselin S, et al.
Automated segmentation of the thalamus for measurement of MR
diffusion indices: evaluation and application to familial Alzheimer’s dis-
ease [abstract]. Alzheimer’s Demen 2011; 7: 4, S50.
Keihaninejad S, Zhang H, Ryan N, Malone I, Frost C, Cardoso M, et al.
White matter tract changes in mild-to-moderate Alzheimer’s disease
revealed by tensor-based registration analysis of diffusion tensor ima-
ging [abstract]. Alzheimer’s Dement 2012a; 8: 4, P31.
Keihaninejad S, Ryan NS, Malone IB, Modat M, Cash D, Ridgway GR,
et al. The importance of group-wise registration in Tract Based Spatial
Statistics study of neurodegeneration: a simulation study in Alzheimer’s
disease. PLoS One 2012b; 7: e45996.
Klunk WE, Engler H, Nordberg A, Wang Y, Blomqvist G, Holt DP, et al.
Imagingbrain amyloidin Alzheimer’s
Compound-B. Ann Neurol 2004; 55: 306–19.
Klunk WE, Price JC, Mathis CA, Tsopelas ND, Lopresti BJ, Ziolko SK,
et al. Amyloid deposition begins in the striatum of presenilin-1 muta-
tion carriers from two unrelated pedigrees. J Neurosci 2007; 27:
Knight WD, Kim LG, Douiri A, Frost C, Rossor MN, Fox NC. Acceleration
of cortical thinning in familial Alzheimer’s disease. Neurobiol Aging
2011a; 32: 1765–73.
Knight WD, Okello AA, Ryan NS, Turkheimer FE, Rodrı ´guez Martinez de
Llano S, Edison P, et al. Carbon-11-Pittsburgh compound B positron
emission tomography imaging of amyloid deposition in presenilin 1
mutation carriers. Brain 2011b; 134: 293–300.
Lee GJ, Lu PH, Medina LD, Rodriguez-Agudelo Y, Melchor S,
Coppola G, et al. Regional brain volume differences in symptomatic
and presymptomatic carriers of familial Alzheimer’s disease mutations.
J Neurol Neurosurg Psychiatry 2013; 84: 154–62.
Leung KK, Barnes J, Modat M, Ridgway GR, Bartlett JW, Fox NC, et al.
Brain MAPS: an automated, accurate and robust brain extraction tech-
nique using a template library. Neuroimage 2011; 55: 1091–108.
LeungKK, BarnesJ, Ridgway
Macdonald K, et al. Automated cross-sectional and longitudinal hip-
pocampal volume measurement in mild cognitive impairment and
Alzheimer’s disease. Neuroimage 2010; 51: 1345–59.
Lifshitz J, Kelley BJ, Povlishock JT. Perisomatic thalamic axotomy after
diffuse traumatic brain injury is associated with atrophy rather than cell
death. J Neuropathol Exp Neurol 2007; 66: 218–29.
Mattson MP, Magnus T. Ageing and neuronal vulnerability. Nat Rev
Neurosci 2006; 7: 278–94.
McKinstry RC, Mathur A, Miller JH, Ozcan A, Snyder AZ, Schefft GL,
et al. Radial organization of developing preterm human cerebral cortex
revealed by non-invasive water diffusion anisotropy MRI. Cereb
Cortex 2002; 12: 1237–43.
Middleton FA, Strick PL. Basal ganglia output and cognition: evidence
from anatomical, behavioral, and clinical studies. Brain Cogn 2000; 42:
Modat M, Ridgway GR, Taylor ZA, Lehmann M, Barnes J, Hawkes DJ,
et al. Fast free-form deformation using graphics processing units.
Comput Methods Programs Biomed 2010; 98: 278–84.
GR, Bartlett JW, ClarksonMJ,
Mori S, Wakana S, Nagae-Poetscher L, van Zijl P. MRI atlas of human
white matter. Amsterdam. Elsevier, 2005.
Neave N, Nagle S, Aggleton JP. Evidence for the involvement of the
mammillary bodies and cingulum bundle in allocentric spatial process-
ing by rats. Eur J Neurosci 1997; 9: 941–55.
Nelson O, Supnet C, Liu H, Bezprozvanny I. Familial Alzheimer’s disease
mutations in presenilins: effects on endoplasmic reticulum calcium
homeostasis and correlation with clinical phenotypes. J Alzheimers
Dis 2010; 21: 781–93.
Nestor PJ, Fryer TD, Smielewski P, Hodges JR. Limbic hypometabolism in
Alzheimer’s disease and mild cognitive impairment. Ann Neurol 2003;
O’Dwyer L, Lamberton F, Matura S, Tanner C, Scheibe M, Miller J, et al.
Reduced hippocampal volume in healthy young ApoE4 carriers: an
MRI study. PLoS One 2012; 7: e48895.
O’Riordan S, McMonagle P, Janssen JC, Fox NC, Farrell M, Collinge J,
et al. Presenilin-1 mutation (E280G), spastic paraparesis, and cranial
MRI white-matter abnormalities. Neurology 2002; 59: 1108–10.
Olabarria M, Noristani HN, Verkhratsky A, Rodriguez JJ. Concomitant
astroglial atrophy and astrogliosis in a triple transgenic animal model of
Alzheimer’s disease. Glia 2010; 58: 831–8.
Ourselin S, Roche A, Subsol G, Pennec X, Ayache N. Reconstructing a
serial structure from serial histological sections. Image Vision Comput
2001; 19: 25–31.
Oyanagi K, Takahashi H, Wakabayashi K, Ikuta F. Selective involvement
of large neurons in the neostriatum of Alzheimer’s disease and
senile dementia: a morphometric investigation. Brain Res 1987; 411:
Pappata S, Levasseur M, Gunn RN, Myers R, Crouzel C, Syrota A, et al.
Thalamic microglial activation in ischemic stroke detected in vivo by
PET and [11C]PK1195. Neurology 2000; 55: 1052–4.
Pedro T, Weiler M, Yasuda CL, D’Abreu A, Damasceno BP, Cendes F,
et al. Volumetric brain changes in thalamus, corpus callosum and
medial temporal structures: mild Alzheimer’s disease compared with
amnestic mild cognitive impairment. Dement Geriatr Cogn Disord
2012; 34: 149–55.
Price JL, Morris JC. Tangles and plaques in nondemented aging and
‘preclinical’ Alzheimer’s disease. Ann Neurol 1999; 45: 358–68.
Turkheimer FE, Kinnunen KM, et al. Inflammation after trauma: micro-
glial activation and traumatic brain injury. Ann Neurol 2011; 70:
Reiman EM, Langbaum JB, Tariot PN. Alzheimer’s prevention initiative: a
proposal to evaluate presymptomatic treatments as quickly as possible.
Biomark Med 2010; 4: 3–14.
Reiman EM, Quiroz YT, Fleisher AS, Chen K, Velez-Pardo C, Jimenez-
Del-Rio M, et al. Brain imaging and fluid biomarker analysis in young
adults at genetic risk for autosomal dominant Alzheimer’s disease in
the presenilin 1 E280A kindred: a case-control study. Lancet Neurol
2012; 11: 1048–56.
Ridgway G, Barnes J, Pepple T, Fox N. Estimation of total intracranial
volume; a comparison of methods [abstract]. Alzheimer’s Demen
2011; 7: S62.
Ridha BH, Barnes J, Bartlett JW, Godbolt A, Pepple T, Rossor MN, et al.
Tracking atrophy progression in familial Alzheimer’s disease: a serial
MRI study. Lancet Neurol 2006; 5: 828–34.
Ringman JM, O’Neill J, Geschwind D, Medina L, Apostolova LG,
Rodriguez Y, et al. Diffusion tensor imaging in preclinical and pre-
symptomatic carriers of familial Alzheimer’s disease mutations. Brain
2007; 130: 1767–76.
Rodriguez JJ, Witton J, Olabarria M, Noristani HN, Verkhratsky A.
Increase in the density of resting microglia precedes neuritic plaque
Alzheimer’s disease. Cell Death Dis 2010; 1: e1.
Ryan NS, Rossor MN. Correlating familial Alzheimer’s disease gene mu-
tations with clinical phenotype. Biomark Med 2010; 4: 99–112.
Schmahmann JD. Vascular syndromes of the thalamus. Stroke 2003; 34:
Thalamus and caudate in familial ADBrain 2013: Page 15 of 16 |
at University College London on April 16, 2013
Smith CD, Chebrolu H, Andersen AH, Powell DA, Lovell MA, Xiong S, Download full-text
et al. White matter diffusion alterations in normal women at risk of
Alzheimer’s disease. Neurobiol Aging 2010; 31: 1122–31.
Smith SM, Jenkinson M, Johansen-Berg H, Rueckert D, Nichols TE,
Mackay CE, et al. Tract-based spatial statistics: voxelwise analysis of
multi-subject diffusion data. Neuroimage 2006; 31: 1487–505.
Smith SM, Nichols TE. Threshold-free cluster enhancement: addressing
problems of smoothing, threshold dependence and localisation in clus-
ter inference. Neuroimage 2009; 44: 83–98.
Song SK, Sun SW, Ju WK, Lin SJ, Cross AH, Neufeld AH. Diffusion tensor
imaging detects and differentiates axon and myelin degeneration in
mouse optic nerve after retinal ischemia. Neuroimage 2003; 20:
Spain A, Daumas S, Lifshitz J, Rhodes J, Andrews PJ, Horsburgh K, et al.
Mild fluid percussion injury in mice produces evolving selective axonal
pathology and cognitive deficits relevant to human brain injury. J
Neurotrauma 2010; 27: 1429–38.
Stebbins GT, Murphy CM. Diffusion tensor imaging in Alzheimer’s
disease and mild cognitive impairment. Behav Neurol 2009; 21: 39–49.
Sun SW, Song SK, Harms MP, Lin SJ, Holtzman DM, Merchant KM,
et al. Detection of age-dependent brain injury in a mouse model of
brain amyloidosis associated
magnetic resonance diffusion tensor imaging. Exp Neurol 2005; 191:
Tovar-Moll F, Evangelou IE, Chiu AW, Richert ND, Ostuni JL,
Ohayon JM, et al. Thalamic involvement and its impact on clinical
disability in patients with multiple sclerosis: a diffusion tensor imaging
study at 3T. AJNR Am J Neuroradiol 2009; 30: 1380–6.
Van der Werf YD, Jolles J, Witter MP, Uylings HB. Contributions of
thalamic nuclei to declarative memory functioning. Cortex 2003; 39:
Wang Y, Gupta A, Liu Z, Zhang H, Escolar ML, Gilmore JH, et al. DTI
registration in atlas based fiber analysis of infantile Krabbe disease.
Neuroimage 2011; 55: 1577–86.
Warfield SK, Zou KH, Wells WM. Simultaneous truth and performance
level estimation (STAPLE): an algorithm for the validation of image
segmentation. IEEE Trans Med Imaging 2004; 23: 903–21.
Warren JD, Rohrer JD, Hardy J. Disintegrating brain networks: from syn-
dromes to molecular nexopathies. Neuron 2012; 73: 1060–2.
Neuropsychological profiles of familial Alzheimer’s disease associated
with mutations in the presenilin 1 and amyloid precursor protein
genes. J Neurol 2001; 248: 45–50.
Weiskopf N, Lutti A, Helms G, Novak M, Ashburner J, Hutton C. Unified
segmentation based correction of R1 brain maps for RF transmit field
inhomogeneities (UNICORT). Neuroimage 2011; 54: 2116–24.
Wiegell MR, Tuch DS, Larsson HBW, Wedeen VJ. Angular differentiation
of thalamic nuclei by quantitative DTI. Proc Int Soc Mag Reson Med
2000; 8: 481.
Wilson S, Raghupathi R, Saatman KE, MacKinnon MA, McIntosh TK,
Graham DI. Continued in situ DNA fragmentation of microglia/macro-
phages in white matter weeks and months after traumatic brain injury.
J Neurotrauma 2004; 21: 239–50.
Xuereb JH, Perry RH, Candy JM, Perry EK, Marshall E, Bonham JR. Nerve
cell loss in the thalamus in Alzheimer’s disease and Parkinson’ disease.
Brain 1991; 114 (Pt 3): 1363–79.
Zhang H, Yushkevich PA, Alexander DC, Gee JC. Deformable registration
of diffusion tensor MR images with explicit orientation optimization.
Med Image Anal 2006; 10: 764–85.
Zhang H, Schneider T, Wheeler-Kingshott CA, Alexander DC. NODDI:
practical in vivo neurite orientation dispersion and density imaging of
the human brain. Neuroimage 2012; 61: 1000–16.
Brain 2013: Page 16 of 16 N. S. Ryan et al.
at University College London on April 16, 2013