Epilepsia, 52(9):1715–1724, 2011
doi: 10.1111/j.1528-1167.2011.03117.x
FULL-LENGTH ORIGINAL RESEARCH
Microstructural and volumetric abnormalities of the putamen
in juvenile myoclonic epilepsy
*Simon S. Keller, *Tobias Ahrens, ySiawoosh Mohammadi, *Gabriel Möddel, zHarald Kugel,
*E. Bernd Ringelstein, and *Michael Deppe
*Department of Neurology, University of Münster, Münster, Germany; yWellcome Trust Centre for Neuroimaging, UCL Institute of
Neurology, University College London, London, United Kingdom; and zDepartment of Radiology, University of Münster, Münster,
Germany
SUMMARY
Purpose: Patients with juvenile myoclonic epilepsy (JME)
show evidence of microstructural white matter (WM)
damage of thalamocortical fiber tracts and changes of
blood oxygen level dependent (BOLD) signal in a striato-
thalamocortical network. The objective of the present
study was to investigate microstructural and volumetric
alterations of the putamen in patients with JME using dif-
fusion tensor imaging (DTI) and conventional magnetic
resonance imaging (MRI).
Methods: We performed DTI and MRI for 10 patients
with JME and 59 age-matched neurologically healthy vol-
unteers. Evaluation of microstructural damage was inves-
tigated using calculation of mean fractional anisotropy
(FA) values in a priori regions of interest (ROIs) for the
putamen, frontal lobe, and a thalamocortical region, after
application of an improved eddy current correction
method and a new statistical parametric mapping (SPM)–
compatible toolbox incorporating intensive multicontrast
FA image registration. Stereologic analysis on MRI was
performed to estimate macroscopic volume of the puta-
men in both cerebral hemispheres for all subjects.
Juvenile myoclonic epilepsy (JME) is an electroclinical
syndrome (Berg et al., 2010) with a strong genetic basis
(Medina et al., 2008; Zifkin et al., 2005), typically begin-
ning in puberty (Genton & Gelisse, 2001; Montalenti et al.,
2001). Patients with JME frequently have neurocognitive
deficits consistent with frontal lobe dysfunction (Devinsky
et al., 1997; Piazzini et al., 2008; Pulsipher et al., 2009),
which is consistent with the frequent presentation of bifron-
tal accentuated epileptiform discharges (Holmes et al.,
2010; Santiago-Rodriguez et al., 2002). Neuroanatomic
Accepted April 7, 2011; Early View publication June 2, 2011.
Address correspondence to Michael Deppe, The Department of Neurol-
ogy, University of Mnster, Albert-Schweitzer-Street 33, D-48129 Mnster,
Germany. E-mail: deppe@uni-muenster.de
Simon S. Keller and Tobias Ahrens contributed equally.
Wiley Periodicals, Inc.
ª 2011 International League Against Epilepsy
1715
Key Findings: Relative to controls, patients had signifi-
cantly reduced FA in the frontal lobe (p = 0.01) and
thalamocortical fiber WM (p < 0.001). In contrast,
putamen FA was bilaterally increased (p = 0.01) and
correlated with decreasing putamen volume (r2 = )0.63,
p = 0.004) in patients only. Putamen FA correlated nega-
tively with onset of JME (total: r2 = )0.50, p = 0.01),
duration of JME (r2 = 0.52, p = 0.01), and thalamocortical
fiber FA (r2 = )0.47, p = 0.01).
Significance: This is the first evidence of combined
microstructural and macrostructural putamen abnor-
malities in patients with JME, with early age of onset and
a longer duration of epilepsy being significant predictors
for greater architectural alterations. These findings are
consistent with studies indicating neurophysiologic
abnormalities of frontostriatal networks in patients with
JME, and may contribute to explain the frequent presen-
tation of executive dysfunction in these patients. Confir-
mation and further exploration of the increase in
putamen FA in patients with JME is required in larger
samples.
KEY WORDS: Basal ganglia, Diffusion tensor imaging,
Epilepsy, Fractional anisotropy, White matter.
abnormalities in patients with JME have been manifested as
volume atrophy of the thalamus (Kim et al., 2007; Pulsipher
et al., 2009), microstructural white matter (WM) damage of
a thalamocortical network (Deppe et al., 2008), and
alterations of gray matter (GM) concentration in the frontal
lobe (Woermann et al., 1999a,b; Tae et al., 2006; Kim
et al., 2007). Furthermore, a recent 18F-Fallypride ([18F]FP)
positron emission tomography (PET) study demonstrated
reduced dopamine binding on the D2/D3 receptor of the
posterior putamen in patients with JME (Landvogt et al.,
2010), although another study reported no alterations in
[11C]PE2I putamen dopamine receptors but did in the sub-
stantia nigra and frontal lobe (Ciumas et al., 2008). There
is, therefore, convergent evidence implicating involvement
of the putamen in JME given the above reported neurophys-
iologic abnormalities of the putamen (Landvogt et al.,
2010), the substantial connectivity (Leh et al., 2007)
1716
S. S. Keller et al.
between the putamen and the previously reported abnormal-
ities of the thalamus and prefrontal cortex (Woermann
et al., 1999b; Tae et al., 2006; Kim et al., 2007; Deppe
et al., 2008; Pulsipher et al., 2009), and the neurocognitive
deficits associated with frontostriatal networks observed in
JME (Devinsky et al., 1997; Piazzini et al., 2008; Pulsipher
et al., 2009). The primary goal of the present study was to
investigate microstructural and volumetric alterations of the
putamen in patients with JME using magnetic resonance
(MR) diffusion tensor imaging (DTI) and conventional MR
imaging.
DTI is a sensitive in vivo MR imaging method for the
investigation of microstructural cerebral lesions that cannot
be detected using conventional MR imaging (Deppe et al.,
2007, 2008; Simonyan et al., 2008; Gattellaro et al., 2009).
DTI permits quantitative analysis of WM and GM fractional
anisotropy (FA), which provides information on the integ-
rity of brain tissue and disruption to normal axonal organi-
zation in neurologic disorders. Given that DTI-determined
alterations of subcortical nuclei are notoriously difficult to
quantify due to noise inherent in DTI data acquisition and
problems associated with image registration, we used a
recently developed eddy current correction approach and
statistical parametric mapping (SPM)–compatible DTI tool-
box that incorporates intensive multicontrast image registra-
tion to prospectively determine subcortical FA alterations in
patients with JME. We sought to investigate architectural
and morphologic alterations of the putamen in patients with
JME using complementary DTI and conventional MR imag-
ing approaches. In particular, we performed quantitative
analyses of mean putamen FA using DTI and of putamen
volume using stereologic analysis on MR images to investi-
gate the nature of putamen abnormalities in JME, whether
microstructural alterations in FA co-occurred with macro-
scopic changes, and whether any changes were related to
clinical factors including the duration and age of onset of
JME.
Methods
Participants
Ten patients (six female and four male, mean age
28 € 8.7 SD, range 18–41 years) with JME were studied
(Deppe et al., 2008). All available medical records were
examined for each patient, including information on seizure
semiology, seizure frequency, and neuropsychological
performance. Diagnosis of JME relied on the clinical and
seizure semiology information consistent with International
League Against Epilepsy (ILAE) guidelines (ILAE, 1989).
Mean onset of habitual seizures was 13.7 € 2.7 years. All
patients showed bilateral myoclonic seizures, seven patients
had at least one generalized tonic–clonic seizure (GTCS),
and five patients had typical dialeptic seizures. Patients had
at least one electroencephalography (EEG) with a typical
generalized polyspike wave complex. At the time of scan-
Epilepsia, 52(9):1715–1724, 2011
doi: 10.1111/j.1528-1167.2011.03117.x
ning all patients were under anticonvulsant therapy with a
mean duration of 12.3 € 11.01 years (range 0.2–31 years);
four patients were completely seizure free for >2 years and
all but one patient were free of GTCS for almost 3 years.
The last myoclonic or absence seizure was at least 1 week
prior to assessment (mean 23.1 € 27.5 months). All patients
had normal conventional magnetic resonance imaging
[MRI; T1-, T2-weighted, and fluid-attenuated inversion
recovery (FLAIR)] and showed no signs of abnormality on
neurologic examination. Patient data are provided in
Table 1. We additionally studied a neurologically and
psychiatrically healthy control group that was composed of
59 adult volunteers [30 women and 29 men, mean age
27.5 € 4.0 standard deviation (SD), range 21–43], all of
whom had normal neurologic examination and normal MRI
(T1-, T2-weighted, and FLAIR). There was no statistical dif-
ference in age between patients and controls (t = )0.64,
p = 0.53). All subjects gave written informed consent and
the local ethics committee approved this study.
DTI and MRI acquisition
All participants underwent DTI and high-resolution
MRI (T1-weighted, T2-weighted, and FLAIR imaging at
3T; Philips Intera T30; Philips, Best, The Netherlands). All
MRI modalities were used to exclude the possibility of brain
lesions in patients and controls. Only DTI and T1-weighted
MRI acquisitions were used for quantitative analyses that
are reported herein. For DTI we used single shot echo planar
imaging (EPI) with 20 diffusion directions without cardiac
gating [two b-factors, 0 and 1,000 s/mm2, TR = 9.8 s/
TE = 95 ms, acquisition matrix: 128 · 128, voxel size:
1.8 · 1.8 · 3.6 mm3 (reconstructed to 2.0 · 2.0 · 2.0 mm3
after image processing), two averages, scanning time
7:46 min]. For T1-weighted structural MRI we used a high-
resolution three-dimensional (3D) turbo-field-echo sequence
(matrix 256 · 256 · 160 over a field of view of 25.6 ·
25.6 · 16 cm3 reconstructed after zero filling to 512 ·
512 · 320 cubic voxels with an edge length of 0.5 mm).
DTI image preprocessing and analysis
The analysis of DTI data in patients with neurologic
syndromes can be complicated by many factors, including
geometric distortions during long-image acquisition periods
and image registration insufficiencies. In an attempt to over-
come these problems, we have developed a new improved
eddy current correction technique (Mohammadi et al.,
2010) and an SPM-compatible toolbox that incorporates a
multicontrast image registration algorithm for the optimum
spatial preprocessing of DTI data prior to statistical analysis
(Glauche et al., 2010; Mohammadi et al., 2009). These
improvements facilitate detection of focal FA alterations in
brain regions notoriously difficult to quantify using DTI,
such as within small subcortical nuclei. The development of
this toolbox was part of a wider project, the focus of which
is to develop and sensitize quantitative DTI techniques for
 1717
DTI of JME

Table 1. JME patient data

Current Last Last Duration
Age at Duration Seizure anticonvulsant GTCS seizure Number of ACT
Age epilepsy of disease Epileptogenic types medication All ACTever (months (months of GTCS therapy
Patient (year) Gender onset (years) zone (EEG) ever had (per day) used ago) ago) ever had (years)
1 18 M 17 1 Gen Mcl VPA 1,000 mg VPA n.a. 2 0 0.2
2 33 M 12 21 Gen Mcl, GTCS VPA 2,500 mg VPA 10 10 4 6.1
3 20 M 17 3 Gen Mcl, abs VPA 1,000 mg VPA n.a. 15 0 0.2
4 19 F 14 5 Gen Mcl, abs LTG 200 mg VPA, LTG n.a. 4 0 3.7
5 22 F 16 5 Gen Mcl, GTCS VPA 1,250 mg VPA, CBZ 63 63 2 4.3
6 25 M 12 15 Gen Mcl, GTCS VPA 900 mg VPA 60 60 4 10.3
7 40 F 14 26 Gen Mcl, GTCS LTG 500 mg LTG, PHT, 35 29 9 21.9
VPA, GBP
8 27 F 15 12 Gen Mcl, GTCS, LTG 600 mg, LTG, ETH 60 60 15 10.5
abs ETH 250 mg
9 41 F 10 31 Gen Mcl, GTCS, LTG 450 mg LTG, CBZ, 38 1 12 2.6
abs PRD, PHB
10 35 F 12 23 Gen Mcl, GTCS, LTG 200 mg, LTG, VPA, 77 3 10 1.9
abs TOP 600 mg TOP, ETH,
PHB
F, female; m, male; rt, right; lft, left; hem, hemispheric; gen, generalized; temp, temporal; bilat, bilateral; nonloc, nonlocalizing; mcl, myoclonic seizure; abs,
absence seizure; automot, automotor seizure; sm mot, simple motor seizure; GTCS, generalized tonic–clonic seizure; VPA, valproate; LTG, lamotrigine; TOP,
topiramate; CBZ, carbamazepine; OXC, oxcarbazepine; LEV, levetiracetam; GBP, gabapentin; PHT, phenytoin; ETH, ethosuximide; PHB, phenobarbital; VGB,
vigabatrin; PRD, primidone; ACT, anticonvulsant therapy.

the identification of microscopic brain alterations in smaller
samples of patients and individual patients with neurologic
disorders (Deppe et al., 2007, 2008). All DTI image pro-
cessing using the toolbox was performed with the ‘‘Mnster
Neuroimaging Evaluation System (EVAL).’’ All time con-
suming calculations, for example, eddy currents correction
and normalizations, were carried out on a 64-bit 64-proces-
sor parallel computer (Sun Microsystems, Inc., Palo Alto,
CA, U.S.A.). The registration toolbox provides iterative
multicontrast registrations steps based on FA and b0 con-
trasts, using two customized templates, one FA and one b0
contrast template. Prior to the iterative registration, the EPI
images measured without diffusion gradient (b0 images)
were registered to the SPM EPI template using affine trans-
formations. The multicontrast registration was then itera-
tively applied to obtain normalized FA images. Registered
FA images correspond to Montreal Neurological Institute
(MNI) coordinate space. The combination of the eddy cur-
rent correction and iterative multicontrast registration
enables the detection of very focal alterations in water diffu-
sion parameters.
For analysis of the optimally spatially registered FA
images, we generated regions of interest (ROIs) to obtain
mean FA values that were defined a priori using ROI masks
on the output images from the registration toolbox for all
patients and controls. Based on previous work reporting
morphologic and physiologic abnormalities in patients with
JME, we defined ROIs including the putamen (Fig. 1A) and
frontal lobe WM (Fig. 1C). We also used an ROI for the
WM of thalamocortical tracts adjacent to the thalamus
(referred to here as thalamocortical fiber FA) that was
previously reported to be abnormal in JME patients using
voxel-based statistics by Deppe et al. (2008) (Fig. 1B).
MRI analysis
We obtained volume of global WM and GM, relative
brain size, and cerebrospinal fluid (CSF) automatically
from 3D T1-weighted images of all patients and controls
by using the SIENAX protocol (Smith et al., 2002) inte-
grated into FSL software (http://www.fmrib.ox.ac.uk/fsl/
siena/index.html). Individual brain tissue volumes were
normalized for subject head size by a volumetric scaling
factor (VSCALE), derived by affine registration of the
skull image to the MNI152 space.
The Cavalieri method of design-based stereology in con-
junction with point counting (Roberts et al., 2000) was used
to estimate the volume of the left and right putamen from the
T1-weighted images of all subjects. By using the Cavalieri
method, volume is directly estimated from equidistant and
parallel MR images of the brain starting with a uniform
random starting position. A second level of sampling is
required to estimate the section area from each image by
applying point counting within the ROI. The mathematical
justification and implementation of the methodology is
simple and it can be applied to structures of arbitrary shape
(Garcia-Finana et al., 2009). The putamen ROI followed
those described by Pulsipher et al. (2007) based on
guidelines freely available on the Internet (http://www.
psychiatry.uiowa.edu/mhcrc/pdf/papers/putamen.pdf). The
anterior limb of the internal capsule separated the anterior
portion of the putamen from the caudate nucleus, and tissue
connecting the caudate and putamen on the most rostral

Epilepsia, 52(9):1715–1724, 2011

doi: 10.1111/j.1528-1167.2011.03117.x
1718
S. S. Keller et al.
B
C
sections were omitted from measurements and classified as
caudate (Pulsipher et al., 2007). The medial and lateral bor-
ders of the putamen were the internal capsule and external
capsule, respectively. MRICRO (http://www.cabiatl.com/
mricro/mricro/mricro.html) was used initially for image
reorientation into the coronal plane, smoothing to improve
GM–WM demarcation, and brightness and contrast optimi-
zation. Images were subsequently imported into EASY-
MEASURE software (Keller et al., 2007) for stereologic
point counting. Randomly orientated test probes were
removed with a mouse click within the putamen ROI sepa-
rately for the left and right hemispheres of each participant
(Fig. 1D). Stereologic parameters were optimized to
achieve a coefficient of error of less than 5% (Roberts et al.,
2000). Separation between test points on the square grid
used for point counting was 0.5 cm (i.e., 5 pixels) and slice
interval was 2.5 mm (every fifth MR section). Unbiased
estimates of the putamen transect area were obtained by
multiplying the total number of points recorded by the area
corresponding to each test point. An unbiased estimate of
putamen volume was obtained as the sum of the estimated
areas of the structure transects on consecutive systematic
sections multiplied by the distance between sections.
Epilepsia, 52(9):1715–1724, 2011
doi: 10.1111/j.1528-1167.2011.03117.x
Figure 1.
(A) Putamen FA ROI. (B) Thalamo-
cortical white matter FA ROI. (C)
Frontal lobe FA ROI. (D) Putamen
volume estimated using point
counting on a randomly selected
T1-weighted coronal MRI section.
Epilepsia ILAE
Approximately 250 points were recorded on approximately
12–15 systematic random sections, consistent with stereo-
logic volume estimation of other brain structures (Keller &
Roberts, 2009). An interrater study was undertaken in 10
subjects and revealed a high level of reproducibility (intra-
class correlation = 0.97).
Statistical analysis
Given the possible influence of nonnormally distributed
data in our sample, we used a Kruskal-Wallis ANOVA
including correction for multiple comparisons (Siegel &
Castellan, 1988) to test for intergroup differences in
FA and volume. The analysis of relationships between
volumetric and DTI variables was performed by using
Pearson’s correlation coefficients. For the percentage
change of volume and FA between controls and patients
we used the formula (b)a)/a · 100, where a = control
and b = patient value. Right–left putamen FA asymmetry
was calculated using the formula (c)d)/(c+d), where
c = FA putamen right and d = FA putamen left. All statis-
tical calculations were performed using STATISTICA
version 9.1 (2010; Stat Soft. Inc, Tulsa, OK, U.S.A.;
http://www.statsoft.com).
 1719
DTI of JME

Table 2. ROI values and SIENAX data

Mean CTRL Mean JME SD CTRL SD JME H (1, N = 69) p-value Change in %
FA putamen total* 133.65 148.30 19.39 27.59 5.95 0.014 10.96
FA putamen left* 133.06 151.02 20.90 27.48 9.05 0.002 13.50
FA putamen right* 134.24 145.57 20.30 29.76 4.54 0.031 8.44
Vol putamen total (cm3)* 9.31 7.91 0.99 1.16 10.82 <0.001 15.02
Vol putamen left (cm3)* 4.59 3.95 0.53 0.58 11.38 <0.001 13.78
Vol putamen right (cm3)* 4.72 3.95 0.48 0.59 10.59 <0.001 16.23
FA thalamo-cortical* 459.52 421.82 27.11 38.80 11.58 <0.001 8.20
FA frontal lobe* 334.24 320.63 17.94 15.51 5.76 0.012 4.07
WM absolute vol (cm3) 486.57 485.12 50.59 37.11 0.22 0.928 0.30
GM absolute vol (cm3)* 709.13 659.52 55.13 53.19 5.65 0.013 7.00
WM normalized vol (cm3) 623.73 626.48 45.13 26.88 0.48 0.812 0.43
GM normalized vol (cm3)* 911.06 852.01 62.77 47.81 6.44 0.006 6.48
VSCALE 1.29 1.29 0.15 0.11 0.04 0.964 0.00
CSF vol (cm3) 270.91 269.83 49.24 36.50 0.07 0.947 0.40
Rel. brain size 0.82 0.81 0.02 0.02 0.86 0.393 0.79
Statistics for ROI FA values (mean FA value · 1,000).
*Significantly changed in patients relative to healthy controls (p < 0.05).

A B

Figure 2.
Differences in left (blue) and right (red) putamen volume (A) and mean FA (B) between controls (ctrl) and patients (JME).
Epilepsia ILAE

patients (r2 = )0.47, p = 0.01) (Fig. 3A), but no relation-

Results
DTI
All descriptive and inferential statistics for FA differ-
ences between patients and controls are provided in
Table 2. We observed significantly (p = 0.01) decreased
FA of frontal lobe WM in patients relative to controls,
which represented a 4.07% reduction. There was a highly
significant (p < 0.001) decrease of FA in the thalamocorti-
cal fiber ROI in patients relative to controls, representing a
percentage change of 8.2%. Putamen FA was significantly
increased in patients relative to controls. This increase was
10.96% (p = 0.01) bilaterally, 13.50% (p = 0.002) for the
left putamen, and 8.44% (p = 0.03) for the right putamen
(Fig. 2).
There was a negative correlation between the total
putamen FA and FA of the thalamocortical fiber ROI in
ship between increased total putamen FA and decreased FA
of the frontal lobe (r2 = )0.11, p = 0.75). Only right
putamen FA (r2 = )0.6, p = 0.008) and not left putamen
FA (r2 = )0.3, p = 0.06) was significantly negatively
correlated with the thalamocortical fiber ROI in patients.
Significant correlations were not observed between puta-
men FA and thalamocortical fiber ROI FA (r2 = )0.017,
p = 0.89) or between total putamen FA and frontal lobe FA
(r2 = )0.011, p = 0.56) in controls.
Volume
All descriptive and inferential statistics for volume dif-
ferences between patients and controls are provided in
Table 2. Putamen volume was significantly decreased
in patients relative to controls bilaterally (p < 0.001), and
in the left (p < 0.001) and right (p < 0.001) hemispheres

Epilepsia, 52(9):1715–1724, 2011

doi: 10.1111/j.1528-1167.2011.03117.x
1720
S. S. Keller et al.
A B
Figure 3.
Significant relationships observed in the present study. (A) Bilateral (Bi.) thalamocortical (Th-C) WM FA and putamen FA are nega-
tively correlated in patients. (B) Bilateral putamen FA and putamen volume are negatively correlated in patients (red circles) but not in
controls (blue circles). (C) Bilateral putamen FA is negatively correlated with age of onset of JME (and duration of JME, not shown).
The horizontal line indicates the mean putamen FA in controls.
Epilepsia ILAE
separately (Fig. 2). Normalized (p = 0.006) and absolute
(p = 0.01) GM volume, but not normalized (p = 0.81) or
absolute (p = 0.93) WM volume, was significantly
decreased in patients relative to controls. There were no
differences in CSF volume (p = 0.95), relative brain size
(p = 0.39), and VSCALE (p = 0.96) between patients and
controls, confirming that no systematic biologic difference
between patients and controls had affected the results.
DTI-volume correlations
When all patients and controls were considered together,
no correlation was observed between putamen volume
bilaterally and total putamen FA (r2 = )0.08), left putamen
volume/FA (r2 = )0.07), or right putamen volume/FA
(r2 = )0.06). However, when patients were analyzed sepa-
rately, there was a strong correlation between putamen FA
and volume bilaterally (r2 = )0.63, p = 0.004), and in the
left (r2 = )0.62, p = 0.004) and right (r2 = )0.54,
p = 0.009) hemispheres separately. Significant correlations
did not exist in controls. These relationships are shown in
Fig. 3B. There were no significant correlations between
total putamen volume and frontal lobe FA (patients:
r2 = 0.004, p = 0.86; controls: r2 = 0.001, p = 0.79) or
thalamocortical fiber ROI FA (patients: r2 = 0.4, p = 0.03;
controls: r2 = 0.002, p = 0.35).
Correlations with clinical parameters
Putamen FA was positively correlated with duration of
JME (total: r2 = 0.52, p = 0.011; right: r2 = 0.45, p = 0.02;
left: r2 = 0.50, p = 0.01) and negatively correlated with
onset of JME (total: r2 = )0.50, p = 0.01; right: r2 = )0.40,
p = 0.03; left: r2 = )0.52, p = 0.01) (Fig. 3C). Intuitively,
onset and duration were significantly correlated (r2 = 0.69,
p = 0.002). Putamen volume showed a borderline correla-
tion with duration of JME (total: r2 = )0.33, p = 0.05; right:
r2 = )0.30, p = 0.06; left: r2 = )0.50, p = 0.01), but not
Epilepsia, 52(9):1715–1724, 2011
doi: 10.1111/j.1528-1167.2011.03117.x
C
onset of JME (r2 = 0.18, p = 0.12; right: r2 = 0.19,
p = 0.12; left: r2 = 0.16, p = 0.13).
Discussion
The main finding of this study was the significant
inverse relationship between FA and volume alterations of
the putamen of patients with JME relative to a neurologi-
cally healthy control group. This co-occurrence of morpho-
logic atrophy with increased anisotropy of the putamen in
JME is a new finding. Reduced volume of the putamen has
been documented in previous studies of GM atrophy in
partial epilepsy (Dreifuss et al., 2001; Keller & Roberts,
2008), but our demonstration of increased FA of these
nuclei is a new finding in patients with JME and requires
careful consideration. The observed inverse relationship
between FA and volume alterations of the putamen are
counterintuitive given that the pathologic processes caus-
ing macroscopic atrophy would have been thought to also
cause a loss of anisotropy of brain tissue. One explanation
may be that the atrophic GM within the putamen
excessively constrains the space where numerous myelin-
ated axons intersperse. Myelinated axons have a much
greater FA compared to GM, and the spatial constriction
of myelinated axons may lead to a localized increased
anisotropy of water diffusion, given that this process would
occur beyond the voxel resolution of DTI. This hypotheti-
cal situation is illustrated in Fig. 4. However, such an
explanation would need to be reconciled with the possibil-
ity of volume atrophy and FA reduction of the thalamus in
patients with JME, a structure also consisting of myelin-
ated fibers. Thalamic volume reduction using conventional
MRI has previously been reported in patients with JME
(Kim et al., 2007; Pulsipher et al., 2009), although to our
knowledge there has been no report of FA alterations of
thalamic nuclei in patients with JME to date [PubMed

A B
Figure 4.
Potential microstructural changes that may affect average water diffusion behavior in the putamen (FA increase). (A) microscopic
image of a subvoxel-sized region within the putamen. (B) The putamen can roughly be divided into isotropic gray matter (GM) with
low FA and regions of myelinated axons (MA) with high FA. (C) The loss of isotropic gray matter leads to a relative increase of aniso-
tropic tissue (MA), which is potentially much less affected. The average tissue anisotropy increases and thus the mean FA estimated
within a voxel.
Epilepsia ILAE
search of ‘‘DTI’’ and ‘‘juvenile myoclonic epilepsy’’ yields
three studies in January 2011: Deppe et al. (2008),
O’Muircheartaigh et al. (2011), Vulliemoz et al. (2010)].
We are currently investigating the relationship between
measures of anisotropy, diffusivity, and volume of the thal-
amus in patients with JME.
An alternative explanation relates to possible alterations
of iron concentration within the putamen in patients with
epilepsy. Given that there may be excessive iron concentra-
tions in patients with epilepsy (Ikeda, 2001) and that normal
age-related alterations of DTI-based putamen measures
have been associated with local cerebral iron concentration
(Pfefferbaum et al., 2010), it may be plausible to associate
the increase of putamen FA with an increase in local iron
concentration in our patients. Further support for this
hypothesis comes from studies of neurodegenerative dis-
orders. For example, patients with Huntington’s disease
have been shown to have excessive concentrations of iron in
the putamen (Dexter et al., 1991, 1992), and show evidence
of atrophy (Ruocco et al., 2006) and increased FA (Douaud
et al., 2009; Kloppel et al., 2008) of the putamen. Con-
versely, patients with Parkinson’s disease have been shown
to have no increased iron concentration in the putamen
(Dexter et al., 1991, 1992; Griffiths et al., 1999; Graham
et al., 2000)—and even reduced iron concentration in the
putamen (Graham et al., 2000)—and to show evidence of
atrophy (Geng et al., 2006) but no increased FA (Sitburana
& Ondo, 2009) of the putamen. Furthermore, neurologic
disorders primarily characterized by excessive iron
accumulation in basal ganglia nuclei but not the putamen,
such as Hallervorden-Spatz disease, have been shown to be
associated with FA increases in the substantia nigra and glo-
bus pallidus, and not the putamen (Awasthi et al., 2010).
Increased iron concentration may cause death of c-aminobu-
tyric acid (GABA)ergic neurons (Zhang et al., 1989) and
dysfunction of dopaminergic neurons (Wesemann et al.,
1721
DTI of JME
C
1995), and may explain the altered putamen dopaminergic
neurotransmission in patients with epilepsy (Bouilleret
et al., 2005, 2008; Landvogt et al., 2010). Although
speculative, the suggestion of increased putamen iron
concentrations in patients with epilepsy causing increased
putamen FA is a plausible theory for a DTI finding that has
previously remained unexplained in some neurologic dis-
orders (Kloppel et al., 2008; Douaud et al., 2009).
It was important to confirm that the increased putamen
FA in patients was not due to a systematic error in image reg-
istration. In particular, it was conceivable that WM adjacent
to the atrophic putamen in patients could be incorporated
into the putamen ROI and increase the mean FA of the ROI
given the higher FA values of WM. However, we assessed
the projection of the putamen ROI on the corrected and reg-
istered FA images of all patients, and in no patient did the
ROI include adjacent WM. The putamen ROI mask did not
encompass the entire nucleus, thereby allowing for atrophy
and imperfect registration. Figure 5 compares a randomly
selected control and patient FA image, illustrating the
increased putamen FA typically seen in our patients with
JME relative to controls. Given that the inverse correlation
between putamen FA and volume needs to be confirmed in
larger samples of patients with JME, and that it is not
yet known whether such an inverse correlation exists in the
thalamus, the findings presented herein with respect to
FA-volume correlations should be considered as prelimin-
ary. However, strong support for the putamen FA increase
being biologically relevant comes from the significant cor-
relations observed between putamen FA and the clinical
variables of the patients (see below).
We also reported that increasing putamen FA correlates
with decreasing thalamocortical WM FA. This is also a
new finding and may be due to the involvement of both
structures and connecting pathways in an executive
network (Haber, 2003; Haznedar et al., 2005). The
Epilepsia, 52(9):1715–1724, 2011
doi: 10.1111/j.1528-1167.2011.03117.x
1722
S. S. Keller et al.
observed reduced FA of thalamocortical WM is consistent
with previous reports of thalamic volume loss in JME
(Kim et al., 2007; Pulsipher et al., 2009). Macroscopic
atrophy and microscopic anisotropy alterations might also
reflect metabolic alterations of putamen neurons as
described by a [18F]FP PET study, which reported reduced
dopamine binding on the D2/D3 receptor of the posterior
putamen in patients with JME (Landvogt et al., 2010).
These metabolic and structural alterations of the putamen
in JME are consistent with animal models of epilepsy that
have shown that modulation of striatal dopaminergic and
GABAergic circuits that have a modulating effect on
absence and generalized seizures (Deransart et al., 1998;
Hikosaka, 1998; Ono et al., 1987).
Putamen FA in particular showed a significant correlation
with the duration and age of onset of JME. The putamen,
and more generally the lenticular nuclei and frontal lobe,
undergo complex development during adolescence (Sowell
et al., 1999), the time at which JME is typically diagnosed
(Genton & Gelisse, 2001; Montalenti et al., 2001). It is,
therefore, likely that the pathophysiologic processes associ-
ated with the onset of JME adversely affect the normal
development of frontostriatal structures and connections,
and may explain the deficits in executive functions ordinar-
ily mediated by these brain regions (Devinsky et al., 1997;
Piazzini et al., 2008; Pulsipher et al., 2009). However,
there is also the possibility that chronic use of antiepileptic
medication may interact with brain development. A
previous animal study has shown that valproate, an
antiepileptic drug administered to 80% of our sample of
patients (see Table 1), can cause neurodegeneration in the
developing brain (Bittigau et al., 2003).
Epilepsia, 52(9):1715–1724, 2011
doi: 10.1111/j.1528-1167.2011.03117.x
Figure 5.
Illustration of increased FA within
the putamen ROI in a randomly
selected patient relative to a
randomly selected healthy control.
The top row indicates orthogonal
sagittal, axial, and coronal sections
of an FA image of the control
subject with the putamen ROI
(green) projected onto the image.
The comparison between the
control and patient images on the
bottom row shows the increased
FA (lighter blue) within the
putamen ROI (dotted circle). The
patient image also shows evidence
of more widespread WM FA
reduction.
Epilepsia ILAE
Conclusion
We have found that patients with JME show concomi-
tant and inverse microstructural and macrostructural
changes of the putamen in both cerebral hemispheres,
which correlate with the duration and age of onset of JME.
We have also provided evidence indicating frontal and tha-
lamocortical WM deterioration in JME, the latter of which
correlates significantly with the pathologic status of the
putamen. The observed microscopic and macroscopic
structural abnormalities observed in the present study are
consistent with results of previous neurophysiologic studies
indicating preferential abnormalities of frontostriatal
networks in JME, and may contribute to explain the
frequent presentation of executive dysfunctions in this
patient group. Further work is required to determine the
pathologic determinants of increased FA of the putamen in
patients with epilepsy.
Acknowledgments
This work was supported by the Transregional Collaborative Research
Centre SFB/TR 3 (Project A8) of the Deutsche Forschungsgemeinschaft
(DFG) and by the Neuromedical Foundation (Stiftung Neuromedizin),
Mnster. SM (and open access to this article) were supported by the Well-
come Trust. The authors thank Prof. Dr. Tanja Kuhlmann for providing the
microscopic image of the putamen.
Disclosure
None of the authors has any conflict of interest to disclose. We confirm
that we have read the Journal’s position on issues involved in ethical
publication and affirm that this report is consistent with those guidelines.
 1723
DTI of JME

Graham JM, Paley MN, Grunewald RA, Hoggard N, Griffiths PD. (2000)
References Brain iron deposition in Parkinson’s disease imaged using the PRIME
magnetic resonance sequence. Brain 123(Pt 12):2423–2431.
Awasthi R, Gupta RK, Trivedi R, Singh JK, Paliwal VK, Rathore RK.
Griffiths PD, Dobson BR, Jones GR, Clarke DT. (1999) Iron in the basal
(2010) Diffusion tensor MR imaging in children with pantothenate
ganglia in Parkinson’s disease. An in vitro study using extended X-ray
kinase-associated neurodegeneration with brain iron accumulation and
absorption fine structure and cryo-electron microscopy. Brain
their siblings. AJNR Am J Neuroradiol 31:442–447.
122(Pt 4):667–673.
Berg AT, Berkovic SF, Brodie MJ, Buchhalter J, Cross JH, van Emde
Haber SN. (2003) The primate basal ganglia: parallel and integrative net-
Boas W, Engel J, French J, Glauser TA, Mathern GW, Moshe SL,
works. J Chem Neuroanat 26:317–330.
Nordli D, Plouin P, Scheffer IE. (2010) Revised terminology and con-
Haznedar MM, Roversi F, Pallanti S, Baldini-Rossi N, Schnur DB, Li-
cepts for organization of seizures and epilepsies: report of the ILAE
calzi EM, Tang C, Hof PR, Hollander E, Buchsbaum MS. (2005)
Commission on Classification and Terminology, 2005–2009. Epilepsia
Fronto-thalamo-striatal gray and white matter volumes and anisotropy
51:676–685.
of their connections in bipolar spectrum illnesses. Biol Psychiatry
Bittigau P, Sifringer M, Ikonomidou C. (2003) Antiepileptic drugs and
57:733–742.
apoptosis in the developing brain. Ann N Y Acad Sci 993:103–114.
Hikosaka O. (1998) Neural systems for control of voluntary action – a
Discussion 123-104.
hypothesis. Adv Biophys 35:81–102.
Bouilleret V, Semah F, Biraben A, Taussig D, Chassoux F, Syrota A, Ribe-
Holmes MD, Quiring J, Tucker DM. (2010) Evidence that juvenile myo-
iro MJ. (2005) Involvement of the basal ganglia in refractory epilepsy:
clonic epilepsy is a disorder of frontotemporal corticothalamic net-
an 18F-fluoro-L-DOPA PET study using 2 methods of analysis. J Nucl
works. Neuroimage 49:80–93.
Med 46:540–547.
Ikeda M. (2001) Iron overload without the C282Y mutation in patients with
Bouilleret V, Semah F, Chassoux F, Mantzaridez M, Biraben A, Trebossen
epilepsy. J Neurol Neurosurg Psychiatry 70:551–553.
R, Ribeiro MJ. (2008) Basal ganglia involvement in temporal lobe epi-
ILAE. (1989) Proposal for revised classification of epilepsies and epileptic
lepsy: a functional and morphologic study. Neurology 70:177–184.
syndromes. Commission on Classification and Terminology of the
Ciumas C, Wahlin TB, Jucaite A, Lindstrom P, Halldin C, Savic I. (2008)
International League against Epilepsy. Epilepsia 30:389–399.
Reduced dopamine transporter binding in patients with juvenile myo-
Keller SS, Roberts N. (2008) Voxel-based morphometry of temporal lobe
clonic epilepsy. Neurology 71:788–794.
epilepsy: an introduction and review of the literature. Epilepsia
Deppe M, Duning T, Mohammadi S, Schwindt W, Kugel H, Knecht S,
49:741–757.
Ringelstein EB. (2007) Diffusion-tensor imaging at 3 T: detection of
Keller SS, Roberts N. (2009) Measurement of brain volume using MRI:
white matter alterations in neurological patients on the basis of normal
software, techniques, choices and prerequisites. J Anthropol Sci
values. Invest Radiol 42:338–345.
87:127–151.
Deppe M, Kellinghaus C, Duning T, Moddel G, Mohammadi S, Deppe K,
Keller SS, Highley JR, Garcia-Finana M, Sluming V, Rezaie R, Roberts N.
Schiffbauer H, Kugel H, Keller SS, Ringelstein EB, Knecht S. (2008)
(2007) Sulcal variability, stereological measurement and asymmetry of
Nerve fiber impairment of anterior thalamocortical circuitry in juvenile
Broca’s area on MR images. J Anat 211:534–555.
myoclonic epilepsy. Neurology 71:1981–1985.
Kim JH, Lee JK, Koh SB, Lee SA, Lee JM, Kim SI, Kang JK. (2007) Regio-
Deransart C, Vercueil L, Marescaux C, Depaulis A. (1998) The role of basal
nal grey matter abnormalities in juvenile myoclonic epilepsy: a voxel-
ganglia in the control of generalized absence seizures. Epilepsy Res
based morphometry study. Neuroimage 37:1132–1137.
32:213–223.
Kloppel S, Draganski B, Golding CV, Chu C, Nagy Z, Cook PA, Hicks SL,
Devinsky O, Gershengorn J, Brown E, Perrine K, Vazquez B, Luciano D.
Kennard C, Alexander DC, Parker GJ, Tabrizi SJ, Frackowiak
(1997) Frontal functions in juvenile myoclonic epilepsy. Neuropsychia-
RS. (2008) White matter connections reflect changes in voluntary-
try Neuropsychol Behav Neurol 10:243–246.
guided saccades in pre-symptomatic Huntington’s disease. Brain
Dexter DT, Carayon A, Javoy-Agid F, Agid Y, Wells FR, Daniel SE, Lees
131:196–204.
AJ, Jenner P, Marsden CD. (1991) Alterations in the levels of iron,
Landvogt C, Buchholz HG, Bernedo V, Schreckenberger M, Werhahn KJ.
ferritin and other trace metals in Parkinson’s disease and other neurode-
(2010) Alteration of dopamine D2/D3 receptor binding in patients with
generative diseases affecting the basal ganglia. Brain 114(Pt 4):1953–
juvenile myoclonic epilepsy. Epilepsia 51:1699–1706.
1975.
Leh SE, Ptito A, Chakravarty MM, Strafella AP. (2007) Fronto-striatal con-
Dexter DT, Jenner P, Schapira AH, Marsden CD. (1992) Alterations in lev-
nections in the human brain: a probabilistic diffusion tractography
els of iron, ferritin, and other trace metals in neurodegenerative diseases
study. Neurosci Lett 419:113–118.
affecting the basal ganglia. The Royal Kings and Queens Parkinson’s
Medina MT, Suzuki T, Alonso ME, Duron RM, Martinez-Juarez IE, Bailey
Disease Research Group. Ann Neurol 32(Suppl):S94–S100.
JN, Bai D, Inoue Y, Yoshimura I, Kaneko S, Montoya MC, Ochoa A,
Douaud G, Behrens TE, Poupon C, Cointepas Y, Jbabdi S, Gaura V, Gole-
Prado AJ, Tanaka M, Machado-Salas J, Fujimoto S, Ito M, Hamano S,
stani N, Krystkowiak P, Verny C, Damier P, Bachoud-Levi AC, Han-
Sugita K, Ueda Y, Osawa M, Oguni H, Rubio-Donnadieu F, Yamakawa
traye P, Remy P. (2009) In vivo evidence for the selective subcortical
K, Delgado-Escueta AV. (2008) Novel mutations in Myoclonin1/
degeneration in Huntington’s disease. Neuroimage 46:958–966.
EFHC1 in sporadic and familial juvenile myoclonic epilepsy. Neurol-
Dreifuss S, Vingerhoets FJ, Lazeyras F, Andino SG, Spinelli L, Delavelle J,
ogy 70:2137–2144.
Seeck M. (2001) Volumetric measurements of subcortical nuclei in
Mohammadi S, Glauche V, Deppe M. (2009) SPM normalization tool-
patients with temporal lobe epilepsy. Neurology 57:1636–1641.
box for voxel-based statistics on fractional anisotropy images. The
Garcia-Finana M, Keller SS, Roberts N. (2009) Confidence intervals for
Organisation of Human Brain Mapping, Neuroimage, San Fransisco,
the volume of brain structures in Cavalieri sampling with local errors.
CA, pp. 122.
J Neurosci Methods 179:71–77.
Mohammadi S, Moller HE, Kugel H, Muller DK, Deppe M. (2010) Correct-
Gattellaro G, Minati L, Grisoli M, Mariani C, Carella F, Osio M, Ciceri E,
ing eddy current and motion effects by affine whole-brain registrations:
Albanese A, Bruzzone MG. (2009) White matter involvement in
evaluation of three-dimensional distortions and comparison with slice-
idiopathic Parkinson disease: a diffusion tensor imaging study. AJNR
wise correction. Magn Reson Med 64:1047–1056.
Am J Neuroradiol 30:1222–1226.
Montalenti E, Imperiale D, Rovera A, Bergamasco B, Benna P.
Geng DY, Li YX, Zee CS. (2006) Magnetic resonance imaging-based volu-
(2001) Clinical features, EEG findings and diagnostic pitfalls in
metric analysis of basal ganglia nuclei and substantia nigra in patients
juvenile myoclonic epilepsy: a series of 63 patients. J Neurol Sci
with Parkinson’s disease. Neurosurgery 58:256–262. Discussion 256–
184:65–70.
262.
O’Muircheartaigh J, Vollmar C, Barker GJ, Kumari V, Symms MR,
Genton P, Gelisse P. (2001) Juvenile myoclonic epilepsy. Arch Neurol
Thompson P, Duncan JS, Koepp MJ, Richardson MP. (2011) Focal
58:1487–1490.
structural changes and cognitive dysfunction in juvenile myoclonic epi-
Glauche V, Mohammadi S, Deppe M. (2010) SPM normalization toolbox
lepsy. Neurology 76:34–40.
for diffusion weighted images. In: Proceedings of the 18th Annual
Ono K, Mori K, Baba H, Wada JA. (1987) A role of the striatum in premo-
Meeting of the International Society of Magnetic Resonance in Medi-
tor cortical seizure development. Brain Res 435:84–90.
cine, Montreal, Canada, Abstract 1648.

Epilepsia, 52(9):1715–1724, 2011

doi: 10.1111/j.1528-1167.2011.03117.x
1724
S. S. Keller et al.
Pfefferbaum A, Adalsteinsson E, Rohlfing T, Sullivan EV. (2010) Diffu-
sion tensor imaging of deep gray matter brain structures: effects of age
and iron concentration. Neurobiol Aging 31:482–493.
Piazzini A, Turner K, Vignoli A, Canger R, Canevini MP. (2008) Frontal
cognitive dysfunction in juvenile myoclonic epilepsy. Epilepsia
49:657–662.
Pulsipher DT, Seidenberg M, Morton JJ, Geary E, Parrish J, Hermann B.
(2007) MRI volume loss of subcortical structures in unilateral temporal
lobe epilepsy. Epilepsy Behav 11:442–449.
Pulsipher DT, Seidenberg M, Guidotti L, Tuchscherer VN, Morton J, Sheth
RD, Hermann B. (2009) Thalamofrontal circuitry and executive dys-
function in recent-onset juvenile myoclonic epilepsy. Epilepsia
50:1210–1219.
Roberts N, Puddephat MJ, McNulty V. (2000) The benefit of stereology for
quantitative radiology. Br J Radiol 73:679–697.
Ruocco HH, Lopes-Cendes I, Li LM, Santos-Silva M, Cendes F. (2006)
Striatal and extrastriatal atrophy in Huntington’s disease and its
relationship with length of the CAG repeat. Braz J Med Biol Res
39:1129–1136.
Santiago-Rodriguez E, Harmony T, Fernandez-Bouzas A, Hernandez-Bal-
deras A, Martinez-Lopez M, Graef A, Carlos Garcia J, Fernandez T.
(2002) Source analysis of polyspike and wave complexes in juvenile
myoclonic epilepsy. Seizure 11:320–324.
Siegel S, Castellan NJ. (1988) Nonparametric statistics for the behavioural
sciences, 2nd ed. McGraw-Hill, New York.
Simonyan K, Tovar-Moll F, Ostuni J, Hallett M, Kalasinsky VF, Lewin-
Smith MR, Rushing EJ, Vortmeyer AO, Ludlow CL. (2008) Focal
white matter changes in spasmodic dysphonia: a combined diffusion
tensor imaging and neuropathological study. Brain 131:447–459.
Sitburana O, Ondo WG. (2009) Brain magnetic resonance imaging
(MRI) in parkinsonian disorders. Parkinsonism Relat Disord 15:165–
174.
Epilepsia, 52(9):1715–1724, 2011
doi: 10.1111/j.1528-1167.2011.03117.x
Smith SM, Zhang Y, Jenkinson M, Chen J, Matthews PM, Federico
A, De Stefano N. (2002) Accurate, robust, and automated
longitudinal and cross-sectional brain change analysis. Neuroimage
17:479–489.
Sowell ER, Thompson PM, Holmes CJ, Jernigan TL, Toga AW. (1999) In
vivo evidence for post-adolescent brain maturation in frontal and stria-
tal regions. Nat Neurosci 2:859–861.
Tae WS, Hong SB, Joo EY, Han SJ, Cho JW, Seo DW, Lee JM, Kim IY,
Byun HS, Kim SI. (2006) Structural brain abnormalities in juvenile
myoclonic epilepsy patients: volumetry and voxel-based morphometry.
Korean J Radiol 7:162–172.
Vulliemoz S, Vollmar C, Koepp MJ, Yogarajah M, O’Muircheartaigh J,
Carmichael DW, Stretton J, Richardson MP, Symms MR, Duncan JS.
(2010) Connectivity of the supplementary motor area in juvenile myo-
clonic epilepsy and frontal lobe epilepsy. Epilepsia 52:507–514.
Wesemann W, Solbach M, Nafe R, Grote C, Sontag KH, Riederer P, Jellin-
ger K, Mennel HD, Clement HW. (1995) Effect of lazaroid U-74389G
on iron-induced reduction of striatal dopamine metabolism. J Neural
Transm Suppl 46:175–182.
Woermann FG, Free SL, Koepp MJ, Ashburner J, Duncan JS. (1999a)
Voxel-by-voxel comparison of automatically segmented cerebral gray
matter – a rater-independent comparison of structural MRI in patients
with epilepsy. Neuroimage 10:373–384.
Woermann FG, Free SL, Koepp MJ, Sisodiya SM, Duncan JS. (1999b)
Abnormal cerebral structure in juvenile myoclonic epilepsy demon-
strated with voxel-based analysis of MRI. Brain 122(Pt 11):2101–2108.
Zhang ZH, Zuo QH, Wu XR. (1989) Effects of lipid peroxidation on
GABA uptake and release in iron-induced seizures. Chin Med J (Engl)
102:24–27.
Zifkin B, Andermann E, Andermann F. (2005) Mechanisms, genetics, and
pathogenesis of juvenile myoclonic epilepsy. Curr Opin Neurol
18:147–153.