Brain Advance Access originally published online on December 22, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Brain, Vol. 127, No. 2, 408-419, 2004
© 2004 Guarantors of Brain
doi: 10.1093/brain/awh053
Subthalamic nucleus stimulation modulates motor cortex oscillatory activity in Parkinson’s disease
1 Department of Neurology, 2 Department of Clinical Neurophysiology and 3 Department of Neurosurgery, EA2683, Lille University Medical Centre, Lille, France
Correspondence to: Dr David Devos, Hôpital R. Salengro, Clinique Neurologique, CHRU, F-59037 Lille cedex, France E-mail: d-devos{at}chru-lille.fr
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
In Parkinson’s disease, impaired motor preparation has been related to an increased latency in the appearance of movement-related desynchronization (MRD) throughout the contralateral primary sensorimotor (PSM) cortex. Internal globus pallidus (GPi) stimulation improved movement desynchronization over the PSM cortex during movement execution but failed to improve impaired motor preparation. PET studies indicate that subthalamic nucleus (STN) stimulation partly reverses the abnormal premotor pattern of brain activation during movement. By monitoring MRD, we aimed to assess changes in premotor and PSM cortex oscillatory activity induced by bilateral STN stimulation and to compare these changes with those induced by L-dopa. Ten Parkinson’s disease patients and a group of healthy, age-matched controls performed self-paced wrist flexions in each of four conditions: without either stimulation or L-dopa (the ‘off’ condition), with stimulation and without L-dopa (On Stim), with L-dopa and without stimulation (‘on drug’), and with both stimulation and L-dopa (On Both). Compared with the Off condition, in both the On Stim and the On Drug condition the Unified Parkinson’s Disease Rating Scale (UPDRS) III score decreased by about 60% and in the On Both condition it decreased by 80%. The desynchronization latency over central regions contralateral to movement and the movement desynchronization over bilateral central regions were significantly increased by stimulation and by L-dopa, with a maximal effect when the two were associated. Furthermore, desynchronization latency significantly decreased over bilateral frontocentral regions in the three treatment conditions compared with the Off condition. In Parkinson’s disease, STN stimulation may induce a change in abnormal cortical oscillatory activity patterns (similar to that produced by L-dopa) by decreasing the abnormal spreading of desynchronization over frontocentral regions and increasing PSM cortex activity during movement preparation and execution, with a correlated improvement in bradykinesia. Parkinsonians under treatment displayed a desynchronization pattern close to that seen in healthy, age-matched controls, although central latencies remained shorter. The study indicates that it is possible to influence cortical reactivity related to the planning and execution of voluntary movement through the basal ganglia, and furthermore that the oscillatory activity of the PSM cortex (in addition to that of premotor areas) could be of major importance in the control of movement-associated, neural activity in Parkinson’s disease.
Key Words: Parkinson’s disease; deep brain stimulation; subthalamic nucleus; EEG; event-related desynchronization
Abbreviations: CAPIT = Core Assessment Program for Intracerebral Transplantations; GPi = globus pallidus internus; MRD = movement-related desynchronization; PSM = primary sensorimotor; STN = subthalamic nucleus; UPDRS = United Parkinson’s Disease Rating Scale
Received April 14, 2003. Revised August 14, 2003. Accepted October 10, 2003.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
The basal ganglia circuitry processes signals that flow from a range of structures including the thalami and the cortex (DeLong et al., 1990
), and constitutes an integrative system of critical importance for the planning and execution of motor behaviour (Graybiel et al., 1995). In Parkinson’s disease, dopaminergic denervation leads to an abnormal pattern of neuronal discharge in the basal ganglia and, in particular, to abnormal synchronization between the cortical and subcortical oscillatory activities. This latter phenomenon has been demonstrated in animal models of Parkinson’s disease (Tseng et al., 2001; Magill et al., 2001) and in human Parkinson’s disease using coherence analysis and movement-related frequency-specific changes in synchronization (Cassidy et al., 2002; Williams et al., 2002). Abnormal, cortico-subcortical oscillatory activity can give rise to tremor and akinesia induced by aberrant information which codes for a state of cerebral cortex activation (Brown and Marsden, 1998). The cortical changes in synchronization have been studied in more detail in human Parkinson’s disease by monitoring the movement-related desynchronization (MRD) of the mu rhythm, in the 
band (7–12 Hz). A number of investigators have demonstrated the delayed appearance of mu rhythm MRD (prior to movement onset) over the central area covering the primary sensorimotor (PSM) cortex contralateral to movement, compared with age-matched controls (Defebvre et al., 1994; Magnani et al., 1998; Wang et al., 1999). In contrast, a contralateral mu rhythm appears for nearly 2 s over the PSM cortex during movement preparation in healthy subjects, and appears bilaterally over this region during movement execution (Pfurtscheller et al., 1989; Derambure et al., 1997; Guieu et al., 1999). This decrease in desynchronization latency was considered to represent the motor programming delay observed in even the earliest stages of the disease in untreated hemiparkinsonians (Defebvre et al., 1996), and which can be partially corrected by chronic (Defebvre et al., 1998) or acute administration of L-dopa (Magnani et al., 1997; Wang et al., 1999). In advanced Parkinson’s disease with motor complications, a variety of functional, neurosurgical treatments of the basal ganglia and the thalamus have been proposed as complements to or replacements for L-dopa therapy (Limousin-Dowsey et al., 1999; Lozano and Lang, 2001). We have previously demonstrated that globus pallidus internus (GPi) stimulation decreases the abnormal spread of mu rhythm desynchronization during movement preparation over frontocentral regions and induces a selective and focal increase in PSM cortex desynchronization during movement execution. However, GPi stimulation required association with L-dopa in order to significantly improve the desynchronization latency over the PSM cortex (Devos et al., 2002). In the present study, we measured the mu rhythm MRD in order to assess whether subthalamic nucleus (STN) stimulation, which is currently preferred to GPi stimulation because of its higher clinical efficiency (Limousin et al., 1997; Krack et al., 1998; Deep Brain Stimulation Study Group, 2001), allows the restoration of normal, PSM cortex motor programming. The action of STN stimulation on the basal ganglia circuitry and its relation to motor symptom improvement remain largely unknown. There is a wealth of data suggesting that the STN plays a key role in the control of movement by exerting an excitatory influence on the output structures of basal ganglia, i.e. the pars reticulata substantia nigra and the GPi (Parent and Hazrati, 1995). It has been suggested that stimulation of the hyperactive STN increases basal ganglia-thalamic outflow, thus increasing cortex activation during voluntary movement (Obeso et al., 2000). Most of the STN stimulation studies to date have been based on a hypothesis which stresses the critical importance of the premotor areas in the impairment of movement control in Parkinson’s disease, and have thus tended to focus on these regions. This type of research has revealed either an increase in blood flow by PET (Limousin et al., 1997; Ceballos-Baumann et al., 1999; Brooks et al., 2000) or an increased amplitude of the contingent negative variation (Gerschlager et al., 1999) and the frontal N30 component of somatosensory-evoked potentials (Pierantozzi et al., 1999). However, the effect of STN stimulation on the primary sensorimotor (PSM) cortex during movement has not been established, and thus could be reliably studied by movement-related changes in synchronization (Pfurtscheller and Berghold, 1989; Derambure et al., 1997; Guieu et al., 1999). The aim of the present study was to compare the influence of STN stimulation and L-dopa on PSM and premotor cortex oscillatory activity during movement preparation and execution in patients with severe Parkinson’s disease by using MRD. In order to assess whether these treatments normalize the pattern of movement-related desynchronization, we compared Parkinson’s disease patients with healthy controls. We focused on spectral analysis of the mu rhythm, a delay in the appearance of which is well established in Parkinson’s disease and which is thought to reflect the degree of cortical activation during movement phases.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Subjects
We studied 10 consecutive Parkinson’s disease patients (Gibb and Lees, 1988
) who were undergoing treatment with bilateral STN stimulation (with quadripolar electrodes) in order to control L-dopa-related motor complications together with 10 healthy, age-matched controls [median age 66 years (range 56–69 years); four males and six females]. All patients and controls were right-handed. Our study was carried out 12 months after implantation of bilateral STN stimulation electrodes. The characteristics of the patients are summarized in Table 1. All patients received L-dopa treatment: the daily, median L-dopa dose equivalent was 745 mg after 1 year of STN stimulation. Patients 2, 4, 5 and 8 received ropinirole and patients 3 and 6 took bromocriptine. None of the other patients received agonist treatment.
|
In order to avoid circadian variations in the motor state, clinical evaluations and electroencephalogram recordings were performed during the same morning and in four sets of conditions for each patient, in the following order.
(i) Without any stimulation or L-dopa for at least the previous 15 h; the condition was designated ‘off’ and was necessarily the first condition in the morning after a night of treatment withdrawal.
(ii) With stimulation and without L-dopa (designated On Stim).
(iii) Without stimulation and after an acute administration of L-dopa (designated On Stim), where the L-dopa dose was the usual first morning dose used by each patient to relieve their symptoms prior to the date of surgery. The On Drug condition was initiated 2 h after switching off the STN stimulation. Indeed, Temperli et al. (2003) found that 90% of the improvement in the United Parkinson’s Disease Rating Scale (UPDRS) motor score was lost in the 2 h following discontinuation of STN stimulation (with 10% related to axial signs). Before starting the on-drug recordings, we checked that UPDRS motor scores were indeed back at the baseline after 2 h.
(iv) With stimulation and after an acute administration of L-dopa (designated On Both). The On Both condition was defined as the best parkinsonian triad control without (or at most with mild) dyskinesias, according to the CAPIT (Core Assessment Program for Intracerebral Transplantations) committee’s criteria (Langston et al., 1992). The two conditions involving L-dopa were performed last because the end of the L-dopa effect cannot be assessed with certainty: earlier L-dopa use would have entailed a longer time delay before initiating the following condition. The L-dopa dose was the same in the On Drug and On Both conditions (Table 1). Definition of the best motor control was based on a long-term survey (over several months) performed in our neurological department prior to the study, in order to assess the best stimulation settings and L-dopa doses (Table 1). For each condition, clinical evaluation was carried out using the UPDRS motor ratings (Fahn et al., 1987) and its bradykinesia subscore (items 23, 24, 25) for the upper limb performing the movements. The CAPIT tapping test was also analysed (Langston et al., 1992).
Neurosurgical procedure
The neurosurgical procedure was performed in the Neurosurgery Department of the Lille University Medical Centre, and was similar to that previously described by Benabid et al. (2000). The procedure was performed in a single session under general anaesthesia. Antiparkinsonian medication was stopped the evening before surgery. The patient was placed in a Talairach stereotactic frame. Next, a double contrast ventriculography was performed: this allowed definition of the anterior commissure, the posterior commissure and the midline of the third ventricle, which served as references for determination of the target coordinates according to Guiot’s geometrical diagram. The target was defined according to the stereotactic coordinates which correspond to the STN area in standard reference atlases. Quadripolar electrodes (Medtronic, Minneapolis, MN, USA) were then implanted using a control stylet into each of the two STN via a frontal anterior double obliquity, and were adjusted as a function of intraoperative microrecordings and stimulation (the patients were awake throughout this latter procedure). The electrodes (with four contacts, contact 0 ventral to contact 3 dorsal, each measuring 1.5 mm in length and 0.5 mm apart) were initially connected to an external pulse generator (Extrel; Medtronic) for stimulation testing. A few days later, programmable pulse generators (Itrel II; Medtronic) were implanted into the subclavicular region and connected to the electrodes. The electrode contacts were selected and the electrical parameters (pulse width, voltage and frequency) were optimized by telemetry in order to provide the best relief of parkinsonian symptoms in the ‘on’ condition and with the fewest possible side effects in the ‘off’ condition (Table 1). No adverse effects of bilateral STN stimulation were noticed. Patients and controls gave their written, informed consent and the study was approved by the Local Ethics Committee of Lille.
Experimental procedure
The study was carried out using bipolar STN stimulation in order to avoid artefacts induced by unipolar systems. The methods for recording MRD were the same as those employed in our previous studies (Derambure et al., 1993; Defebvre et al., 1994, 1996, 1998; Guieu et al., 1999; Devos et al., 2002). During EEG recording, the subject sat comfortably in an armchair and stared at a fixed point. The arm was held to the side, with the forearm resting on the armrest and each hand in a sheath. The wrists were straightened in an intermediate prosupination position (thumb on top) and the fingers were held in a straight, resting position. Other than wrist flexion, no movement of the studied limb was possible. During each recording, patients were instructed to perform the same unilateral, brisk, self-paced, 45° wrist flexion movements with the more akinetic hand (Table 1), under the four sets of conditions described above. Control subjects performed the same movement in a single session and with the same (right) hand. The movement was executed while maintaining the other arm at rest. Correct execution of movement was controlled by an online video display. Each movement was repeated with a variable time interval of at least 10 s (time intervals of less than 10 s were eliminated), followed by a slow return to the resting position. The movement amplitude was limited by two abutments. For each condition, we recorded (over a 20 min period) at least 50 EEG epochs lasting 10 s surrounding the movement and free of artefacts. The patients had a break of 45 min between each session in order to limit the effect of fatigue and to prepare the following condition, with the exception of the interval between the second and the third conditions, where a 2 h break was accorded.
Data recording
Bipolar EMGs were recorded using surface electrodes placed on the flexor carpi radialis. EEG activity was recorded from 37 scalp electrodes located according to the international 10–20 system, referenced to a prefrontal ground placed in front of AFz. Impedances were kept below 5 k
. EMG and EEG signals were sampled at 256 Hz and amplified with a time constant of 0.03 s for EMG and 0.3 s for EEG signals, together with a low-pass filter fixed at 128 Hz.
Data processing
The same EEG electrodes were retained for all four sets of recordings, resulting in the same spatial conditions. MRD was computed from 21source derivations covering the frontal (F) [F1, Fz, F2], frontocentral (FC) [FC1, FC3, FCz, FC2, FC4], central (C) [C1, C3, Cz, C2, C4], parietocentral (CP) [CP1, CP3, CPz, CP2, CP4] and parietal (P) [P1, Pz, P2] areas. The location of the source derivations was chosen so as to explore the left and right premotor and PSM areas. Because the movements were performed with either the right or the left hand, i.e. corresponding to the more akinetic hand of each patient (Table 1), we labelled source derivations in each area which were contralateral to the movements with the letter ‘c’ and those that were ipsilateral with the letter ‘i’. For further analysis, each of the areas defined above was divided into contralateral and ipsilateral regions: frontocentral (FCc and FCi), central (Cc and Ci) and centroparietal (CPc and CPi).
Movement onset and duration were determined from the rectified, electromyographic activity. EMG burst duration was evaluated as the time interval ranging from the onset of the EMG burst to a point after the maximum at which the voltage corresponded to 5% of the peak voltage. By using this measure of EMG duration, it was possible to avoid the imprecision of the EMG burst end.
MRD was analysed from 5 s before the onset of movement to 5 s after, and was computed in a narrow frequency band for mu rhythm (peak ±1 Hz) using a sixth-order digital FIR filter (Pfurtscheller et al., 1989; Guieu et al., 1999). The mu rhythm was established for each subject by comparing the power spectral densities of the EEG at electrodes C3/C1 or C2/C4 during the interval prior to movement onset (0 to –2 s) and during the reference interval (–3.5 to – 4.5 s). The samples were squared and averaged over all trials to obtain the mean power change time-course. To obtain a normalized measure of MRD, the power decrease was expressed as a percentage of a reference value computed within the time interval ranging from 4.5 to 3.5 s before EMG onset. To reduce the variance, temporal resolution was diminished so as to obtain one power value every 125 ms. The significance of the differences between mean powers observed during the reference period and those measured during subsequent 125 ms intervals was also expressed as a probability value (P) using non-parametric statistics (Wilcoxon’s signed rank test).
Data analysis
The following parameters were computed for each subject and each condition.
(i) The desynchronization latency prior to the movement onset (expressed in s) was measured for the first of three consecutive significant values (P < 0.01). For each contralateral and ipsilateral region, the source derivation showing the earliest MRD was selected.
(ii) Movement desynchronization was defined as the sum of the significant desynchronization values (cumulated percentages) for each region during execution of the movement. The results for both source derivations of each contralateral and ipsilateral region were pooled. In order to take into account the fact that EMG duration varied between subjects and conditions, and to avoid an effect of movement duration on this measure, we divided the movement desynchronization by the EMG duration to allow expression per time unit (%/s). This measure represented each region’s degree of cortical activation during the execution phase.
Statistical analysis
In order to assess the effects of STN stimulation and L-dopa on clinical scores, EMG duration, latencies and movement desynchronization, we performed an analysis of variance for repeated measures with two factors: STN stimulation (On, Off) and L-dopa (with, without). The second objective was to compare the latencies and movement desynchronization over each region pair-wise for the four conditions, by using a one-way analysis of variance with the factor condition (Off, On Drug, On Stim, On Both). In the light of the number of subjects, Conover’s non-parametric test was performed for both analyses (Conover and Imam, 1981). Principal effects (as well as interactions between factors) were analysed by contrast studies. According to the repeated measures, the Greenhouse–Geisser epsilon correction of degrees of freedom was applied if required. We computed Spearman correlations between latencies, movement desynchronizations and clinical parameters. The third objective was to compare the three treatment conditions in Parkinson’s disease patients and controls, using the Mann–Whitney U test. A significance level of 0.05 was chosen. We used SPSS 9.0® software for statistical analysis.
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
EMG activity of the flexor carpi radialis
A principal, significant effect of STN stimulation on EMG duration was demonstrated [F(1,9) = 27.78, P = 0.001], showing a shortening of EMG duration under STN stimulation with or without L-dopa when compared with conditions lacking STN stimulation with or without L-dopa. A trend towards shorter EMG durations was also noted under L-dopa, both with and without STN stimulation (Table 2). EMG durations did not differ significantly between controls and the On Both, On Stim and On Drug conditions.
|
Spectral analysis of the mu rhythm peak
For each patient, the same peak was observed in each of the four conditions. STN stimulation thus did not influence the cortical mu rhythm’s peak, allowing us to use the same mu rhythm band in the MRD analysis in all of the four conditions for each patient. The peaks observed in the normal controls were in the same range as those of the patients, i.e. between 8 and 12 Hz.
Effect of STN stimulation and L-dopa therapy
Principal, significant effects of STN stimulation [F(1,9) = 156.13, P = 0.0001] and of L-dopa [F(1,9) = 197.81, P = 0.0001] on the bradykinesia subscore were observed, revealing an improvement under STN stimulation with or without L-dopa and under L-dopa with or without STN stimulation. There was no significant interaction between the two treatments; the degree of improvement under STN stimulation was similar to that seen with L-dopa. The same statistical effects were obtained for the UPDRS motor ratings, the CAPIT tapping test scores, the latencies over ipsilateral and contralateral frontocentral and contralateral central regions and the movement desynchronization over ipsilateral and contralateral central regions (Table 2). All interactions were insignificant except for UPDRS motor rating, for which L-dopa associated with STN stimulation displayed a lesser effect than L-dopa alone. We also observed a trend of latency changes over ipsilateral central regions but it remained insignificant (Table 2).
The mu rhythm MRD patterns
In normal controls, the initial desynchronization appeared first over the contralateral central region, then over the contralateral centroparietal region and the ipsilateral central region before movement onset and finally over ipsilateral centroparietal, frontocentral and parietal regions at movement onset, followed by a major bilateral central desynchronization during movement (Figs 1 and 2).
|
|
In the Off condition for the Parkinson’s disease patients, desynchronization was first observed over contralateral and ipsilateral frontocentral regions and next over central and centroparietal regions. In contrast, the On Drug, On Stim and On Both conditions seemed to share a pattern close to that of normal controls, except for earlier desynchronization over contralateral and ipsilateral central derivations (Figs 1 and 2). For Parkinson’s disease patients, the earliest contralateral central latency was observed in the On Both condition.
These results were confirmed by the one-way Conover analysis, which revealed a principal significant effect of the condition on latency over the contralateral frontocentral region [F(3,27) = 4.36, P = 0.033], the ipsilateral frontocentral region [F(3,27) = 21.3, P = 0.0001] and the contralateral central region [F(3,7) = 7.77, P = 0.012], and on movement desynchronization over contralateral [F(3,27) = 93.97, P = 0.0001] and ipsilateral central [F(3,27) = 6.18, P = 0.002] regions. Contrast studies showed a decrease in bilateral frontocentral latency, an increase in contralateral central latency and an increase in bilateral central movement desynchronization in the On Stim, On Drug and On Both conditions compared with the Off condition (Figs 3 and 4, Table 2). No significant differences appeared when comparing On Stim and On Drug conditions, whereas significant increases in contralateral central latency and movement desynchronization were noted when comparing On Both and both the On Stim and On Drug conditions (Figs 3 and 4, Table 2). The frontal, median frontocentral (FCz and Cz) and parietal regions did not display any desynchronization before movement onset in Parkinson’s disease patients.
|
|
The contralateral and ipsilateral central latencies were significantly greater in controls than in Parkinson’s disease patients in the On Drug (U = 4, P = 0.0005 and U = 11, P = 0.003 respectively), On Stim (U = 9, P = 0.002 and U = 6.5, P = 0.001 respectively) and On Both conditions (U = 18, P = 0.015 and U = 9, P = 0.002 respectively), whereas desynchronization latencies over frontocentral, centroparietal and parietal regions were not significantly different. For all regions, movement desynchronization did not significantly differ between the On Both condition and controls.
Relation between cortical oscillatory activity changes and clinical improvement
Latencies over the contralateral, central region seemed to be inversely correlated (10 patients with four conditions per patient) with the UPDRS motor rating (r = – 0.59, P < 0.001) (Fig. 5) and the bradykinesia subscore (r = –0.62, P < 0.001). The central, contralateral movement desynchronization also seemed to be inversely correlated with the UPDRS motor rating (r = –0.66, P < 0.001) (Fig. 5), the bradykinesia subscore (r = –0.64, P < 0.001) and the CAPIT tapping test score (r = –0.51, P < 0.001).
|
In each condition, the EMG duration appeared to be correlated with the bradykinesia subscore (r = +0.57, P < 0.001) and the UPDRS motor rating (r = +0.58, P < 0.001).
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
During movement preparation and in the absence of STN stimulation or L-dopa, our patients with advanced Parkinson’s disease displayed abnormal cortical oscillatory activity patterns with very reduced and delayed desynchronization over the contralateral central region with, in contrast, increased desynchronization over bilateral frontocentral regions. STN stimulation improved motor symptoms by 58% and partially restored the normal pattern of cortical activation by (i) increasing contralateral PSM cortex activation during movement preparation and execution and (ii) decreasing the spread of desynchronization prior to movement onset over the bilateral frontocentral regions. The effect of STN stimulation on cortical mu rhythm desynchronization during movement preparation and execution appeared to be the same as that of acute administration of L-dopa. PSM cortex oscillatory reactivity changes seemed to be correlated with clinical improvement.
Effects of STN stimulation
STN stimulation alone improved the bradykinesia subscore by about 47% and the UPDRS motor score by 58%, which is similar to previous studies, in which the UPDRS motor score decrease ranged from about 45% (41–51%) (Ceballos-Baumann et al., 1999; Moro et al., 1999; Pinter et al., 1999; Deep Brain Stimulation Study Group, 2001; Simuni et al., 2002; Vingerhoets et al., 2002) to about 60% (58–74%) (Limousin et al., 1997; Krack et al., 1998; Kumar et al., 1998; Houeto et al., 2000; Rodriguez-Oroz et al., 2000; Molinuevo et al., 2000; Figueiras-Méndez et al., 2002; Thobois et al., 2002).
Equally, STN stimulation appeared to influence the abnormal pattern of cortical oscillatory reactivity to movement observed in advanced Parkinson’s disease by increasing the desynchronization over the contralateral central derivations during movement preparation and execution, which in turn leads to the establishment of a near-normal pattern of cortical oscillatory activity when compared with previous studies in healthy subjects (Guieu et al., 1999) and to the results from our control group. However, compared with healthy subjects, the central latencies remained shorter under STN stimulation with or without L-dopa (Figs 1 and 2).
Previous studies with electrocorticographic (Crone et al., 1998) and intracerebral recordings (Szurhaj et al., 2003) have shown that the signals recorded by the central derivations measured on the scalp do accurately represent the activity of the PSM cortex in the anterior bank of the central sulcus. STN stimulation may therefore partly restore PSM cortex activation during movement in advanced Parkinson’s disease. The original concept of the basal ganglia model seems to explain our results clearly, since cortical hypoactivation was considered a major feature of bradykinesia in the parkinsonian state (Obeso et al., 2000). The dynamic organization of oscillatory activities into different frequency bands between basal ganglia and the motor cortex was first suggested by Brown and Marsden (1998) and was subsequently shown in human Parkinson’s disease via the coherence between EEG and local potentials from STN and GPi (Cassidy et al., 2002; Williams et al., 2002) and in animal models of Parkinson’s disease (Magill et al., 2001; Tseng et al., 2001). The correlations between bradykinesia and mu rhythm MRD parameters over the PSM cortex suggest that this structure could also be of major importance for the control of neural activity associated with movement, in addition to the much-studied, non-primary motor cortical areas (Limousin et al., 1997; Ceballos-Baumann et al., 1999; Gerschlager et al., 1999; Pierantozzi et al., 1999; Obeso et al., 2000). Paradoxically, STN stimulation at rest decreases the motor cortex blood flow in PET studies (Limousin et al., 1997; Ceballos-Baumann et al., 1999) and increases cortical inhibition in transcranial magnetic stimulation experiments (Cunic et al., 2002). The PSM cortex activity decrease at rest under STN stimulation was thought to be related to improvements in rigidity and off-dystonia (Limousin et al., 1997; Ceballos-Baumann et al., 1999), whereas the increases in PSM cortex activity during movement observed in our study may be related to improvements in bradykinesia. Thus, these apparently opposite effects of STN stimulation on the PSM cortex activity may in fact depend on the condition of rest or movement.
Performance of the four non-randomized conditions during the same morning may induce a possible effect of fatigue on the results. However, in view of the large differences in mu rhythm desynchronization between the various conditions and the fact that the best results were obtained during the On Both condition (which was the last of the four), we suggest that this potential effect did not significantly interfere with our findings. Moreover, the relatively short recording periods were separated by long rest periods.
Movement duration was clearly decreased by STN stimulation and tended to decrease under L-dopa therapy, even though patients were instructed to perform the same brisk, self-paced 45° wrist flexion movements in the four conditions. One could legitimately ask whether these modifications might induce variations of mu MRD latency during movement preparation. However, it has been demonstrated that brisk versus slow finger movements (Stancak and Pfurtscheller, 1996) and brisk versus sustained wrist movements (Cassim et al., 2000) do not induce any significant changes in mu rhythm desynchronization before movement. Thus, in our study, movement duration changes can justifiably be eliminated as an independent variable determining the cortical activation pattern prior to movement, and should rather be attributed to bradykinesia improvement, as suggested by the correlation between EMG duration and the bradykinesia subscore. However, in other frequency bands, such as the gamma and beta bands, (de)synchronization differences over the motor cortex during phasic versus tonic movement execution may exist (Brown et al., 1998).
Furthermore, the percentage of mu rhythm MRD during movement also increased under STN stimulation compared with the Off condition. This could be considered as another parameter for improvement of oscillatory reactivity in the PSM cortex. However, this could also be related to variation in afferent feedback as a function of the decreased movement duration, following the improvement in bradykinesia under STN stimulation with or without L-dopa.
STN stimulation reduced the latency desynchronization over the bilateral frontocentral regions, the latter probably corresponding to the premotor cortex. This frontocentral spread of desynchronization in the absence of any treatment was not observed in our normal controls and has also never been noted in mild Parkinson’s disease under the same movement and analysis conditions (Defebvre et al., 1994, 1996): however, it was noticed previously in advanced Parkinson’s disease under conditions similar to our Off situation (Devos et al., 2002). This abnormal cortical activity could reflect impaired activation, as suggested by the reduced amplitude of the contingent negative variation (Gerschlager et al., 1999) and the decrease in premotor cortex blood flow (Limousin et al., 1997; Ceballos-Baumann et al., 1999; Brooks and Samuel, 2000). However, we should bear in mind that the PET scan technique and MRD each reveal different phenomena and possess different time and spatial resolutions. The supplementary motor area and cingular cortex have an internal frontal topography which is not detected with precision by the scalp recordings in MRD, and PET studies are unable to detect the rapid changes that occur during preparation and execution phases.
Comparison of STN stimulation and L-dopa effects
In mildly impaired Parkinson’s disease patients, desynchronization latency was significantly increased over the contralateral PSM cortex after chronic (Defebvre et al., 1998) and acute administration of L-dopa with either finger movement (Magnani et al., 1997) or simple and sequential hand and elbow movements (Wang et al., 1999). Here, in advanced Parkinson’s disease, we observed the same improvement in desynchronization parameters during the preparation and execution phases. The lack of any significant difference in desynchronization parameters between STN stimulation and acute administration of L-dopa suggested the same effect on the cortical oscillatory activity changes during preparation and execution of a simple, self-initiated, distal movement. They seemed to induce the same partial restoration of normal pattern. Using transcranial magnetic stimulation, STN stimulation displayed the same effect on short-interval intracortical inhibition as pharmacological dopaminergic therapy (Cunic et al., 2002; Pierantozzi et al., 2002). However, it has no effect on the silent period duration and long-interval intracortical inhibition, which are also influenced by dopaminergic drugs (Cunic et al., 2002). Furthermore, many other factors—for example, the difficult STN targeting and the age and dopa sensitivity of the patients (Welter et al., 2002)—could influence STN stimulation efficiency. Hence, mechanisms of improvement should always be considered in the light of the clinical results. In our study, all patients were highly dopa-sensitive, with a 59% mean decrease in UPDRS score. The rate of UPDRS decrease under STN stimulation was also close to the mean dopa sensitivity observed in previous studies (Krack et al., 1998; Deep Brain Stimulation Study Group, 2001; Welter et al., 2002).
During the On Both condition, we observed a mild but significant improvement of clinical scores, latency and movement desynchronization over the PSM cortex compared with the On Stim and On Drug conditions. The complementary effect of STN stimulation and L-dopa suggested two different potential mechanisms of action: L-dopa acts on the nigro-striatal dopaminergic projections (i.e. allowing restoration of inhibitory input to the striatum) (Gibb, 1997), whereas STN stimulation appears to modulate STN hyperactivity through the STN or near the STN by the spread of current to the large fibre bundles via either functional blockade or postsynaptic facilitation (Ashby et al., 1999). STN stimulation prevents the transfer of certain neuronal information, particularly the abnormal firing pattern related to parkinsonian signs (Beurrier et al., 2001; Hashimoto et al., 2003).
Comparison of the effects of STN and GPi stimulation
Like STN stimulation, GPi stimulation reduced the abnormal spread of desynchronization over frontocentral regions and increased desynchronization during movement over PSM cortex, whereas, unlike STN stimulation and l-dopa, GPi stimulation failed to increase PSM cortex activity during the preparation phase when the same method and movement paradigm were used (Devos et al., 2002). GPi stimulation seemed to induce a more restricted temporal effect on the PSM cortex during movement. The lack of an enhancing cortical effect of GPi stimulation during the preparation phase of a distal movement was also found in PET studies, which either failed to reveal any modification (Limousin et al., 1997) or indicated only a significant increase in ipsilateral supplementary motor area activity and a decrease in activity of the contralateral cingulate motor area (Fukuda et al., 2001). Furthermore, the association of GPi stimulation and l-dopa also led to a more limited temporal and spatial improvement of the PSM cortex oscillatory activity than did the association of l-dopa and STN stimulation, the only significant increase in activity being that of Cc derivations during the preparation phase. Both STN and GPi stimulation studies displayed good concordance between PSM cortex oscillatory activity and the clinical outcomes (Devos et al., 2002). The lesser efficiency of GPI stimulation versus STN stimulation has also previously been observed clinically (Limousin et al., 1997; Krack et al., 1998; Deep Brain Stimulation Study Group, 2001; Krause et al., 2001). However, in another study, both STN and GPi stimulation induced the same improvement in bradykinesia (Brown et al., 1999). Thus, in addition to hypothetical differences in the ‘dynamic organization’ of the basal ganglia, this observation principally suggests dissimilarities in the electrode placement relative to the higher volume of stimulation of the pallidum and its functional, topographic differences. Indeed, contrasting anti-akinetic or anti-dyskinetic effects (Bejjani et al., 1997; Krack et al., 1998 a, b; Yelnik et al., 2000) have been observed when the electrode was applied to the external or the internal pallidum, respectively.
Cortical oscillatory activity patterns in Parkinson’s disease
Comparison of our results with previous work using the same movement paradigm suggests that the desynchronization latency over the PSM cortex decreases with disease severity and clinical disability (Defebvre et al., 1996, 1998). Indeed, in the Off condition, the latency value was 380 ms for patients with a disease duration of 14 years and a UPDRS motor score of 45, whereas it was 1250 ms in de novo untreated hemiparkinsonian patients with a disease duration of 1.65 years and a UPDRS motor rating of 19 (Defebvre et al., 1996), and has ranged from 1500 to 2000 ms in normal subjects in previous studies (Guieu et al., 1999) and in the present study. Moreover, desynchronization parameters in general (and latencies in particular) seemed to be correlated with bradykinesia scores. Consequently, our observation of decreased desynchronization over the PSM cortex might well support the hypothesis of cortical hypoactivation, increasing as the disease worsens and thus contributing to bradykinesia. The decrease in cortical desynchronization observed during movement phases might be explained by an abnormal state of hypersynchronized thalamocortical activity in Parkinson’s disease (Brown and Marsden, 1998). Indeed, a pronounced tendency towards synchronization at a frequency below 20 Hz has been demonstrated in the STN of Parkinsonian patients (Levy et al., 2000): pathologically exaggerated movement-related synchronization was also recorded in the STN, the GPi and the motor cortex of untreated parkinsonians at frequencies of 20 Hz (Cassidy et al., 2002). Our results suggest that the movement-related, frequency-specific decrease in desynchronization recorded at the cortical level may concern the mu rhythm band below 20 Hz and may be exacerbated with worsening disease.
In conclusion, our study displayed a partial restoration of cortical oscillatory activity patterns in Parkinson’s disease by STN stimulation and showed that it is possible to influence movement-related cortical reactivity through the basal ganglia. We intend to analyse the relationship between movement-related oscillatory reactivity in both the PSM cortex and the STN by recording both structures in patients immediately after surgery.
|
Acknowledgements |
|---|
We wish to thank David Fraser for his help. This work was supported by PHRC grants from the French Ministry of Health and by the France Parkinson charitable association.
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Ashby P, Kim YJ, Kumar R, Lang AE, Lozano AM. Neurophysiological effects of stimulation through electrodes in the human subthalamic nucleus. Brain 1999; 122: 1919–31.
Bejjani B, Damier P, Arnulf I, Bonnet AM, Vidailhet M, Dormont D, et al. Pallidal stimulation for Parkinson’s disease. Two targets? Neurology 1997; 49: 1564–9.
Benabid AL, Krack P, Benazzouz A, Limousin P, Koudsie A, Pollak P. Deep brain stimulation of the subthalamic nucleus for Parkinson’s disease: methodologic aspects and clinical criteria. Neurology 2000; 55 (12 Suppl 6): S40–S44.[ISI][Medline]
Beurrier C, Bioulac B, Audin J, Hammond C. High-frequency stimulation produces a transient blockade of voltage-gated currents in subthalamic neurons. J Neurophysiol 2001; 85: 1351–6.
Brooks DJ, Samuel M. The effects of surgical treatment of Parkinson’s disease on brain function: PET findings. [Review]. Neurology 2000; 55 (12 Suppl 6): S52–9.[Medline]
Brown P, Marsden CD. What do the basal ganglia do? [Review]. Lancet 1998; 351: 1801–4.[CrossRef][ISI][Medline]
Brown P, Salenius S, Rothwell JC, Hari R. Cortical correlate of the Piper rhythm in humans. J Neurophysiol 1998; 80: 2911–7.
Brown RG, Dowsey PL, Brown P, Jahanshahi M, Pollak P, Benabid AL, et al. Impact of deep brain stimulation on upper limb akinesia in Parkinson’s disease. Ann Neurol 1999; 45: 473–88.[CrossRef][ISI][Medline]
Cassidy M, Mazzone P, Oliviero A, Insola A, Tonali P, Di Lazzaro V, et al. Movement-related changes in synchronization in the human basal ganglia. Brain 2002; 125: 1235–46.
Cassim FC, Szurhaj W, Sediri H, Devos D, Bourriez JL, Poirot I, et al. Brief and sustained movements: differences in event-related (de)synchronization (ERD/ERS) patterns. Clin Neurophysiol 2000; 111: 2032–9.[Medline]
Ceballos-Baumann AO, Boecker H, Bartenstein P, von Falkenhayn I, Riescher H, Conrad B, et al. A positron emission tomographic study of subthalamic nucleus stimulation in Parkinson disease: enhanced movement-related activity of motor-association cortex and decreased motor cortex resting activity. Arch Neurol 1999; 56: 997–1003.
Conover WJ, Iman RL. Rank transformation as a bridge between parametric and nonparametric statistics. Am Statistician 1981; 35: 124–9.[CrossRef]
Crone NE, Miglioretti DL, Gordon B, Sieracki JM, Wilson MT, Uematsu S, et al. Functional mapping of human sensorimotor cortex with electrocorticographic spectral analysis. I. Alpha and beta event-related desynchronization. Brain 1988; 121: 2271–99.
Cunic D, Roshan L, Khan FI, Lozano AM, Lang AE, Chen R. Effects of subthalamic nucleus stimulation on motor cortex excitability in Parkinson’s disease. Neurology 2002; 58: 1665–72.
Deep Brain Stimulation for Parkinson’s Disease Study Group. Deep-brain stimulation of the subthalamic nucleus or the pars interna of the globus pallidus in Parkinson’s disease. New Engl J Med 2001; 345: 956–63.
Defebvre L, Bourriez JL, Dujardin K, Derambure P, Destée A, Guieu JD. Spatiotemporal study of Bereitschaftspotential event-related desynchronization during voluntary movement in Parkinson’s disease. Brain Topogr 1994; 6: 237–44.[Medline]
Defebvre L, Bourriez JL, Destée A, Guieu JD. Movement-related desynchronisation pattern preceding voluntary movement in untreated Parkinson’s disease. J Neurol Neurosurg Psychiatry 1996; 60: 307–12.[Abstract]
Defebvre L, Bourriez JL, Derambure P, Duhamel A, Guieu JD, Destée A. Influence of chronic administration of L-DOPA on event-related desynchronization of mu rhythm preceding voluntary movement in Parkinson’s disease. Electroencephalogr Clin Neurophysiol 1998; 109: 161–7.[CrossRef][Medline]
DeLong MR. Primate models of movement disorders of basal ganglia origin. Trends Neurosci 1990; 13: 281–5.[CrossRef][ISI][Medline]
Derambure P, Defebvre L, Dujardin K, Bourriez JL, Jacquesson JM, Destée A, et al. Effect of aging on the spatio-temporal pattern of event-related desynchronization during a voluntary movement. Electroencephalogr Clin Neurophysiol 1993; 89: 197–203.[CrossRef][ISI][Medline]
Derambure P, Bourriez JL, Defebvre L, Cassim F, Josien E, Duhamel A, et al. Abnormal cortical activation during planning of voluntary movement in patients with epilepsy focal motor seizures: event-related desynchronization study of electroencephalographic mu rhythm. Epilepsia 1997; 38: 655–62.[Medline]
Devos D, Derambure P, Bourriez JL, Cassim DF, Blond S, Guieu JD, et al. Influence of internal globus pallidus stimulation on motor cortex activation pattern in Parkinson’s disease. Clin Neurophysiol 2002; 113: 1110–20.[Medline]
Fahn S, Elton RL and members of the UPDRS Development Committee. Unified Parkinson’s Disease Rating Scale. In: Fahn S, Marsden CD, Caln DB, Goldstein M, editors. Recent developments in Parkinson’s disease. Florham Park (NJ): MacMillan Healthcare Information; 1987. p. 153–63.
Figueiras-Méndez R, Regidor I, Riva-Meana C, Magarinos-Ascone CM. Further supporting evidence of beneficial subthalamic stimulation in Parkinson’s patients. Neurology 2002; 58: 469–70.
Fukuda M, Mentis M, Ghilardi MF, Dhawan V, Antonini A, Hammerstad J, et al. Functional correlates of pallidal stimulation for Parkinson’s disease. Ann Neurol 2001; 49: 155–64.[CrossRef][ISI][Medline]
Gerschlager W, Alesch F, Cunnington R, Deecke L, Dirnberger G, Endl W, et al. Bilateral subthalamic nucleus stimulation improves frontal cortex function in Parkinson’s disease. An electrophysiological study of the contingent negative variation. Brain 1999; 122: 2365–73.
Gibb WR. Functional neuropathology in Parkinson’s disease. Eur Neurol 1997; 38 Suppl 2: 21–5.
Gibb WRG, Lees AJ. The relevance of the Lewy body to the pathogenesis of idiopathic Parkinson’s disease. J Neurol Neurosurg Psychiatry 1988; 51: 745–52.[Abstract]
Graybiel AM. Building action repertoires: memory and learning functions of the basal ganglia. [Review]. Curr Opin Neurobiol 1995; 5: 733–41.[CrossRef][ISI][Medline]
Guieu JD, Bourriez JL, Derambure P, Defebvre L, Cassim F. Temporal and spatial aspects of event-related desynchronization and movement-related cortical potential. In: Pfurtscheller G, Lopes da Silva FH, editors. Event-related desynchronization. Handbook of electroencephalography and clinical neurophysiology. Revised series, Vol. 6. Amsterdam: Elsevier; 1999. p. 279–90.
Hashimoto T, Elder CM, Okun MS, Patrick SK, Vitek JL. Stimulation of the subthalamic nucleus changes the firing pattern of pallidal neurons. J Neurosci 2003; 23: 1916–23.
Houeto JL, Damier P, Bejjani PB, Staedler C, Bonnet AM, Arnulf I, et al. Subthalamic stimulation in Parkinson disease: a multidisciplinary approach. Arch Neurol 2000; 57: 461–5.
Krack P, Pollak P, Limousin P, Hoffmann D, Xie J, Benazzouz A, et al. Subthalamic nucleus or internal pallidal stimulation in young onset Parkinson’s disease. Brain 1998; 121: 451–7.
Krause M, Fogel W, Heck A, Hacke W, Bonsanto M, Trenkwalder C, et al. Deep brain stimulation for the treatment of Parkinson’s disease: subthalamic nucleus versus globus pallidus internus. J Neurol Neurosurg Psychiatry 2001; 70: 464–70.
Kumar R, Lozano AM, Kim YJ, Hutchison WD, Sime E, Halket E, et al. Double-blind evaluation of subthalamic nucleus deep brain stimulation in advanced Parkinson’s disease. Neurology 1998; 51: 850–5.
Langston JW, Widner H, Goetz CG, Brooks D, Fahn S, Freeman T, et al. Core Assessment Program for Intracerebral Transplantations (CAPIT). Mov Disord 1992; 7: 2–13.[ISI][Medline]
Levy R, Hutchison WD, Lozano AM, Dostrovsky JO. High-frequency synchronization of neuronal activity in the subthalamic nucleus of parkinsonian patients with limb tremor. J Neurosci 2000; 20: 7766–75.
Limousin P, Greene J, Pollak P, Rothwell J, Benabid AL, Frackowiak R. Changes in cerebral activity pattern due to subthalamic nucleus or internal pallidum stimulation in Parkinson’s disease. Ann Neurol 1997; 42: 283–91.[CrossRef][ISI][Medline]
Limousin-Dowsey P, Pollak P, Van Blercom N, Krack P, Benazzouz A, Benabid A. Thalamic, subthalamic nucleus and internal pallidum stimulation in Parkinson’s disease. [Review]. J Neurol 1999; 246 Suppl 2: II42–5.
Lozano AM, Lang AE. Pallidotomy for Parkinson’s disease. Adv Neurol 2001; 86: 413–20.[Medline]
Magill PJ, Bolam JP, Bevan MD. Dopamine regulates the impact of the cerebral cortex on the subthalamic nucleus-globus pallidus network. Neuroscience 2001; 106: 313–30.[CrossRef][ISI][Medline]
Magnani G, Cursi M, Leocani L, Volonté MA, Locatelli T, Elia A, et al. Event-related desynchronization parkinson’s disease: effect of acute administration of L-dopa [abstract]. Abstracts of the 14th International Congress of EEG and Clinical Neurophysiology. Electroencephalogr Clin Neurophysiol 1997; 103: 156.
Magnani G, Cursi M, Leocani L, Volonté MA, Locatelli T, Elia A, et al. Event-related desynchronization to contingent negative variation and self-paced movement paradigms in Parkinson’s disease. Mov Disord 1998; 13: 653–60.[CrossRef][ISI][Medline]
Molinuevo JL, Valldeoriola F, Tolosa E, Rumia J, Valls-Sole J, Roldan H, et al. Levodopa withdrawal after bilateral subthalamic nucleus stimulation in advanced Parkinson disease. Arch Neurol 2000; 57: 983–8.
Moro E, Scerrati M, Romito LM, Roselli R, Tonali P, Albanese A. Chronic subthalamic nucleus stimulation reduces medication requirements in Parkinson’s disease. Neurology 1999; 53: 85–90.
Obeso JA, Rodriguez-Oroz MC, Rodriguez M, Macias R, Alvarez L, Guridi J, et al. Pathophysiologic basis of surgery for Parkinson’s disease. Neurology 2000; 55 (12 Suppl 6): S7–12.[CrossRef][ISI][Medline]
Parent A, Hazrati L. Functional anatomy of the basal ganglia. II. The place of subthalamic nucleus and external pallidum in basal ganglia circuitry. Brain Res Brain Res Rev 1995; 20: 128–54.[CrossRef][Medline]
Pfurtscheller G, Berghold A. Patterns of cortical activation during planning of voluntary movement. Electroencephalogr Clin Neurophysiol 1989; 72: 250–8.[CrossRef][ISI][Medline]
Pierantozzi M, Mazzone P, Bassi A, Rossini PM, Peppe A, Altibrandi MG, et al. The effect of deep brain stimulation on the frontal N30 component of somatosensory evoked potentials in advanced Parkinson’s disease patients. Clin Neurophysiol 1999; 110: 1700–7.[Medline]
Pierantozzi M, Palmieri MG, Mazzone P, Marciani MG, Rossini PM, Stefani A, et al. Deep brain stimulation of both subthalamic nucleus and internal globus pallidus restores intracortical inhibition in Parkinson’s disease paralleling apomorphine effects: a paired magnetic stimulation study. Clin Neurophysiol 2002; 113: 108–13.[CrossRef][ISI][Medline]
Pinter MM, Alesch F, Murg M, Seiwald M, Helscher RJ, Binder H. Deep brain stimulation of the subthalamic nucleus for control of extrapyramidal features in advanced idiopathic parkinson’s disease: one year follow-up. J Neural Transm 1999; 106: 693–709.[Medline]
Rodriguez-Oroz MC, Gorospe A, Guridi J, Ramos E, Linazasoro G, Rodriguez-Palmero M, et al. Bilateral deep brain stimulation of the subthalamic nucleus in Parkinson’s disease. Neurology 2000; 55 (12 Suppl 6): S45–51.[ISI][Medline]