A localized pallidal physiomarker in cervical dystonia

Article (PDF Available)inAnnals of Neurology 82(6) · November 2017with 503 Reads 
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
DOI: 10.1002/ana.25095
Cite this publication
Abstract
Objective: Deep brain stimulation (DBS) allows for direct recordings of neuronal activity from the human basal ganglia. In Parkinson's disease, a disease-specific physiomarker was identified that is now used to investigate adaptive closed-loop stimulation in first studies. In dystonia such a physiomarker is missing. Methods: Pallidal oscillations were recorded from 153 contact pairs in 27 patients. We investigated whether power amplitudes in theta and beta bands correlate with dystonic symptom severity across patients. We then projected theta power from each contact pair onto standard subcortical anatomy. This way, we defined a theta hot spot on a group level and investigated whether proximity of the active DBS contacts to it correlates with clinical improvement. Results: Dystonic symptom severity significantly correlated with theta, but not beta oscillatory amplitudes (ρ=0.4; P=0.009) and interhemispheric coherence (ρ=0.5; P=0.002). The sweet spot of theta activity localized to the posterior third of the internal pallidum and theta power correlated with proximity to this location (ρ=0.23; P=0.002), which coincided with three previously published coordinates describing optimal stimulation targets. Finally, motor improvement through pallidal long term DBS correlated with theta peak amplitude (ρ=0.38; P=0.018). Interpretation: Our findings suggest that theta oscillations in the internal pallidum are robustly associated with dystonic symptoms in cervical dystonia and may be a useful biomarker for adaptive closed-loop stimulation. Furthermore, theta oscillatory activity may have a predictive value for the clinical benefit after chronic DBS that could be used to improve intraoperative neurophysiological target mapping during electrode implantation. This article is protected by copyright. All rights reserved.
Figures - uploaded by Wolf-Julian Neumann
Author content
All content in this area was uploaded by Wolf-Julian Neumann
Content may be subject to copyright.
RESEARCH ARTICLE
A Localized Pallidal Physiomarker
in Cervical Dystonia
Wolf-Julian Neumann, MD ,
1
* Andreas Horn, MD, PhD ,
1
*
Siobhan Ewert, MD,
1
Julius Huebl, MD,
1
Christof Br
ucke, MD, PhD,
1
Colleen Slentz, MSc,
1
Gerd-Helge Schneider, MD,
2
and Andrea A. K
uhn, MD
1,3,4
Objective: Deep brain stimulation (DBS) allows for direct recordings of neuronal activity from the human basal gan-
glia. In Parkinson’s disease, a disease-specific physiomarker was identified that is now used to investigate adaptive
closed-loop stimulation in first studies. In dystonia, such a physiomarker is missing.
Methods: Pallidal oscillations were recorded from 153 contact pairs in 27 patients. We investigated whether power
amplitudes in theta and beta bands correlate with dystonic symptom severity across patients. We then projected
theta power from each contact pair onto standard subcortical anatomy. In this way, we defined a theta hot spot on a
group level and investigated whether proximity of the active DBS contacts to it correlated with clinical improvement.
Results: Dystonic symptom severity significantly correlated with theta but not beta oscillatory amplitudes (q50.4,
p50.009) and interhemispheric coherence (q50.5, p50.002). The sweet spot of theta activity localized to the pos-
terior third of the internal pallidum and theta power correlated with proximity to this location (q50.23, p50.002),
which coincided with 3 previously published coordinates describing optimal stimulation targets. Finally, motor
improvement through pallidal long-term DBS correlated with theta peak amplitude (q50.38, p50.018).
Interpretation: Our findings suggest that theta oscillations in the internal pallidum are robustly associated with dys-
tonic symptoms in cervical dystonia and may be a useful biomarker for adaptive closed-loop stimulation. Further-
more, theta oscillatory activity may have a predictive value for the clinical benefit after chronic DBS that could be
used to improve intraoperative neurophysiological target mapping during electrode implantation.
ANN NEUROL 2017;82:912–924
Deep brain stimulation (DBS) of the internal pal-
lidum has been established as a highly effective
treatment for severe idiopathic dystonia in patients that
are refractory to medical therapy.
1–3
Although the clinical
benefit of DBS in cervical and other focal types of dysto-
nia is well documented,
3
the underlying therapeutic
mechanism remains to be elucidated. Converging evi-
dence points to a modulation of aberrant neural popula-
tion activity in the basal ganglia through high-frequency
stimulation.
4
However, the basal ganglia have been diffi-
cult to characterize electrophysiologically, because of their
anatomical distance from the scalp and their complex
interwoven functional pathways. In recent years, DBS
has offered the unique opportunity to record oscillatory
activity directly from the basal ganglia during surgery
and in a postoperative interval in which the DBS electro-
des are externalized. Here, oscillatory patterns of pallidal
local field potentials (LFPs) were found to differ in a
disease-specific manner.
5
This has inspired the concept of
pathological oscillations in movement disorders, which is
now best established in the form of increased beta-
synchronization (13–35Hz) in Parkinson disease (PD).
6,7
A multitude of studies that explored beta activity in PD
have fostered its current application/impact as a bio-
marker in PD that mirrors motor symptom severity and
can be used for target localization
8
and adaptive DBS.
9
View this article online at wileyonlinelibrary.com. DOI: 10.1002/ana.25095
Received Aug 11, 2017, and in revised form Oct 6, 2017. Accepted for publication Nov 5, 2017.
W.-J.N. and A.H. contributed equally.
Address correspondence to Dr K
uhn, Department of Neurology, Charit
e–Universit
atsmedizin Berlin, Campus Charit
e Mitte, Chariteplatz 1, 10117 Berlin,
Germany. E-mail: andrea.kuehn@charite.de
From the
1
Department of Neurology, Movement Disorders and Neuromodulation Unit, Campus Charite Mitt
e, Charit
e–Universit
atsmedizin Berlin;
2
Department of Neurosurgery, Campus Charite Mitt
e, Charit
e–Universit
atsmedizin Berlin;
3
Berlin School of Mind and Brain, Charit
e–Universit
atsmedizin
Berlin; and
4
NeuroCure, Charit
e–Universit
atsmedizin Berlin, Berlin, Germany
Additional supporting information can be found in the online version of this article.
912 V
C2017 American Neurological Association
A recent study used beta power recorded from a large
cohort to map out the subthalamic nucleus functionally,
demonstrating higher activity in active versus inactive
DBS contacts and in the sensorimotor part versus asso-
ciative part of the nucleus.
10
This mounting evidence
demonstrates great clinical significance in using patholog-
ical oscillations to guide DBS surgery and programming,
as well as adaptive closed-loop applications.
In dystonia, such a physiomarker is missing. Low-
frequency activity in the theta–alpha range has been
reported in the internal pallidum
5
and was also shown in
the subthalamic nucleus.
11–13
Both beta band oscillations
in PD
14,15
and theta activity in dystonia patients with
phasic movements are suppressed by high-frequency
DBS.
16
However, no quantitative relationship between
symptom severity and theta activity in dystonia could be
shown to date. The aim of this study was to investigate
the association of theta oscillations with clinical symptom
severity recorded from dystonic patients, while taking
electrode location into account. Specifically, recording
sites across patients were mapped into standard stereotac-
tic space to reveal relationships between recording sites,
theta power, symptom severity, and DBS-related long-
term clinical outcome in a large cohort of patients with
idiopathic cervical dystonia. Here, we investigate 3 main
questions. First, is theta activity correlated to symptom
severity? Second, is theta activity spatially organized
within the nucleus, that is, is it possible to define a hot
spot of activity in the theta band? Finally, we ask whether
proximity of the active DBS contact to such a hot spot is
predictive of clinical improvement across patients and
whether theta activity correlates to DBS-mediated
improvements.
Patients and Methods
Patients and Surgery
In the present study, we focused on patients with cervical dysto-
nia. The reason was to obtain neurophysiological data in a
cohort of dystonic patients that is as homogenous as possible in
their clinical presentation. We recruited 27 patients (17 female,
age 550 69.5 years, mean 6standard error of the mean
[SEM], disease duration 511 66.0 years) with idiopathic cervi-
cal dystonia. All patients underwent the routine workup for
dystonia patients at our center. This includes thorough clinical
examination with exclusion of any additional neurological
symptoms such as pyramidal signs, ataxia, myoclonus, or other
features and exclusion of abnormalities in magnetic resonance
imaging (MRI). Moreover, standard blood panel testing was
performed in all patients with additional copper and ceruloplas-
min testing to exclude Wilson disease, if appropriate. Genetic
testing was conducted in 10/27 patients with no positive
results. Torticollis was present in all cases (head rotation 5
338658). Laterocollis was present in 18/27 patients.
Additionally, some patients exhibited anterocollis, retrocollis,
and lateral shift. Nine of 27 patients suffered from dystonic
tremor. Further clinical details are summarized in Table 1 (see
online Supplementary Table for DBS parameters). LFPs were
recorded, after patients underwent functional neurosurgery for
the implantation of bilateral DBS electrodes in the globus pal-
lidus internus (GPi). The study was approved by the local
ethics committee, with the standards set by the Declaration of
Helsinki. For all patients, written informed consent was
obtained before inclusion in the study. Dystonic symptom
severity was assessed using the severity subscale of the Toronto
Western Spasmodic Torticollis Rating Scale (TWSTRS) pre-
and postoperatively by an experienced movement disorders spe-
cialist, as part of the clinical routine. The implanted macroe-
lectrodes were of type 3389 (n 517) and 3387 (n 510;
Medtronic Neurological Division, Minneapolis, MN) with
1.27mm diameter and 4 cylindrical platinum iridium contacts
of 1.5mm length and 0.5mm/1.5mm (3389/3387) spacing.
Contacts 0 and 3 were the lowermost and uppermost contacts,
respectively. Electrode placement was guided by preoperative
T2- and T1-weighted MRI and intraoperative microelectrode
recordings.
Experiments and Recordings
All LFP recordings took place 2 to 7 days after electrode
implantation, while the electrode lead was still externalized
before the implantation of the pulse generator. All data were
recorded while patients were at rest, sitting comfortably in an
armchair. Patients were asked not to make voluntary move-
ments and to keep their eyes directed on a fixation cross that
was shown on a paper sheet. LFPs were recorded bipolarly from
3 adjacent contact pairs (01, 12, 23) in each GPi (see Fig 1C–
F for a schematic illustration and an example of raw data). A
D360 (Digitimer, Hertfordshire, UK) amplifier was used,
amplification and hardware filter settings were standardized
with a gain of 350.000, and open hardware filters in the range
of 1 to 500Hz were used for LFP recordings in all patients.
Additional surface electromyography (EMG) of dystonic muscle
activity was recorded unilaterally from the sternocleidomastoid
muscle with pairs of Ag/AgCl electrodes in 15/27 cases (Cases
1, 4, 5, 6, 7, 10, 12, 15, 16, 18, 19, 20, 23, 24, 27). The ster-
nocleidomastoid muscle was chosen because it was affected in
all patients, could be accessed in a standardized way, and was
used in previous studies on pallidomuscular coherence in dys-
tonia.
17,18
Amplification for EMG recordings was standardized
at 310,000 and filtered with a high-pass filter at 10Hz. All sig-
nals were digitized at a fixed sampling rate of 1,000Hz using a
1401 A-D converter (CED, Cambridge, UK) and recorded in
Spike2 software (CED) on a personal computer for offline
analysis.
Spectral Analysis
All data from pallidal LFP recordings were imported to MAT-
LAB (Mathworks, Natick, MA) and analyzed using custom
code based on SPM12 (Wellcome Department of Cognitive
Neurology, London, UK; www.fil.ion.ucl.ac.uk/spm/)
19
and
Neumann et al: Cervical Dystonia Physiomarker
December 2017 913
FieldTrip (Donders Center for Cognitive Neuroimaging, Uni-
versity of Nijmegen, Nijmegen, the Netherlands; http://field-
trip.fcdonders.nl/).
20
Before spectral analysis, all data segments
with artifacts were rejected after visual inspection. Average
recording length after artifact rejection was 422 662 seconds
(mean 6SEM). Nine contact pairs from 5 patients (Cases 8,
10, 13, 14, 15) had to be excluded from all analysis because of
continuous contamination by artifacts. The left electrode in
Patient 10 had to be rejected completely because of artifacts.
Thus, 153 pallidal LFP contact pairs from 53 electrodes in 27
patients were taken to the spectral analysis. All continuous data
were epoched into arbitrary segments of 1.024-second length
(1,024 samples) and taken to the frequency domain using fast
Fourier transform–based methods. The resulting power spectra
TABLE 1. Clinical Details of the Patients
No. Age,
yr/Sex
Disease
Duration, yr
TWSTRS,
preop
a
TWSTRS,
postop
Relative
Improvement, %
Electrode
Model
EMG of
SCM
1 41/F 30 17 13 23.5 3389 L
2 70/F 7 23 4 82.6 3389
3 47/F 4 19 10 47.4 3387
4
b
62/F 23 26 na na 3387 R
5 58/F 8 23 3 87.0 3389 R
6 49/F 9 25 10 60.0 3389 L
7
b
65/M 9 18 na na 3387 R
8 52/F 15 21 10 52.4 3387
9 56/F 9 20 9 55.0 3389
10 47/M 18 25 9 64.0 3389 L
11 47/F 5 18 13 30.0 3389
12 45/M 11 23 18 21.7 3389 L
13
b
44/M 7 16 na na 3389
14 51/F 3 24 9 62.5 3387
15 40/M 16 23 11 52.2 3389 R
16 54/M 9 22 12 45.5 3389 L
17 49/F 14 14 11 21.4 3387
18 42/F 8 18 10 44.4 3387 L
19 40/M 10 24 1 95.8 3387 R
20 65/F 10 16 11 31.3 3389 L
21 33/M 15 17 9 47.1 3389
22 53/F 13 23 17 26.1 3387
23 64/F 12 23 5 78.3 3389 L
24 33/M 2 23 5 78.3 3389 L
25 51/F 12 16 11 31.3 3389
26 44/M 11 14 4 71.4 3389
27 49/F 9 20 5 75.0 3389 R
a
TWSTRS severity score.
b
No postop TWSTRS score available in archival dataset.
EMG 5electromyography; F 5female; L5left; M 5male; na 5not available; postop5postoperative; preop 5preoperative; R 5right;
SCM 5sternocleidomastoid muscle; TWSTRS5Toronto Western Spasmodic Torticollis Rating Scale.
ANNALS of Neurology
914 Volume 82, No. 6
were normalized to the standard deviation of the spectrum
from 5–45 and 55–95 Hz as described previously,
16,21
to
exclude spectral features prone to movement artifacts and 50 Hz
noise, respectively. Relative rather than absolute power was ana-
lyzed, because it is less dependent on contact impedances and
therefore allows across-subject analysis. Individual power spectra
from all channels were visually inspected for peaks in the theta
(4–12 Hz) and beta (13–35 Hz) bands, as described in the
recent methodological discussion.
22,23
Relative peak amplitude
was determined by averaging across 3 Hz surrounding the peak
(peak frequency 61.5 Hz; 3 bins). For each electrode, the con-
tact pair with the highest peak amplitude was subjected to a
subsequent analysis. Contact pairs with maximum peak power
were chosen for each frequency band independently. To investi-
gate the spectral distribution of correlative analyses, we aligned
all normalized power spectra to the highest theta and beta peak,
FIGURE 1: Pallidal oscillatory activity. (A) Average spectral power across all patients revealed peaks surrounded by increased
variance (red shaded area indicates standard error of the mean) in the theta (4–12Hz) and beta (13–35Hz) frequency ranges.
(B) Focal peaks shown were present in all patients across these frequency bands, as can be seen in the number of peaks per
frequency bin (peak histogram). (C, D) All local field potentials were recorded bipolarly from adjacent contact pairs as shown
in the schematic (C) from contact pair GPiR23 from Case 19, in which theta oscillations could easily identified in the raw data
(D; blue line is unfiltered raw signal and red line is filtered in the 4–12Hz range). (E, F) This led to peaks in the power spectra
that can fluctuate in amplitude along the recording (E; time–frequency plot) but result in a clear spectral peak after averaging
across time (F). (G) Each DBS electrode was localized based on postoperative magnetic resonance imaging and projected into
standardized stereotactic (Montreal National Institute) space using Lead-DBS. Theta power values from each contact pair (C, D)
were projected to their midpoints and resulted in a 3-dimensional point-cloud in anatomical space. A scattered interpolant was
fitted to the group of data points and projected onto an equidistant grid that could be exported as volumetric data. GPi 5glo-
bus pallidus internus.
Neumann et al: Cervical Dystonia Physiomarker
December 2017 915
respectively (Fig 2E, F), including 3 Hz below and 8 Hz above
the peak frequency (frequency 0 on the x-axis). Because we
were most interested in the theta band, which falls directly into
the 1/f component of the power spectra, we did a supplemen-
tary analysis to avoid contamination of peak amplitudes by dif-
ferences in the noise floor. Furthermore, this procedure does
not impair spatial analysis of different contact pair locations, as
the peak height remains dependent on the proximity to the
LFP source. To linearize the spectra, a logarithmic transforma-
tion was performed for both frequency and power and the
least-squares line was fitted on the basis of the data in the 1 to
3 Hz and 35 to 40 Hz frequency ranges to remove the linear
trend by subtraction.
24
This procedure results in linear spectra,
and peak amplitudes can be analyzed without confound of the
1/f component in the power spectrum. Analysis of linearized
and normalized spectra was kept identical for all further proce-
dures. To investigate potential interhemispheric oscillatory cou-
pling of the pallidal nuclei and to rule out the notion that our
peak power findings are associated with electrode cable move-
ment artifacts in patients who are more severely affected, we
calculated coherence between left and right GPi contact pairs.
Coherence provides a frequency domain measure of the linear
phase and amplitude relationships between signals and can
reveal spectrally specific functional connectivity between neuro-
nal populations. Because the distance of the electrodes across
hemispheres is relatively small, coherence (COH) can be con-
founded by volume conduction that spreads from a common
source without time lag. Imaginary part of coherency
(iCOH)
25,26
is robust against signals that share a zero-lag com-
ponent and thus rules out confound from volume conduction.
However, iCOH does not provide a reliable quantitative esti-
mate of coupling. Therefore, after ensuring that coherence was
present in both iCOH and COH measures, we used COH val-
ues for further quantitative analyses. Both coherence and iCOH
were analyzed between all possible contact pair combinations (9
possible combinations) and averaged per patient separately for
the theta (4–12 Hz) and beta frequency (13–35 Hz) bands.
Like linearized spectra, coherence spectra lack a prominent 1/f
component and do not require additional normalization for
across-patient analysis. Patient 10 had to be excluded from all
interhemispheric coherence analyses, because of the excluded
artifactual left electrode. Finally, pallidomuscular coupling was
analyzed by calculating coherence between recorded EMG activ-
ity (15/27 cases, see Table 1) and the bilateral GPi-LFP contact
pairs with maximum theta peak power.
Mapping Oscillatory Activity in Montreal
National Institute Standard Space
Preoperative MRI was obtained in all patients on a 1.5T scan-
ner using a T2-weighted fast spin-echo. Twenty-five patients
underwent postoperative MRI within 5 days after implantation
of the electrodes. In the remaining 2 patients, postoperative
high-resolution computed tomography was acquired. All DBS
electrodes were localized using Lead-DBS software (http://www.
lead-dbs.org).
27
In brief, linear coregistration of postoperative
to preoperative images was performed using SPM12 and
BRAINSFit software (https://www.nitrc.org/projects/multimo-
dereg/; http://hdl.handle.net/1926/1291). Consecutively, all
images were nonlinearly warped into standard stereotactic
(Montreal National Institute [MNI]; ICBM152 2009b nonlin-
ear) space using a diffeomorphic image registration (DARTEL).
Electrode trajectories are prelocalized by Lead-DBS and were
FIGURE 2: Pallidal theta peak amplitude is correlated with
dystonic symptom severity. (A, B) Pallidal theta peak ampli-
tude (A) averaged across contact pairs, but not beta peak
amplitude (B), correlated significantly with the preoperative
Toronto Western Spasmodic Torticollis Rating Scale
(TWSTRS) severity score (n 527, Spearman q50.4, p5
0.009). (C) The correlation became even stronger when only
the contact pair with the highest theta amplitude was cho-
sen per hemisphere (n 527, Spearman q50.4, p50.009).
(D) Selection of the contact pair with the highest beta peak
power did not reveal significant results. To investigate the
spectral focality of the described correlation with preopera-
tive TWSTRS scores, power spectra were aligned to the
peak frequency and correlations were conducted for each
frequency bin (23to18 Hz surrounding the peak). (E)
Three significant frequency bins were revealed (21to11Hz
surrounding the peak; n 527, p<0.05 uncorrected; light
gray shaded area), but the most robust result was found for
the peak amplitude itself (n 527, p<0.05, false discovery
rate corrected for multiple comparisons; dark gray shaded
area), indicating high spectral focality of the correlation
within the theta band. (F) No significant results were
revealed for the beta band. [Color figure can be viewed at
www.annalsofneurology.org]
ANNALS of Neurology
916 Volume 82, No. 6
manually adjusted to optimally fit the visible artifact in the
postoperative image. Based on electrode localizations, recorded
power values were mapped into standard stereotactic space
using the subcortical electrophysiology mapping approach (see
Fig 1G) introduced in Horn et al.
10
Theta/beta peak ampli-
tudes derived from linearized spectra were mapped to the
Euclidean midpoint between the coordinates representing the 2
electrode contacts from which the signal was recorded. All data
from the left hemisphere were projected to the right hemisphere
and pooled after confirmation that peak amplitudes did not dif-
fer between hemispheres. This led to a scattered point cloud in
standard space, each point associated with an electrophysiologi-
cal value that was projected onto an equidistant grid using nat-
ural neighbor interpolation. The resulting volume represents
electrophysiological estimates for each coordinate that is covered
by the group of electrodes across patients. It was smoothed
with a Gaussian kernel with a full width at half maximum of
0.7mm for visualization. To control for the effect of electrode
localization, we investigated an association between theta power
and distances to (1) a peak site of theta activity and (2) an opti-
mal stimulation site for DBS in dystonia. Thus, distances from
each recording site to 4 points in space were calculated. First,
the spatial theta peak was defined as the centroid of contact
pair locations ranging above the 95th percentile in theta ampli-
tude to test for spatial focality. Second, the centroid between all
active stimulation contacts was calculated in anatomical space.
Third and fourth, 2 literature coordinates described to represent
optimal stimulation sites for DBS in dystonia were used,
28,29
the latter of which was transformed from anterior commissure -
posterior commissure (AC/PC) space into MNI space following
the probabilistic approach defined in Horn et al.
30
This offers
indirect evidence for theta power as a biomarker for optimal
stimulation target. To control for the spectral specificity of the
results, we repeated all analyses on beta peak power.
Statistical Analysis
To assess a potential association between pallidal oscillatory
activity, spatial distance to the 4 coordinates, and clinical scores,
nonparametric rank-based Spearman correlations were calcu-
lated for all analyses. To search for potential theta asymmetry,
we compared ipsilateral and contralateral pallidal theta activity
in relation to the direction of head rotation with paired permu-
tation tests. Moreover, we correlated the difference in ipsi- and
contralateral pallidal theta activity with the degree of head rota-
tion. To address potential differences of theta activity in sub-
jects with dystonic tremor, averaged peak power in patients
with and without dystonic tremor (n 59) was compared. Statis-
tical significance of correlations was determined by Monte
Carlo permutation.
31
Therefore, a test statistic was generated by
calculating 10,000 replications of Spearman correlations from
averaged spectral power and clinical scores with positions of
clinical score values randomly exchanged. Probability values are
reported as the position of the original Spearman correlation
coefficient (q) value in the distribution of the test statistic.
7
For
clinical correlations, peak amplitudes were either averaged over
all available contact pairs (n 56) of both hemispheres or
maximum peak amplitudes per electrode (n 52) were averaged
over both hemispheres per patient. For coherence correlations,
the average 4 to 12Hz coherence values were used either across
all possible contact pair combinations (n 59 per patient) or for
the contact pair combinations with the highest 4 to 12Hz
coherence per electrode (n 51 per patient). To test for a corre-
lation with preoperative clinical symptom severity, correlations
were conducted between the abovementioned oscillatory phe-
nomena and the preoperative TWSTRS severity score in all
patients. To test for a potential predictive value of the peak
amplitude and coherence, we also tested for a correlation with
clinical improvement after at least 3 months of chronic stimula-
tion as the percentage change of the postoperative from the pre-
operative TWSTRS severity score (%TWSTRS
improvement
5
[TWSTRS
preop
2TWSTRS
postop
] / TWSTRS
preop
* 100) and
as the absolute TWSTRS severity change (DTWSTRS 5
TWSTRS
preop
2TWSTRS
postop
). Three patients had to be
excluded from the clinical improvement correlations, because
the postoperative TWSTRS severity score was missing in the
archival dataset. All significant correlations survived correction
for multiple comparisons by controlling the false discovery rate
(FDR)
32
for all reported correlation probability values, except
where explicitly stated otherwise. To test for statistical signifi-
cance of coherence in the 4 to 12Hz band, original data were
compared to shuffled surrogates using permutation tests across
patients. In detail, for each coherence pair a shuffled surrogate
was generated by randomly exchanging the position of an arbi-
trary epoch in only 1 of the 2 signals. This leads to changes in
the phase relationship between the signals and diminishes physi-
ological coherence.
26
Pairwise permutation tests were carried
out to test whether averaged 4 to 12Hz coherence of the origi-
nal signal is significantly greater than the shuffled data across
patients.
21,26
Results
Spectral Distribution of Peaks in the Power
Spectra from Pallidal Recordings
All pallidal LFP power spectra (grand average in Fig 1A)
had peaks in the theta (4–12Hz, n 5148/153 contact
pairs) and beta (13–35Hz, n 5151/153 contact pairs)
frequency ranges (see histogram in Fig 1B). Average peak
frequencies were 7.2 60.2Hz (mean 6SEM) for theta
and 17.5 60.5 Hz for beta frequency ranges. Peaks with
amplitude maxima in the theta range were most often
found in contact pair 12 (contact pair 01: 23%, 12/53
electrodes; contact pair 12: 47%, 25/53 electrodes; con-
tact pair 23: 30%, 16/53 electrodes). The average gradi-
ent in peak amplitude from maximum to remaining
contact pairs was 30 62.7%. Average difference in peak
frequency between contact pairs with peak maxima to
the remaining contact pairs was 0.27 60.25 Hz, which
indicates that they were likely picked up from the same
source and differences in peak amplitudes across contact
pairs may be most dependent on the spatial proximity of
Neumann et al: Cervical Dystonia Physiomarker
December 2017 917
the contact pair to the source location of the local oscil-
lating pallidal neuronal populations. The chosen stimula-
tion contact for chronic stimulation was congruent with
at least one of the contacts displaying the maximum
theta peak in 94% of the electrodes (5/53 stimulation
contacts differed from the maximum peak contact pair).
No difference in theta laterality was revealed (p>0.1),
and no direct correlation between degree of rotation
(including direction) and theta asymmetry was found
(Spearman q520.1, p50.3). Furthermore, theta peak
power was not significantly different in patients who
exhibited dystonic tremor (n 59) when compared to the
remaining cohort (p>0.1).
Correlation of Pallidal Peak Amplitude
and Preoperative Dystonic Symptom Severity
in Cervical Dystonia
The average preoperative TWSTRS I (motor) severity
score was 20 60.7 (mean 6SEM) in our patient cohort.
Pallidal theta peak power averaged across all contact pairs
in each patient correlated significantly with the preopera-
tive clinical symptom severity across all patients (n 527,
Spearman q50.40, p50.005; see Fig 2). This correla-
tion improved when only contact pairs with maximum
peak amplitude per electrode averaged across hemispheres
were subjected to the analysis (n 527, Spearman
q50.48, p50.005). Neither averaged nor maximum
beta peak amplitudes correlated significantly with the
preoperative TWSTRS (n 527, Spearman q520.12/
20.05, p>0.3). The same analyses were conducted for
the linearized power spectra that lack the 1/f component,
which further improved the correlations (n 527, Spear-
man q50.58/0.67, for averaged/maximum peak ampli-
tude, p<0.002). Again, no significant correlations were
revealed for beta peak amplitude (n 527, Spearman
q50.12/0.11, p>0.2; data not shown).
Pallidal Peak Amplitude Correlations with
Clinical Symptom Severity Are Spectrally Focal
To investigate the spectral focality of the significant cor-
relation, we aligned pallidal power spectra of contact
pairs with maximum peak amplitudes to the peak fre-
quency, by stacking spectral power values of the fre-
quency band of 23 to 8 Hz surrounding the theta and
beta peaks, respectively. Correlations with preoperative
TWSTRS scores were then calculated for each frequency
bin (n 512) independently. A narrow band of 3 Hz
width surrounding the peak was significantly correlated
(p<0.05), but only the correlation at the peak center
survived FDR correction for multiple comparisons
(p<0.003; see Fig 2E). For the beta peaks, no significant
frequency bin was revealed (see Fig 2F).
Correlation of Pallidal Interhemispheric
Coherence with Preoperative Symptom Severity
in Patients with Cervical Dystonia
Prominent COH and iCOH were present in pallidal
LFP contact pairs between hemispheres in all patients
(p<0.0001; Fig 3). Again, theta coherence averaged
across all possible contact pairs (n 59) was significantly
correlated with preoperative symptom severity (n 526,
Spearman q50.51, p50.0024; Patient 10 was excluded
from all coherence analyses because of an artifactual left
electrode), but no significant correlation was found for
the beta band (n 526, Spearman q50.17, p50.11).
Correlation with only the maximum coherence contact
pair combination robustly reproduced the result, but did
not improve the correlation for the theta band (n 526,
Spearman q50.51, p50.0024) and also failed to reveal
significant results for the beta band (n 526, Spearman
q50.11, p>0.3).
Pallidal Theta Band Activity Is Coherent
with Dystonic Muscle Activity
To investigate the association of pallidal activity with dys-
tonic symptoms as shown in previous studies,
16
we ana-
lyzed pallidomuscular coupling between dystonic EMG
from sternocleidomastoid muscles and contralateral GPi-
LFP contact pairs with maximum peak amplitude in 15
patients (see Table 1). We found highly significant GPi-
LFP–sternocleidomastoid muscle (SCM)-EMG coupling
in the 4 to 12 Hz band in these patients (see Fig 3D;
p50.008). No significant correlation of 4 to 12 Hz
GPi-LFP–SCM EMG coupling was found with preoper-
ative symptom severity or postoperative therapeutic bene-
fit (p>0.3).
Spatial Localization of Theta Activity in MNI
Space and Correlation of Peak Amplitude
to Theta Peak Location
All trajectories traversed the posterior third of the inter-
nal pallidum as well as the intersection between GPi and
globus pallidus externus previously described as optimal
stimulation targets.
29,30
The trajectory span on the ven-
trodorsal z-axis was 212.4 mm to 10.2 mm, whereas
the internal pallidum spans from 28mmto10.5
mm.
33
Given the predictive value of theta activity with
symptom severity, we focused on theta activity in consec-
utive steps. To investigate whether theta activity was
higher in a specific subregion of the GPi, power values
recorded from each contact pair were projected to the
spatial domain following the approach defined in Horn
et al (see Fig 1G).
10
The spatial centroid of recording
sites exhibiting top 5% theta peak amplitudes was
located in a central position within the internal pallidum
(MNI coordinates: x 5621.35 mm, y 525.98 mm,
ANNALS of Neurology
918 Volume 82, No. 6
z524.46 mm; Fig 4). This location was further con-
firmed by projecting all theta values to the surface of the
GPi without using an a priori percentile. The mean
location of active DBS contacts was situated close
(1.3 mm distance) to this coordinate: x 5620.89 mm,
y526.92 mm, z 525.27 mm. Table 2 and Figure 4A
summarize these coordinates and 2 literature-based coor-
dinates that were reported to represent optimal DBS
stimulation targets in dystonia. Distance from each
recording site to each of these 4 points was calculated
and correlated with theta power. The distance between
the absolute spatial maximum derived from the contact
pair location with maximum peak power and each of the
4 coordinates is also given in Table 2. Crucially, proxim-
ity to any of the 4 coordinates from all contact pair loca-
tions revealed a low but significant correlation with theta
peak amplitude values, indicating the association of lead
position with theta activity (see Fig 4C). The only excep-
tion was the centroid of coordinates for active contacts,
which showed with a probability value of 0.06. When
the experiments were repeated with beta power, none of
the correlations was significant, indicating the spectral
and spatial specificity of pallidal theta activity and its
role in dystonia.
Correlation of Pallidal Theta Peak Amplitude
and Theta Band Coherence with Clinical
Improvement after Chronic Pallidal DBS
All patients improved with chronic (3 months) pallidal
DBS. Average improvement in the TWSTRS across all
patients was 53.5% 64.3% from 20.4 60.7 to 9.2 60.8
points in the severity score (n 524; Cases 4, 7, and 13
were excluded because of missing postoperative
TWSTRS). The maximum theta peak amplitude corre-
lated significantly with relative clinical improvement
(n 524, Spearman q50.38, p50.0229; Fig 5A). The
correlation increased by using linearized spectra (n 524,
Spearman q50.48, p50.01). Average interhemispheric
theta coherence revealed a similarly robust correlation
FIGURE 3.
FIGURE 3: Interhemispheric and pallidomuscular coherence
(COH). (A) Interhemispheric imaginary part of coherency
(iCOH) averaged across all contact pair combinations was sig-
nificant in the theta but not the beta band when compared
to shuffled surrogate data across patients (n 526, averaged
coherence in the 4–12Hz band in inset, p <0.001). (B, C) Simi-
lar to theta peak amplitude, the preoperative Toronto West-
ern Spasmodic Torticollis Rating Scale (TWSTRS) severity
score also correlated significantly with theta coherence (B;
n526, Spearman q50.51, p50.0024) but not beta coher-
ence (C). (D) Pallidal local field potentials were significantly
coupled to dystonic muscle activity recorded through surface
electromyographic electrodes of the sternocleidomastoid
muscle in the theta band (n 515; p50.008). Bar plots in
insets in A and D depict the averaged 4 to 12Hz coherence
for original and shuffled surrogate data (error bars indicate
standard error of the mean) *p<0.01. [Color figure can be
viewed at www.annalsofneurology.org]
Neumann et al: Cervical Dystonia Physiomarker
December 2017 919
with relative improvement (n 523, Spearman q50.45,
p50.0086; see Fig 5B). Both correlations remained
significant when using the absolute change in
DTWSTRS
postoperative–preoperative
(n 524/23, Spearman q5
0.39/0.56 for peak amplitude/coherence, p50.0229/
0.0043). No significant correlations with relative or absolute
change in TWSTRS scores were revealed for the beta band.
Discussion
There are 3 conclusions from this study. First, we show
that theta (4–12Hz) peak amplitude and coherence are
significantly correlated with dystonic symptom severity in
patients with cervical dystonia. Second, maximum theta
band activity was localized within the internal pallidum,
with higher theta activity recorded closer to optimal DBS
target sites for dystonia that have been defined in the lit-
erature, suggesting a direct relationship between the ori-
gin of theta activity and an optimal stimulation site.
Finally, the evidence that peak theta activity seems both
localized to a specific region within GPi and associated
with clinical outcome suggests its potential role as a
physiomarker that may be used in closed-loop applica-
tions. In contrast, there was no clinical significance of
beta activity in GPi-DBS for dystonia.
Role of Pallidal Theta Activity
in the Pathophysiology of Dystonia
Characterizing oscillatory patterns in the basal ganglia of
dystonia patients may be key to further understand the
underlying dysfunction in dystonia. Pallidal theta oscilla-
tions likely originate from theta bursting cells in the GPi
that were described in single unit recordings during DBS
surgery.
34
Although an association of pallidal activity
with dystonic EMG was previously described,
17
the spec-
tral specificity of pallidal oscillations in the pathophysiol-
ogy of dystonia is currently under debate.
13
The presence
of both beta and theta oscillations in pallidal LFP record-
ings has cast doubt on the pathophysiological implication
of either of these rhythms in dystonia. Crucially, our
results demonstrate a spectrally highly focal association of
peak theta activity with clinical symptom severity in a
large cohort of cervical dystonia patients. Furthermore,
interhemispheric theta band coherence (previously
described by Moll et al
34
robustly replicates the clinical
correlation with dystonic symptoms and integrates our
finding with the observation that pallidomuscular net-
work activity is driven by pallidal oscillations.
18
Both
neuroimaging and electrophysiological studies have
revealed structural and functional network changes in
dystonic patients that have resulted in the current opin-
ion that dystonia is a circuit disorder.
35–37
Previous stud-
ies have revealed alterations of pallidocortical coherence
through parallel LFP-MEG recordings with reduced pal-
lidocerebellar connectivity in dystonia patients.
26
Interest-
ingly, pallidal theta activity was found to be negatively
correlated with physiological motor cortical beta and cer-
ebellar alpha band oscillations.
38
Given our findings on
pallidal coherence and the previously described presence
of theta activity in the subthalamic nucleus in dystonia
patients,
11–13
we hypothesize that subcortical theta
FIGURE 4: Relationship between theta power and recording
site. (A) Central landmarks cluster in the ventral internal pal-
lidum. Red circle shows the centroid of the recording sites
fromwhichthetapowerabovethe95thpercentilewas
recorded. (B–D) Raw theta power values mapped to the sur-
faceoftheglobuspallidusinternus(GPi;B)andvisualizedin
axial (planes cut at height of the theta peak at z 524.46mm;
C) and coronal (x 5221.35 mm; D) planes show the overlap of
the landmark cluster with the spatial theta peak distribution.
(E) All recording sites were visualized (white circles).Euclidean
distances between each site and the 4 landmarks from A were
calculated as shown for 2 example sites and the theta power
peak. (F) correlation between distance to theta peak and theta
power of each recording site. For correlations with distances
to the other sites, see Table 2. The same examples as in E are
color coded in the scatter plot. GPe 5globus pallidus externus;
OT 5optic tract; post. 5posterior; Put. 5putamen; STN 5sub-
thalamic nucleus; GPi = globus pallidus internus; colored aster-
isks depict target locations (see legend in A).
ANNALS of Neurology
920 Volume 82, No. 6
synchronization may represent a “noisy signal” as pro-
posed by Marsden and Obeso
39
in the cortex–basal gan-
glia–thalamic loop that may interfere with physiological
motor network function, leading to the generation of
dystonic muscle contractions. However, the underlying
mechanism causing the emergence of theta oscillations in
dystonia remains to be elucidated and could be a key
to integrating findings of increased cortical excitability
and reduced intracortical inhibition
40
in dystonia pa-
tients with the described dystonic network alterations.
Although we propose that theta activity is not an entirely
local phenomenon but represents network communica-
tion, the modulation of single nodes of this network
could lead to restoration of physiological signaling in
parallel with clinical symptom alleviation. DBS has
recently been proposed to stabilize subcortical activation
patterns, which could be more tolerable for downstream
circuit computations.
41
In a previous study, we could
demonstrate that pallidal DBS leads to a rapid decrease
in pallidal theta activity in patients with phasic dysto-
nia,
16
who are known to respond more quickly to short-
term DBS.
42
Although these findings support the notion
that suppression of theta activity may be one potential
mechanism of pallidal DBS in dystonia, long-term
recordings after chronic DBS will have to prove the
stability of theta oscillations as recently reported for beta
activity in PD.
43
A Sweet Spot of Pallidal Theta
Activity in Dystonia
The definition of an optimal stimulation site for dystonia
is critical given the latency of therapeutic effects that
TABLE 2. Correlation of Theta Peak Power with Distance to Pallidal Landmarks
Label Coordinate,
Left, mm
Coordinate,
Right, mm
q/p, Theta vs
Distance
Origin
Peak theta activity x: 221.35 x: 21.35 q50.23, p<0.003 This study
y: 25.98 y: 25.98
z: 24.46 z: 24.46
Average active contact x: 220.87 x: 20.92 q50.12, p50.06 This study
y: 26.94 y: 26.90
z: 25.27 z: 25.27
Optimal GPi-DBS target x: 222.37 x: 22.37 q50.25, p<0.001 Starr 2006
y: 25.57 y: 25.57
z: 24.97 z: 24.97
Optimal GPi-DBS target x: 220.13 x: 20.58 q50.15, p<0.05 Schonecker 2015
28
y: 26.79 y: 26.56
z: 25.38 z: 24.84
Four relevant pallidal landmarks coordinates were isolated from data shown in this study as well as prior literature. Distance between recording sites
and each of these coordinates was correlated with theta power recorded at the site (last column) to reveal a systematic distribution of theta power
in the spatial domain and a relationship between theta power and the optimal DBS stimulation site.
a
Converted to Montreal National Institute coordinates in Horn et al.
30
DBS 5deep brain stimulation; GPi 5globus pallidus internus.
FIGURE 5: Theta amplitude and coherence are correlated
with therapeutic benefit through pallidal deep brain stimula-
tion (DBS). Both maximum theta peak amplitude averaged
across hemispheres (A; n 524, Spearman q50.38, p5
0.0176) and interhemispheric coherence (B; n 523, Spear-
man q50.45, p50.0086) correlated significantly with rela-
tive improvement in the Toronto Western Spasmodic
Torticollis Rating Scale severity score after 3 months of
chronic pallidal DBS. [Color figure can be viewed at www.
annalsofneurology.org]
Neumann et al: Cervical Dystonia Physiomarker
December 2017 921
render patient programming more complex than in other
movement disorders. More and more recent studies have
aimed at predicting patient outcome by means of DBS
computer simulations, but they have mostly focused on
PD (eg, Horn et al,
44
Eisenstein et al
45
. So far, in dysto-
nia, few studies have investigated a clinical or pathophys-
iologic relationship between DBS placement and clinical
outcome, but all concluded that the optimal stimulation
site resides in the posteroventral third of the internal pal-
lidum.
28,29,46–48
This region of the nucleus corresponds
to its sensorimotor functional zone.
30,48
We demonstrate
that the peak of theta activity coincides with a location
that was defined as the optimal stimulation site in the
posterolateral portion of the internal pallidum in 2 retro-
spective studies.
28,29
Moreover, in our sample, proximity
between the active contact and this sweet spot correlated
with clinical outcome. Combined, these results serve as
indirect evidence that a mechanism of action of DBS in
dystonia could be theta suppression. Pathophysiologically,
the stimulation site within the posteroventral GPi could
exert (part of) its therapeutic effect by modulating 2 pal-
lidal outflow tracts, the ansa lenticularis and lenticular
fasciculus.
49
Based on the localization of the sweet spot
in the present and previous studies, no direct association
to either fiber bundle is possible. However, our results
extend the relationship between clinical outcome and
stimulation location within the posteroventral GPi with
electrophysiological evidence for high synchronization in
the theta band.
Crucially, theta peak amplitudes were correlated
with clinical benefit after 3 months of stimulation and
may therefore have a predictive clinical value. Moreover,
we prospectively validated previously reported optimal
DBS targets by means of theta activity. Combined, these
2 findings are of great clinical interest, as theta oscilla-
tions are easily recordable during electrode implanta-
tion
34
and could be used to indicate the target with best
dystonic symptom alleviation. This is of particular rele-
vance, as unlike in PD, intraoperative stimulation testing
is of little informative value for the prediction of clinical
benefit. Thus, indexing theta peak power for all recorded
trajectories while advancing through the globus pallidus
could be a valuable addition to standard neurophysiologi-
cal procedures during DBS electrode surgery in dystonia
patients, but future studies are needed to evaluate its
usefulness.
A Potential Physiomarker for Closed-Loop
DBS in Dystonia
Above paragraphs summarize that theta activity (1) is
associated with symptom severity and (2) likely peaks in
a localized source to which (3) proximity with the active
DBS contact is predictive of clinical improvement. These
findings render theta synchrony a potentially highly rele-
vant physiomarker for dystonia, much like beta activity
in PD, which is equally predictive for motor symptom
severity and is spatially confined to a peak within the
subthalamic nucleus.
6,7,10
How theta oscillations could
be used to improve clinical outcome in dystonia should
be investigated in future studies. For now, we see 3
potential applications. As mentioned above, it could be
used to guide DBS placement intraoperatively by analyz-
ing LFPs from microelectrode recordings. Second, it
could be used to guide DBS programming by establish-
ing algorithms that analyze theta synchronization from
all contact pairs of the electrode. Such algorithms could
become increasingly relevant with use of segmented DBS
lead designs.
4
Relatedly, theta activity could find a poten-
tial use in adaptive closed-loop stimulation for dystonia
similar to successful applications in PD.
9,50
Here, stimu-
lation parameters could be adapted across time in res-
ponse to activity in the theta band. Further studies are
needed to explore and validate these concepts, which
seem promising based on present evidence.
Conclusions
Our data demonstrate that pallidal theta oscillatory activ-
ity is correlated with motor symptom severity in patients
with cervical dystonia. Moreover, we could show that
theta oscillations represent synchronized network activity
that is correlated with clinical symptom severity and also
coherent with dystonic EMG. We demonstrate local con-
finement of theta activity to optimal DBS targets
reported in the literature. These targets spatially coincide
with peak theta activity, establishing a physiological link
between optimal DBS electrode location and clinical
DBS effect. Pallidal theta oscillations may be key to
understanding the pathophysiology of dystonia and could
act as a useful biomarker of motor symptoms for target-
ing and adaptive DBS.
Acknowledgment
This work was funded by German Federal Ministry of
Education and Research grant Dystract 01GM1514D,
German Research Foundation grant KFO247, and Med-
tronic. A.H. received funding from Stiftung Charite, the
Berlin Institute of Health, and the Professor Klaus Thie-
mann Foundation.
Author Contributions
W.-J.N., A.H., and A.A.K. contributed to the design of
the study; all authors contributed to the acquisition and
ANNALS of Neurology
922 Volume 82, No. 6
analysis of data; W.-J.N., A.H., and A.A.K. contributed
to drafting the text and preparing the figures.
Potential Conflicts of Interest
W.-J.N. has received nonfinancial support from Med-
tronic and St Jude. G.-H.S. and A.A.K. have business
relationships with Medtronic, St. Jude, and Boston Sci-
entific, which are makers of DBS devices, but none is
related to the current work. A.A.K. has received research
funding in the context of this study from Medtronic, the
manufacturer of the DBS electrodes reported in Patients
and Methods.
References
1. Vidailhet M, Vercueil L, Houeto JL, et al. Bilateral deep-brain stim-
ulation of the globus pallidus in primary generalized dystonia. N
Engl J Med 2005;352:459–467.
2. Kupsch A, Benecke R, M
uller J, et al. Pallidal deep-brain stimula-
tion in primary generalized or segmental dystonia. N Engl J Med
2006;355:1978–1990.
3. Volkmann J, Mueller J, Deuschl G, et al. Pallidal neurostimulation
in patients with medication-refractory cervical dystonia: a rando-
mised, sham-controlled trial. Lancet Neurol 2014;13:875–884.
4. K
uhn AA, Volkmann J. Innovations in deep brain stimulation meth-
odology. Mov Disord 2017;32:11–19.
5. Silberstein P, K
uhn AA, Kupsch A, et al. Patterning of globus pal-
lidus local field potentials differs between Parkinson’s disease and
dystonia. Brain 2003;126(pt 12):2597–2608.
6. K
uhn AA, Kupsch A, Schneider GH, Brown P. Reduction in subtha-
lamic 8-35 Hz oscill atory activity correlates with clinical improve-
ment in Parkinson’s disease. Eur J Neurosci 2006;23:1956–1960.
7. Neumann WJ, Degen K, Schneider GH, et al. Subthalamic syn-
chronized oscillatory activity correlates with motor impairment in
patients with Parkinson’s disease. Mov Disord 2016;31:1748–1751.
8. Tamir I, Marmor-Levin O, Eitan R, et al. Posterolateral trajectories
favor a longer motor domain in subthalamic nucleus deep brain
stimulation for Parkinson disease. World Neurosurg 2017;106:450–
461.
9. Little S, Pogosyan A, Neal S, et al. Adaptive deep brain stimula-
tion in advanced Parkinson disease. 2013;74:449–457.
10. Horn A, Neumann WJ, Degen K, et al. Toward an electrophysio-
logical “sweet spot” for deep brain stimulation in the subthalamic
nucleus. Hum Brain Mapp 2017 Apr 8. doi:10.1002/hbm.
11. Geng X, Zhang J, Jiang Y, et al. Comparison of oscillatory activity
in subthalamic nucleus in Parkinson’s disease and dystonia. Neu-
robiol Dis 2017;98:100–107.
12. Neumann WJ, Huebl J, Br
ucke C, et al. Enhanced low-frequency
oscillatory activity of the subthalamic nucleus in a patient with dys-
tonia. Mov Disord 2012;27:1063–1066.
13. Wang DD, de Hemptinne C, Miocinovic S, et al. Subthalamic local
field potentials in Parkinson’s disease and isolated dystonia: an
evaluation of potential biomarkers. Neurobiol Dis 2016;89:213–
222.
14. Neumann WJ, Staub F, Horn A, et al. Deep brain recordings using
an implanted pulse generator in Parkinson’s disease. Neuromodu-
lation 2016;19:20–24.
15. Eusebio A, Thevathasan W, Doyle Gaynor L, et al. Deep brain
stimulation can suppress pathological synchronisation in par-
kinsonian patients. J Neurol Neurosurg Psychiatry 2011;82:
569–573.
16. Barow E, Neumann WJ, Br
ucke C, et al. Deep brain stimulation
suppresses pallidal low frequency activity in patients with phasic
dystonic movements. Brain 2014;137(pt 11):3012–3024.
17. Liu X, Wang S, Yianni J, et al. The sensory and motor representa-
tion of synchronized oscillations in the globus pallidus in patients
with primary dystonia. Brain 2008;131(pt 6):1562–1573.
18. Sharott A, Grosse P, K
uhn AA, et al. Is the synchronization between
pallidal and muscle activity in primary dystonia due to peripheral
afferance or a motor drive? Brain 2008;131(pt 2):473–484.
19. Litvak V, Mattout J, Kiebel S, et al. EEG and MEG data analysis in
SPM8. Comput Intell Neurosci 2011;2011:852961.
20. Oostenveld R, Fries P, Maris E, Schoffelen JM. FieldTrip: open
source software for advanced analysis of MEG, EEG, and invasive
electrophysiological data. Comput Intell Neurosci 2011;2011:
156869.
21. Neumann WJ, Huebl J, Br
ucke C, et al. Different patterns of local
field potentials from limbic DBS targets in patients with major
depressive and obsessive compulsive disorder. Mol Psychiatry
2014;19:1186–1192.
22. Beudel M, Oswal A, Jha A, et al. Oscillatory beta power correlates
with akinesia-rigidity in the parkinsonian subthalamic nucleus. Mov
Disord 2017;32:174–175.
23. Neumann WJ, K
uhn AA. Subthalamic beta power-Unified Parkin-
son’s Disease Rating Scale III correlations require akinetic symp-
toms. Mov Disord 2017;32:175–176.
24. Nikulin VV, Brismar T. Phase synchronization between alpha and
beta oscillations in the human electroencephalogram. Neurosci-
ence 2006;137:647–657.
25. Nolte G, Bai O, Wheaton L, et al. Identifying true brain interaction
from EEG data using the imaginary part of coherency. Clin Neuro-
physiol 2004;115:2292–2307.
26. Neumann WJ, Jha A, Bock A, et al. Cortico-pallidal oscillatory
connectivity in patients with dystonia. Brain 2015;138(pt 7):1894–
1906.
27. Horn A, K
uhn AA. Lead-DBS: a toolbox for deep brain stimulation
electrode localizations and visualizations. Neuroimage 2015;107:
127–135.
28. Sch
onecker T, Gruber D, Kivi A, et al. Postoperative MRI localisa-
tion of electrodes and clinical efficacy of pallidal deep brain stimu-
lation in cervical dystonia. J Neurol Neurosurg Psychiatry 2015;86:
833–839.
29. Starr PA, Turner RS, Rau G, et al. Microelectrode-guided implan-
tation of deep brain stimulators into the globus pallidus internus
for dystonia: techniques, electrode locations, and outcomes. J.
Neurosurg. 2006;104:488–501. http://dx.doi.org/10.3171/ jns.
2006.104.4.488.
30. Horn A, K
uhn AA, Merkl A, et al. Probabilistic conversion of neu-
rosurgical DBS electrode coordinates into MNI space. Neuro-
image 2017;150:395–404.
31. Good PI. Permutation, parametric and bootstrap tests of hypothe-
ses. New York, NY: Springer-Verlag, 2005.
32. Benjamini Y, Krieger AM, Yekutieli D. Adaptive linear step-up pro-
cedures that control the false discovery rate. Biometrika 2006;93:
491–507.
33. Ewert S, Plettig P, Li N, et al. Toward defining deep brain stimulation
targets in MNI space: a subcortical atlas based on multimodal MRI,
histology and structural connectivity. Neuroimage 2017 May 20. pii:
S1053-8119(17)30407-X. doi: 10.1016/j.neuroimage.2017.05.015.
34. Moll CK, Galindo-Leon E, Sharott A, et al. Asymmetric pallidal
neuronal activity in patients with cervical dystonia. Front Syst Neu-
rosci 2014;8:15.
35. Lehericy S, Tijssen MA, Vidailhet M, et al. The anatomical basis of
dystonia: current view using neuroimaging. Mov Disord 2013;28:
944–957.
Neumann et al: Cervical Dystonia Physiomarker
December 2017 923
36. Hendrix CM, Vitek JL. Toward a network model of dystonia. Ann
N Y Acad Sci 2012;1265:46–55.
37. Quartarone A, Hallett M. Emerging concepts in the physiological
basis of dystonia. Mov Disord 2013;28:958–967.
38. Neumann WJ, K
uhn AA. Reply: Role of cortico-pallidal connectiv-
ity in the pathophysiology of dystonia. Brain 2016;139(pt 9):e49.
39. Marsden CD, Obeso JA. The functions of the basal ganglia and
the paradox of stereotaxic surgery in Parkinson’s disease. Brain
1994;117(pt 4):877–897.
40. Quartarone A, Bagnato S, Rizzo V, et al. Abnormal associative
plasticity of the human motor cortex in writer’s cramp. Brain 2003;
126(pt 12):2586–2596.
41. Wichmann T, DeLong MR. Deep brain stimulation for movement
disorders of basal ganglia origin: restoring function or functional-
ity? Neurotherapeutics 2016;13:264–283.
42. Chung M, Huh R. Different clinical course of pallidal deep brain
stimulation for phasic- and tonic-type cervical dystonia. Acta Neu-
rochir (Wien) 2016;158:171–180.
43. Neumann WJ, Staub-Bartelt, F, Horn A, et al. Long term correla-
tion of subthalamic beta band activity with motor impairment in
patients with Parkinson’s disease. Clin Neurophysiol 2017;128:
2286–2291.
44. Horn A, Reich M, Vorwerk J, et al. Connectivity predicts deep
brain stimulation outcome in Parkinson disease. Ann Neurol 2017;
82:67–78.
45. Eisenstein SA, Koller JM, Black KD, et al. Functional anatomy of
subthalamic nucleus stimulation in Parkinson disease. Ann Neurol
2014;76:279–295.
46. Cheung T, Noecker AM, Alterman RL, et al. Defining a therapeutic
target for pallidal deep brain stimulation for dystonia. Ann Neurol
2014;76:22–30.
47. Vayssiere N, van der Gaag N, Cif L, et al. Deep brain stimulation
for dystonia confirming a somatotopic organization in the globus
pallidus internus. J Neurosurg 2004;101:181–188.
48. Cif L, Gonzalez-Martinez V, Vasques X, et al. Staged implantation
of multiple electrodes in the internal globus pallidus in the treat-
ment of primary generalized dystonia. J Neurosurg 2012;116:
1144–1152.
49. Tisch S, Rothwell JC, Bhatia KP, et al. Pallidal stimulation modifies
after-effects of paired associative stimulation on motor cortex
excitability in primary generalised dystonia. Exp Neurol 2007;206:
80–85.
50. Priori A, Foffani G, Rossi L, Marceglia S. Adaptive deep brain
stimulation (aDBS) controlled by local field potential oscillations.
Exp Neurol 2013;245:77–86.
ANNALS of Neurology
924 Volume 82, No. 6
  • ... The response to DBS in dystonia is quite variable and might depend on several clinical, genetic and electrophysiological factors. [53,54] The current evidence ranges 5 from class I evidence for isolated generalized dystonia to class IV evidence for some acquired combined dystonia. Thus, considering a patient with dystonia for DBS sometimes may not be straightforward, as reliable predictive factors are still under study. ...
    Thesis
    Full-text available
    Dystonia is a motor disorder characterized by involuntary and sustained muscle contractions which, following their diffusion to a limb, a limb segment or the whole body, cause abnormal attitudes and postures. It is a very serious and disabling disease, affecting the personal, professional, economic and psychosocial lives of patients. However, there is currently no cure for the disease. Medical treatment may decrease or delay symptoms but remains very limited in dystonia. As for the ablative surgery which represents an aggressive technique and having several side effects. Deep brain stimulation is a revolutionary technique that has been shown to be effective in the treatment of dystonia, but it is unfortunately not widely used. This prospective study reports the experience of Ibn Rochd University Hospital’s of Casablanca, which is the first center in Morocco and the Arab world that uses this technique in the treatment of dystonias.
  • ... Additionally, an investigation of VTAs of pallidal electrodes in TOR1A gene mutation carriers attributed the most pronounced clinical improvements to stimulation of the middle aspect of the posterior GPi [18]. Another study correlating theta oscillations with clinical outcome in CD patients localized the optimal spot for stimulation within the posterior third of the GPi [19]. A recent multicenter study in patients with cervical and generalized dystonia applied a probabilistic approach and suggested that the optimal spot for antidystonic effects was the ventro-posterior GPi and the adjacent subpallidal white matter [15]. ...
    Article
    Background Deep brain stimulation (DBS) within the pallidum represents an effective and well-established treatment for isolated dystonia. However, clinical outcome after surgery may be variable with limited response in 10–25% of patients. The effect of lead location on clinical improvement is still under debate.Objective To identify stimulated brain regions associated with the most beneficial clinical outcome in dystonia patients.Methods18 patients with cervical and generalized dystonia with chronic DBS of the internal pallidum were investigated. Patients were grouped according to their clinical improvement into responders, intermediate responders and non-responders. Magnetic resonance and computed tomography images were co-registered, and the volume of tissue activated (VTA) with respect to the pallidum of individual patients was analysed.ResultsVTAs in responders (n = 11), intermediate responders (n = 3) and non-responders (n = 4) intersected with the posterior internal (GPi) and external (GPe) pallidum and the subpallidal area. VTA heat maps showed an almost complete overlap of VTAs of responders, intermediate and non-responders. VTA coverage of the GPi was not higher in responders. In contrast, VTAs of intermediate and non-responders covered the GPi to a significantly larger extent in the left hemisphere (p < 0.01).ConclusionsDBS of ventral parts of the posterior GPi, GPe and the adjacent subpallidal area containing pallidothalamic output projections resulted in favourable clinical effects. Of note, non-responders were also stimulated within the same area. This suggests that factors other than mere lead location (e.g., clinical phenotype, genetic background) have determined clinical outcome in the present cohort.
  • ... This confirmed that elevated beta-power is predominantly expressed within the sensorimotor functional zone of the STN, a finding that was reproduced and extended by two different teams again by the use of a Lead group prototype (Geng et al., 2018;van Wijk et al., 2017). The concept is now referred to as subcortical electrophysiology mapping and the involved research questions shaped functionality and continued development of Lead group while being applied in further studies Neumann et al., 2017;Tinkhauser et al., 2019). ...
    Preprint
    Full-text available
    Deep Brain Stimulation (DBS) is an established treatment option for movement disorders and is investigated to treat a growing number of other brain disorders. It has been shown that DBS effects are highly dependent on exact electrode placement, which is especially important when probing novel indications or stereotactic targets. Thus, considering precise electrode placement is crucial when investigating efficacy of DBS targets. To measure clinical improvement as a function of electrode placement, neuroscientific methodology and specialized software tools are needed. Such tools should have the goal to make electrode placement comparable across patients and DBS centers, and include statistical analysis options to validate and define optimal targets. Moreover, to allow for comparability across different research sites, these need to be performed within an algorithmically and anatomically standardized and openly available group space. With the publication of Lead-DBS software in 2014, an open-source tool was introduced that allowed for precise electrode reconstructions based on pre- and postoperative neuroimaging data. Here, we introduce Lead Group, implemented within the Lead-DBS environment and specifically designed to meet aforementioned demands. In the present article, we showcase the various processing streams of Lead Group in a retrospective cohort of 51 patients suffering from Parkinson's disease, who were implanted with DBS electrodes to the subthalamic nucleus (STN). Specifically, we demonstrate various ways to visualize placement of all electrodes in the group and map clinical improvement values to subcortical space. We do so by using active coordinates and volumes of tissue activated, showing converging evidence of an optimal DBS target in the dorsolateral STN. Second, we relate DBS outcome to the impact of each electrode on local structures by measuring overlap of stimulation volumes with the STN. Finally, we explore the software functions for connectomic mapping, which may be used to relate DBS outcomes to connectivity estimates with remote brain areas. We isolate a specific fiber bundle - which structurally resembles the hyperdirect pathway - that is associated with good clinical outcome in the cohort. The manuscript is accompanied by a walkthrough tutorial through which users are able to reproduce all main results presented in the present manuscript. All data and code needed to reproduce results are openly available.
  • ... Previous research identified abnormal theta oscillations at subcortical and cortical levels in other forms of focal dystonia such as cervical dystonia 38,39 . These abnormal theta oscillations in globus pallidus internus significantly correlate with the severity of symptoms in cervical dystonia 40 . ...
    Article
    Full-text available
    Spasmodic dysphonia (SD) is an incurable focal dystonia of the larynx that impairs speech and communication. Vibro-tactile stimulation (VTS) alters afferent proprioceptive input to sensorimotor cortex that controls speech. This proof-of-concept study examined the effect of laryngeal VTS on speech quality and cortical activity in 13 SD participants who vocalized the vowel /a/ while receiving VTS for 29 minutes. In response to VTS, 9 participants (69%) exhibited a reduction of voice breaks and/or a meaningful increase in smoothed cepstral peak prominence, an acoustic measure of voice/speech quality. Symptom improvements persisted for 20 minutes past VTS. Application of VTS induced a significant suppression of theta band power over the left somatosensory-motor cortex and a significant rise of gamma rhythm over right somatosensory-motor cortex. Such suppression of theta oscillations is observed in patients with cervical dystonia who apply effective sensory tricks, suggesting that VTS in SD may activate a similar neurophysiological mechanism. Results of this feasibility study indicate that laryngeal VTS modulates neuronal synchronization over sensorimotor cortex, which can induce short-term improvements in voice quality. The effects of long-term VTS and its optimal dosage for treating voice symptoms in SD are still unknown and require further systematic study.
  • ... Many studies of human intraoperative localization of DBS targets have focused on the STN alone [13,14]. However, there is increasing interest in identifying pallidal physiomarkers [15][16][17]. The GPi is the primary target for dystonia, and it is increasingly used for PD, based on the finding of equivalent motor benefits in three out of four randomized trials of STN versus GPi-DBS [1][2][3][4]. ...
    Article
    Objective: Deep brain stimulation (DBS) of the internal segment of the globus pallidus (GPi) in patients with Parkinson's disease and dystonia improves motor symptoms and quality of life. Traditionally, pallidal borders have been demarcated by electrophysiological microelectrode recordings (MERs) during DBS surgery. However, detection of pallidal borders can be challenging due to the variability of the firing characteristics of neurons encountered along the trajectory. MER can also be time-consuming and therefore costly. Here we show the feasibility of real-time machine learning classification of striato-pallidal borders to assist neurosurgeons during DBS surgery. Approach: An electrophysiological dataset from 116 trajectories of 42 patients consisting of 11,774 MER segments of background spiking activity in five classes of disease was used to train the classification algorithm. The five classes included awake Parkinson's disease patients, as well as awake and lightly anesthetized genetic and non-genetic dystonia patients. A machine learning algorithm was designed to provide prediction of the striato-pallidal borders, based on hidden Markov models and the L1-distance measure in normalized root mean square (NRMS) and power spectra of the MER. We tested its performance prospectively against the judgment of 3 electrophysiologists in the operating rooms of three hospitals using newly collected data. Main results: The awake and the light anesthesia dystonia classes could be merged. Using MER NRMS and spectra, the machine learning algorithm was on par with the performance of the three electrophysiologists across the striatum-GPe, GPe-GPi, and GPi-exit transitions for all disease classes. Significance: machine learning algorithms enable real-time GPi navigation systems to potentially shorten the duration of electrophysiological mapping of pallidal borders, while ensuring correct pallidal border detection.
  • ... 8,[11][12][13] We have recently shown that pallidal low-frequency activity, but not beta band activity, (that is also present in dystonia 14 ) correlates with the severity of dystonic motor signs, and patients with electrode position close to the maximum low-frequency local field potential (LFP) peak within the pallidum had the best motor outcome with chronic stimulation. 15 Moreover, DBS could suppress low-frequency activity in the short term in patients with predominant phasic, but not tonic, dystonic movements. 16 However, in dystonia the main clinical effect manifests during a much longer time scale of several weeks or months of continuous stimulation, which has been related to plastic changes and cortico-subcortical reorganization induced by DBS. ...
  • ... Various classes of signals have been used to determine when and how much to stimulate, including pathological neural activity 79,82,[88][89][90] and peripheral measurements 91,92 . Biomarkers need not necessarily be directly related to disease mechanisms, but should correlate with the severity of disease symptoms 36,93-95 and track the response to therapeutic interventions 39, 46,93,95,96 . The relevant signals may be relatively unprocessed or subject to several processing steps to extract the information of interest, with or without the aid of machine learning. ...
    Article
    Deep brain stimulation (DBS) is an effective treatment for common movement disorders and has been used to modulate neural activity through delivery of electrical stimulation to key brain structures. The long-term efficacy of stimulation in treating disorders, such as Parkinson’s disease and essential tremor, has encouraged its application to a wide range of neurological and psychiatric conditions. Nevertheless, adoption of DBS remains limited, even in Parkinson’s disease. Recent failed clinical trials of DBS in major depression, and modest treatment outcomes in dementia and epilepsy, are spurring further development. These improvements focus on interaction with disease circuits through complementary, spatially and temporally specific approaches. Spatial specificity is promoted by the use of segmented electrodes and field steering, and temporal specificity involves the delivery of patterned stimulation, mostly controlled through disease-related feedback. Underpinning these developments are new insights into brain structure–function relationships and aberrant circuit dynamics, including new methods with which to assess and refine the clinical effects of stimulation. Advances in deep brain stimulation technologies are poised to improve outcomes in the clinic.
  • Chapter
    In the last decades, neuromodulation, especially deep brain stimulation (DBS), has become an important treatment option in many medical refractory neurological and psychiatric disorders. However, there are still many limitations of DBS especially in terms of efficacy, side effects, and efficiency. A main reason explaining these limitations is the traditional open-loop DBS design, which allows a constant level of stimulation that does not correspond with the fluctuating clinical need. One way to circumvent this limitation is to make DBS act in a responsive way based on the presence of pathological neural activity or other biomarkers. This form of stimulation is called adaptive DBS (aDBS) or closed-loop DBS. At present the only disorder in which aDBS is clinically applied is epilepsy. However, there is an emerging field working on aDBS in other neurological disorders, especially movement disorders, with promising results. In this chapter, an in-depth analysis of the current applications and barriers of aDBS in neurological and psychiatric diseases will be given. The chapter will start with principles of aDBS, followed by indications, possible biomarkers, and evidence for aDBS in a disease-specific way. Finally, future, more data-driven approaches for applying aDBS will be discussed.
  • Article
    Introduction: Deep brain stimulation (DBS) is a well-established treatment of movement disorders; but recently there has been an increasing trend towards the ablative procedure magnetic resonance-guided focused ultrasound (MRgFU). DBS is an efficient neuromodulatory technique but associated with surgical complications. MRIgFUS is an incision-free method that allows thermal lesioning, with fewer surgical complications but irreversible effects. Areas covered: We look at current and prospective aspects of both techniques. In DBS, appropriate patient selection, improvement in surgical expertise, target accuracy (preoperative and intraoperative imaging), neurophysiological recordings, and novel segmented leads need to be considered. However, increased number of older patients with higher comorbidities and risk of DBS complications (mainly intracranial hemorrhage, but also infections, hardware complications) make them not eligible for surgery. With MRgFUS, hemorrhage risks are virtually nonexistent, infection or hardware malfunction are eliminated, while irreversible side effects can appear. Expert opinion: Comparison of the efficacy and risks associated with these techniques, in combination with a growing aged population in developed countries with higher comorbidities and a preference for less invasive treatments, necessitates a review of the indications for movement disorders and the most appropriate treatment modalities.
  • Article
    Deep brain stimulation (DBS) of 3 different targets is the most important therapeutic innovation of the past 30 years for patients with fluctuating Parkinson's disease (PD), disabling dystonia, tremors, and refractory Gilles de la Tourette syndrome. When compared with medical treatment alone, controlled studies have shown better motor, nonmotor, and particularly quality‐of‐life outcomes with large effect sizes for advanced complicated PD that cannot be improved with medication, and also for PD patients with only early fluctuations. Class 1 studies have also shown superiority over medical treatment for generalized, segmental, and botulinum‐toxin refractory focal cervical dystonia. Long‐term efficacy is established for all indications with open studies. For tremors, open studies have shown that DBS is remarkably effective on PD and essential tremor, but efficacy on severe essential tremor and cerebellar tremors is limited by a tendency for tolerance/habituation, including concerns about long‐term efficacy. Open studies of disabling Gilles de la Tourette syndrome show an improvement in tics. New developments hold a promise for further improvement. New hardware with directional stimulation and new stimulation paradigms are further areas of research. The targets of DBS are refined with new imaging processing that will help to diversify the surgical targets. New indications are being explored. Closed‐loop DBS using brain or peripheral sensor signals have shown favorable clinical short‐term results. Long‐term data are lacking, and it is hoped that similar approaches for other movement or behavioral disorders may be developed. Exciting new developments carry the hope for a more pathophysiology‐based approach for DBS for various brain circuit disorders. © 2019 International Parkinson and Movement Disorder Society
  • Article
    Full-text available
    SPM is a free and open source software written in MATLAB (The MathWorks, Inc.). In addition to standard M/EEG preprocessing, we presently offer three main analysis tools: (i) statistical analysis of scalp-maps, time-frequency images, and volumetric 3D source reconstruction images based on the general linear model, with correction for multiple comparisons using random field theory; (ii) Bayesian M/EEG source reconstruction, including support for group studies, simultaneous EEG and MEG, and fMRI priors; (iii) dynamic causal modelling (DCM), an approach combining neural modelling with data analysis for which there are several variants dealing with evoked responses, steady state responses (power spectra and cross-spectra), induced responses, and phase coupling. SPM8 is integrated with the FieldTrip toolbox , making it possible for users to combine a variety of standard analysis methods with new schemes implemented in SPM and build custom analysis tools using powerful graphical user interface (GUI) and batching tools.
  • Article
    Full-text available
    Importance Collective evidence has strongly suggested that deep brain stimulation (DBS) is a promising therapy for Tourette syndrome. Objective To assess the efficacy and safety of DBS in a multinational cohort of patients with Tourette syndrome. Design, Setting, and Participants The prospective International Deep Brain Stimulation Database and Registry included 185 patients with medically refractory Tourette syndrome who underwent DBS implantation from January 1, 2012, to December 31, 2016, at 31 institutions in 10 countries worldwide. Exposures Patients with medically refractory symptoms received DBS implantation in the centromedian thalamic region (93 of 163 [57.1%]), the anterior globus pallidus internus (41 of 163 [25.2%]), the posterior globus pallidus internus (25 of 163 [15.3%]), and the anterior limb of the internal capsule (4 of 163 [2.5%]). Main Outcomes and Measures Scores on the Yale Global Tic Severity Scale and adverse events. Results The International Deep Brain Stimulation Database and Registry enrolled 185 patients (of 171 with available data, 37 females and 134 males; mean [SD] age at surgery, 29.1 [10.8] years [range, 13-58 years]). Symptoms of obsessive-compulsive disorder were present in 97 of 151 patients (64.2%) and 32 of 148 (21.6%) had a history of self-injurious behavior. The mean (SD) total Yale Global Tic Severity Scale score improved from 75.01 (18.36) at baseline to 41.19 (20.00) at 1 year after DBS implantation (P < .001). The mean (SD) motor tic subscore improved from 21.00 (3.72) at baseline to 12.97 (5.58) after 1 year (P < .001), and the mean (SD) phonic tic subscore improved from 16.82 (6.56) at baseline to 9.63 (6.99) at 1 year (P < .001). The overall adverse event rate was 35.4% (56 of 158 patients), with intracranial hemorrhage occurring in 2 patients (1.3%), infection in 4 patients with 5 events (3.2%), and lead explantation in 1 patient (0.6%). The most common stimulation-induced adverse effects were dysarthria (10 [6.3%]) and paresthesia (13 [8.2%]). Conclusions and Relevance Deep brain stimulation was associated with symptomatic improvement in patients with Tourette syndrome but also with important adverse events. A publicly available website on outcomes of DBS in patients with Tourette syndrome has been provided.
  • Article
    See Fujita and Eidelberg (doi:10.1093/brain/awx305) for a scientific commentary on this article. Focal dystonias are the most common type of isolated dystonia. Although their causative pathophysiology remains unclear, it is thought to involve abnormal functioning of the basal ganglia-thalamo-cortical circuitry. We used high-resolution research tomography with the radioligand 11C-NNC-112 to examine striatal dopamine D1 receptor function in two independent groups of patients, writer’s cramp and laryngeal dystonia, compared to healthy controls. We found that availability of dopamine D1 receptors was significantly increased in bilateral putamen by 19.6–22.5% in writer’s cramp and in right putamen and caudate nucleus by 24.6–26.8% in laryngeal dystonia (all P ≤ 0.009). This suggests hyperactivity of the direct basal ganglia pathway in focal dystonia. Our findings paralleled abnormally decreased dopaminergic function via the indirect basal ganglia pathway and decreased symptom-induced phasic striatal dopamine release in writer’s cramp and laryngeal dystonia. When examining topological distribution of dopamine D1 and D2 receptor abnormalities in these forms of dystonia, we found abnormal separation of direct and indirect pathways within the striatum, with negligible, if any, overlap between the two pathways and with the regions of phasic dopamine release. However, despite topological disorganization of dopaminergic function, alterations of dopamine D1 and D2 receptors were somatotopically localized within the striatal hand and larynx representations in writer’s cramp and laryngeal dystonia, respectively. This finding points to their direct relevance to disorder-characteristic clinical features. Increased D1 receptor availability showed significant negative correlations with dystonia duration but not its severity, likely representing a developmental endophenotype of this disorder. In conclusion, a comprehensive pathophysiological mechanism of abnormal basal ganglia function in focal dystonia is built upon upregulated dopamine D1 receptors that abnormally increase excitation of the direct pathway, downregulated dopamine D2 receptors that abnormally decrease inhibition within the indirect pathway, and weakened nigro-striatal phasic dopamine release during symptomatic task performance. Collectively, these aberrations of striatal dopaminergic function underlie imbalance between direct and indirect basal ganglia pathways and lead to abnormal thalamo-motor-cortical hyperexcitability in dystonia.
  • Article
    Full-text available
    Exaggerated basal ganglia beta activity (13-35 Hz) is commonly found in patients with Parkinson's disease and can be suppressed by dopaminergic medication, with the degree of suppression being correlated with the improvement in motor symptoms. Importantly, beta activity is not continuously elevated, but fluctuates to give beta bursts. The percentage number of longer beta bursts in a given interval is positively correlated with clinical impairment in Parkinson's disease patients. Here we determine whether the characteristics of beta bursts are dependent on dopaminergic state. Local field potentials were recorded from the subthalamic nucleus of eight Parkinson's disease patients during temporary lead externalization during surgery for deep brain stimulation. The recordings took place with the patient quietly seated following overnight withdrawal of levodopa and after administration of levodopa. Beta bursts were defined by applying a common amplitude threshold and burst characteristics were compared between the two drug conditions. The amplitude of beta bursts, indicative of the degree of local neural synchronization, progressively increased with burst duration. Treatment with levodopa limited this evolution leading to a relative increase of shorter, lower amplitude bursts. Synchronization, however, was not limited to local neural populations during bursts, but also, when such bursts were cotemporaneous across the hemispheres, was evidenced by bilateral phase synchronization. The probability of beta bursts and the proportion of cotemporaneous bursts were reduced by levodopa. The percentage number of longer beta bursts in a given interval was positively related to motor impairment, while the opposite was true for the percentage number of short duration beta bursts. Importantly, the decrease in burst duration was also correlated with the motor improvement. In conclusion, we demonstrate that long duration beta bursts are associated with an increase in local and interhemispheric synchronization. This may compromise information coding capacity and thereby motor processing. Dopaminergic activity limits this uncontrolled beta synchronization by terminating long duration beta bursts, with positive consequences on network state and motor symptoms.
  • Article
    Objectives: Our goal was to provide a detailed analysis of neurons' electrophysiological activity recorded in sub-territories of Globus pallidus internus (GPi) used as Deep Brain Stimulation (DBS) targets for these clinical conditions to potentially assist electrode targeting. Methods: We used intra-operative microelectrode recording during stereotactic neurosurgery to guide implantation of DBS lead. Results: Units in the medial anterior part of GPi of 7 Tourette's syndrome patients under general anesthesia were firing at mean and median rate of 32.1 and 21 Hz respectively (n = 101), with 45% of spikes fired during bursts and 21.3 bursts per minute. In the latero-posterior part of GPi of 7 dystonic patients under local anesthesia the mean and median activity were 46.1 and 30.6 Hz respectively (n = 27), and a mean of 21.7 bursts per minute was observed, with 30% of all spikes occurring during these bursts. Conclusion: Units activity pattern - slow-regular, fast-irregular or fast-regular were present in different proportions between the two targets. Significance: The electrophysiological characteristics of the medial-anterior part of GPi and its latero-posterior portion can be used to assist DBS electrode targeting and also support the refinement of pathophysiological models of Tourette's syndrome and Dystonia.
  • Article
    Full-text available
    Deep brain stimulation (DBS) has emerged as a promising intervention for the treatment of select movement and neuropsychiatric disorders. Current DBS therapies deliver electrical stimulation continuously and are not designed to adapt to a patient's symptoms. Continuous DBS can lead to rapid battery depletion, which necessitates frequent surgery for battery replacement. Next-generation neurostimulation devices can monitor neural signals from implanted DBS leads, where stimulation can be delivered responsively, moving the field of neuromodulation away from continuous paradigms. To this end, the authors designed and chronically implemented a responsive stimulation paradigm in a patient with medically refractory Tourette syndrome. The patient underwent implantation of a responsive neurostimulator, which is capable of responsive DBS, with bilateral leads in the centromedian-parafascicular (Cm-Pf) region of the thalamus. A spectral feature in the 5- to 15-Hz band was identified as the control signal. Clinical data collected prior to and after 12 months of responsive therapy revealed improvements from baseline scores in both Modified Rush Tic Rating Scale and Yale Global Tic Severity Scale scores (64% and 48% improvement, respectively). The effectiveness of responsive stimulation (p = 0.16) was statistically identical to that of scheduled duty cycle stimulation (p = 0.33; 2-sided Wilcoxon unpaired rank-sum t-test). Overall, responsive stimulation resulted in a 63.3% improvement in the neurostimulator's projected mean battery life. Herein, to their knowledge, the authors present the first proof of concept for responsive stimulation in a patient with Tourette syndrome.
  • Article
    Pharmacological intervention in the substantia nigra is known to induce repetitive behaviors in rodents, but a direct causal relationship between a particular neural circuit and repetitive behavior has not yet been established. Here we demonstrate that acute excitation in dopamine D1 receptor-expressing MSNs terminals in the substantia nigra pars reticulata by optogenetics resulted in sustained and chronic repetitive behaviors. These data show for the first time that activation of the striatonigral direct pathway is sufficient to generate motor stereotypies.
  • Article
    Full-text available
    • Subthalamic beta activity was recorded with an implantable DBS pulse generator over 8 months in 12 patients with Parkinson's disease. • Dopaminergic medication suppresses subthalamic beta activity at operation, 3 and 8 months after DBS. • Beta activity correlates with parkinsonian symptom severity over time. Objectives: To investigate the long term association of subthalamic beta activity with parkinsonian motor signs. Methods: We recruited 15 patients with Parkinson’s disease undergoing subthalamic DBS for local field potential recordings after electrode implantation, and at 3 and 8 months postoperatively using the implantable sensing enabled Activa PC+S (Medtronic). Three patients dropped out leaving 12 patients. Recordings were conducted ON and OFF levodopa at rest. Beta (13 – 35 Hz) peak amplitudes were extracted, compared across time points and correlated with UPDRS-III hemibody scores. Results: Peaks in the beta frequency band (13 – 35 Hz) in the OFF medication state were found in all hemispheres. Mean beta activity was significantly suppressed by levodopa at all recorded time points (P < 0.007) and individual beta power amplitude correlated with parkinsonian motor impairment across time points and dopaminergic states (pooled data; ρ = 0.25, P < 0.001). Conclusions: Our results indicate that beta-activity is correlated with parkinsonian motor signs over a time period of 8 months. Significance: Beta-activity may be a chronically detectable biomarker of symptom severity in PD that should be further evaluated under ongoing DBS.
  • Article
    Objective The clinical outcome of Parkinson's disease (PD) patients that undergo Subthalamic Nucleus Deep Brain Stimulation (STN DBS) is, in part, determined by the length of the electrode trajectory through the motor STN domain, the Dorso-Lateral Oscillatory Region (DLOR). Trajectory length has been found to correlate with the stimulation-related improvement in patients’ motor function (estimated by part III of the United Parkinson’s Disease Rating Score (UPDRS)). Therefore, it seems that ideally trajectories should have maximal DLOR length. Methods We retrospectively studied the influence of various anatomical aspects of PD patients’ brains and the geometry of trajectories planned on the length of the DLOR and STN recorded during DBS surgery. We examined 212 trajectories and 424 electrode MER tracks in 115 patients operated in our center between the years 2010 and 2015. Results We found a strong correlation between the length of the recorded DLOR and STN. Trajectories that were more lateral and/or posterior in orientation had a longer STN and DLOR pass, although the DLOR/STN fraction length (%DLOR) remained constant. The STN target was more lateral when the third ventricle was wider, and the latter correlated with older age and male gender. Conclusions Trajectory angles correlate with the recorded STN and DLOR lengths, and should be altered toward a more posterolateral angle in older patients and atrophied brains in order to compensate for the changes in STN location and geometry. These fine adjustments should yield a longer motor domain pass, thereby improving the patient's predicted outcome.