René Reese

Stimulation site within the MRI-defined STN predicts postoperative motor outcome.

High‐frequency stimulation of the subthalamic nucleus (STN‐HFS) is highly effective in treating motor symptoms in Parkinsons disease (PD) and medication side effects as well as in improving quality of life. Despite preoperative screening for patients as eligible candidates for this treatment, electrode position may furthermore influence treatment quality. Here, we investigated the relationship between the anatomical site of stimulation within the MRI‐defined STN and the outcome of PD patients after STN‐HFS. In 30 PD patients with bilateral STN stimulation, we retrospectively defined the boundaries of the STN within the axial target plane of the stereotactic T2‐weighted MRI and determined the position of the active electrode contact in relation to the border of the STN. The position of the active contact within the STN was the only variable to predict the outcome of STN stimulation. In contrast, covariates such as age, disease duration, symptom severity, and response to levodopa had no effect. The lateral position of the stimulation contact within the STN led to significantly better clinical improvement, lower stimulation parameters, and less need for postoperative dopaminergic medication. The outcome of patients with stimulation contacts within the medial region of the STN was significantly worse. Precise targeting of the lateral region of the STN is essential for achieving sufficient stimulation efficacy. Preoperative T2‐weighted MRI might be a useful component of the targeting procedure to improve the outcome of PD patients.

Short pulse width widens the therapeutic window of subthalamic neurostimulation

We explored the impact of pulse durations <60 μsec on the therapeutic window of subthalamic neurostimulation in Parkinsons disease. Current thresholds for full rigidity control and first muscle contractions were evaluated at pulse durations between 20 and 120 μsec during a monopolar review session in four patients. The average therapeutic window was 2.16 mA at 60 μsec, which proportionally increased by 182% at 30 μsec, while decreasing by 46% at 120 μsec. Measured chronaxies and model data suggest, that pulse durations <60 μsec lead to a focusing of the neurostimulation effect on smaller diameter axons close to the electrode while avoiding stimulation of distant pyramidal tract fibers.

Acquired stuttering after pallidal deep brain stimulation for dystonia

We report two patients, in whom stuttering evolved as an adverse effect of pallidal deep brain stimulation for treating dystonia. Speech dysfluency was observed under conditions that optimally suppressed dystonic symptoms without inducing other extrinsic stimulation effects. This emphasizes a role of the sensorimotor part of the internal globus pallidus in regulating speech fluency.

The central oscillatory network of orthostatic tremor

Orthostatic tremor (OT) is a movement disorder of the legs and trunk that is present in the standing position but typically absent when sitting. The pathological central network involved in orthostatic tremor is still unknown. In this study we analyzed 15 patients with simultaneous high‐resolution electroencephalography and electromyography recording to assess corticomuscular coherence. In 1 patient we were able to simultaneously record the local field potential in the ventrolateral thalamus and electroencephalography. Dynamic imaging of coherent source analysis was used to find the sources in the brain that are coherent with the peripheral tremor signal. When standing, the network for the tremor frequency consisted of unilateral activation in the primary motor leg area, supplementary motor area, primary sensory cortex, two prefrontal/premotor sources, thalamus, and cerebellum for the whole 30‐second segment recorded. The source coherence dynamics for the primary leg area and the thalamic source signals with the tibialis anterior muscle showed that they were highly coherent for the whole 30 seconds for the contralateral side but markedly decreased after 15 seconds for the ipsilateral side. The source signal and the recorded thalamus signal followed the same time frequency dynamics of coherence in 1 patient. The corticomuscular interaction in OT follows a consistent pattern with an initially bilateral pattern and then a segregated unilateral pattern after 15 seconds. This may add to the feeling of unsteadiness. It also makes the thalamus unlikely as the main source of orthostatic tremor.

Subthalamic deep brain stimulation increases pallidal firing rate and regularity

While high-frequency stimulation of the subthalamic nucleus (STN-HFS) is highly effective in the treatment of Parkinsons disease (PD), the mechanisms underlying its therapeutic action remain unclear. Here, we report changes of single-neuron pallidal activity during STN-HFS in a parkinsonian patient. STN-HFS increased firing rate in both segments of the pallidum. Neurons displayed time-locked responses to stimulation pulses, with an early excitation followed by inhibition and late excitation. Finally, pallidal neurons fired more regularly during STN-HFS. The time-locked responses and increased firing regularity may override abnormally patterned pallidal activity, and thereby significantly contribute to the clinical efficacy of STN-HFS in PD.

Deep brain stimulation changes basal ganglia output nuclei firing pattern in the dystonic hamster

Dystonia is a heterogeneous syndrome of movement disorders characterized by involuntary muscle contractions leading to abnormal movements and postures. While medical treatment is often ineffective, deep brain stimulation (DBS) of the internal pallidum improves dystonia. Here, we studied the impact of DBS in the entopeduncular nucleus (EP), the rodent equivalent of the human globus pallidus internus, on basal ganglia output in the dt(sz)-hamster, a well-characterized model of dystonia by extracellular recordings. Previous work has shown that EP-DBS improves dystonic symptoms in dt(sz)-hamsters. We report that EP-DBS changes firing pattern in the EP, most neurons switching to a less regular firing pattern during DBS. In contrast, EP-DBS did not change the average firing rate of EP neurons. EP neurons display multiphasic responses to each stimulation impulse, likely underlying the disruption of their firing rhythm. Finally, neurons in the substantia nigra pars reticulata display similar responses to EP-DBS, supporting the idea that EP-DBS affects basal ganglia output activity through the activation of common afferent fibers.

Subthalamic stimulation increases striatal tyrosine hydroxylase phosphorylation

Subthalamic stimulation enhances striatal tyrosine hydroxylase activity, which is regulated by phosphorylation at different serine residues. Western blotting was performed to investigate phosphorylation at the serine residues 19, 31 and 40 in striatal tissue of rats that had received subthalamic stimulation or sham stimulation for 2 h. In animals that were killed directly after stimulation, the tyrosine hydroxylase protein content was unchanged, whereas phosphorylation at the serine residue 19 was increased and phosphorylation at the serine residues 31 and 40 tended to be higher compared with controls. By contrast, tyrosine hydroxylase protein content and phosphorylation were similar in rats that were killed 24 h after stimulation. Our results suggest that subthalamic stimulation may increase tyrosine hydroxylase activity via increased phosphorylation.

High-Frequency Stimulation of the Subthalamic Nucleus Increases Pallidal Neuronal Firing Rate in a Patient with Parkinson's Disease

High-frequency stimulation of the subthalamic nucleus (STNHFS) has similar clinical effects in Parkinson’s disease (PD) as subthalamotomy but less risks. Surprisingly, a recent study in parkinsonian monkeys showed no inhibitory but an excitatory effect of STN-HFS on GPi neurons. This discrepancy could be explained by a dual effect of STN-HFS consisting of axonal excitation while the soma becomes silenced. We had the opportunity to study the impact of STN-HFS on pallidal neuronal activity in a parkinsonian patient. The 56-year-old man with a 40 year history of PD had previously undergone STN-HFS at our center for treating severe motor fluctuations and dyskinesias at the age of 51. Despite the unusually early onset of clinical symptoms, the patient did not carry a Parkin, PINK or DJ-1 mutation and had no family history of PD. STN-HFS (monopolar, 2 V, 180 Hz, 60 ls pulse width) was highly effective for offperiod motor symptoms but evoked choreatiform dyskinesias at a low threshold. Lower voltages alleviated the dyskinesias but led to recurrent OFF-states. Despite complete withdrawal of dopaminergic drugs and even temporary administration of haloperidol we were unable to control the self-injurious dyskinesias, which dominated the patient’s every day life. We therefore suggested additional bilateral pallidal HFS for their treatment. The study was covered by the local ethics committee (protocol approval A157/04) and the patient provided written consent. Owing to the severe dyskinesias, surgery had to be performed under general anesthesia with i.v. propofol and remifentanyl maintaining a deep sedation in which the patient tolerated artificial ventilation. All procedures for stereotactic targeting of the GPi, neurophysiological monitoring and postoperative confirmation of the correct electrode positioning have been described elsewhere. Microelectrode recordings were started 6 mm above the intended stereotactic target in the ventroposterior GPi. The border between the external (GPe) and internal pallidum was delineated by reduced spontaneous neuronal activity and the presence of border cells. Once stable neuronal activity was detected, we tested the effect of STN-HFS after recording at least 30 s of baseline activity. For this purpose, ipsilateral STN-HFS was switched on for about 10 s followed by another recording period of approximately 30 s after HFS. We used the same parameters as for chronic STN-HFS, but changed the lead configuration to bipolar (most upper contact as anode) to reduce far field electrical artifacts. Neuronal signals were stored for off-line analysis as described elsewhere. We followed the activity of 7 single units (SUA) in the GPe and 5 in the GPi before and after ipsilateral STN-HFS (Fig. 1A). The discharge rate of GPe neurons was significantly increased after STN-HFS [36.6 6 7.4 (mean 6 SEM) spikes/s] compared to baseline (20.9 6 4.8; Wilcoxon signed rank test, P 5 0.047) and a similar trend could be observed in GPi (18.4 6 3.8 spikes/s vs. 21.8 6 4.9; P 5 0.063) (Fig. 1B). These increases remained stable within the period recorded after STN-HFS (data not shown). The discharge pattern was burst-like in all except one neuron (GPi, irregular) at baseline and did not change after STN-HFS. Significant oscillations in the autocorrelation function were found in one GPi neuron at 5 Hz before and after HFS and in three neurons (one in GPi, two in GPe) in the range of 60 to 90 Hz only after HFS. The main finding was an after-effect of STN-HFS on pallidal neuronal activity with increased neuronal firing. We did not observe a consistent change in firing pattern or oscillatory SUA. Because clinical benefits of STN-HFS fade gradually within minutes after switching off stimulation, the physiological after-effect of STN-HFS observed here could be of clinical relevance and may indicate that neuronal excitation is also present during HFS. In animal models of STN-HFS excitation has been postulated as a possible mechanism of action. The mean baseline discharge rate of GPi [37.2 s 6 2.0 (mean 6 SEM) spikes/s] in our case was markedly lower than previously reported rates around 70 spikes/s in the Parkinsonian off-state, but similar to reduced discharge rates during apomorphine induced dyskinesias. The low rates could therefore correspond to the patient’s clinical dyskinetic state, which persisted even hours after reducing or stopping STN-HFS, but may have also been caused by propofol anesthesia. Different limitations in interpreting our data must be considered: The sample of recorded neurons was relatively small and artifacts did not allow reliable spike analysis during stimulation. Moreover, all data were obtained under general anesthesia and could therefore not be correlated to the patient’s clinical condition. Finally, the clinical presentation of PD and the response to STN-HFS with ‘‘non-habituating’’ stimulation induced dyskinesias despite complete withdrawal of dopaminergic drugs was special in this patient. After surgery GPiHFS was in fact helpful to control the disabling dyskinesia for a limited period, but the positive effect faded gradually over weeks. Our finding of increased pallidal neuronal discharges induced by STN-HFS in this patient opposes the classical rate model of basal ganglia disorders, which links reduced Published online 25 July 2008 in Wiley InterScience (www. interscience.wiley.com). DOI: 10.1002/mds.22225

Full Parkinsonian Triad Induced by Pallidal High-Frequency Stimulation in Cervical Dystonia

High-frequency stimulation of the globus pallidus internus (GPi-HFS) is the treatment of choice for patients suffering from generalized or segmental dystonia refractory to conservative medical treatment. Upper-limb bradykinesia or gait freezing have been described as adverse effects of pallidal neurostimulation. Bradykinesia is typically subtle, (e.g., presenting as mild micrographia) and is usually tolerated in return for improved dystonia. Here, we report on a patient with cervical dystonia, in whom GPi-HFS induced a clinically relevant, complete parkinsonian syndrome comprising bradykinesia, hand tremor, and severe freezing of gait (FOG). This 69-year-old man suffered from cervical dystonia since the age of 57. He had no family history of movement disorders, no previous exposure to antidopaminergic drugs, and cervical MRI was normal. After several years of successful botulinum toxine treatment, he became unresponsive as a result of neutralizing antibodies and opted for DBS. Surgical implantation of electrodes for GPi-HFS took place on 16 March 2009 following our published procedure. Intraoperatively, the GPi and its ventral border were identified by microrecordings, and HFS within the target region induced facial muscle contractions at 4 mA current amplitude (120 ls, 130 Hz), indicating a sufficient distance from the pyramidal tract. Permanent electrodes (model 3389; Medtronic Inc., Minneapolis, MN) were bilaterally implanted with the lowest contact at 2 mm below target and connected to an impulse generator (Activa PC; Medtronic). Postoperative MRI confirmed the anatomically correct placement of the electrodes and excluded an asymptomatic intracerebral hemorrhage. Four days later, after a monopolar review session, the most ventral contacts with a threshold for capsular side effects of at least 3 V were chosen for long-term stimulation and programmed to provide monopolar stimulation at an amplitude of approximately 20% below this threshold (GPi left: C+8-, 3.3 V, 120 ls, 130 Hz; GPi right: C+1-, 3.3 V, 120 ls, 130 Hz). After another 2 days of clinical observation, the patient was sent home claiming no adverse effects while the severity of dystonic head rotation was already markedly reduced. A few weeks later, the patient returned complaining about difficulties in walking and handwriting. Clinically, he exhibited clear bradykinesia with slowing and decrement of repetitive movements and a marked resting and kinetic tremor of the right hand, but only mild rigidity. Walking was impaired by difficulties of gait initiation, FOG, and turning difficulties. The video, demonstrating parkinsonian symptoms during stimulation with the initially programmed parameters, was taken 1 year after electrode implantation. Kinematic gait analysis on a motor-driven tread mill (Woodway, Weil am Rhein, Germany) was used to quantify the hypokinetic gait disturbance (Fig. 1C). A cranial MRI revealed no parenchymal abnormalities and a normal DaTSCAN excluded the possiblity of striatal dopaminergic degneration. Oral levodopa did not improve the parkinsonian symptoms. Stopping GPi-HFS or even reducing stimulation intensity led to abrupt worsening of cervical dystonia. Parkinsonian symptoms gradually diminished within days (see Video) after discontinuing HFS, suggesting an effect of stimulation rather than microlesion. Stimulation of dorsal contacts failed to control dystonia, but did not induce parkinsonism. As a compromise, we finally programmed the clinically most effective ventral contacts in bipolar mode with the adjacent contact as anode and titrated stimulation intensity to evoke sufficient dystonia control, but minimal parkinsonian signs (GPi right: 4.5 V, 120 ls, 160 Hz; GPi left: 3.8 V, 90 ls, 160 Hz; Fig. 1A,B). In patients with Parkinson0s disease, HFS of the ventral GPi reduces L-dopa-induced dyskinesias and rigidity, whereas the antiakinetic effect of L-dopa may become antagonized. In contrast, stimulation of the dorsal GPi was found to induce dyskinesia and improve akinesia. This functional segregation may anatomically relate to the two different pallido-fugal fiber tracts, the ansa lenticularis, and the lenticular fasciculus, leaving the nucleus at the ventral or dorsal aspect, respectively. The exact pathophysio-

High frequency stimulation of the entopeduncular nucleus sets the cortico-basal ganglia network to a new functional state in the dystonic hamster.

High frequency stimulation (HFS) of the internal pallidum is effective for the treatment of dystonia. Only few studies have investigated the effects of stimulation on the activity of the cortex-basal ganglia network. We here assess within this network the effect of entopeduncular nucleus (EP) HFS on the expression of c-Fos and cytochrome oxidase subunit I (COI) in the dt(sz)-hamster, a well-characterized model of paroxysmal dystonia. In dt(sz)-hamsters, we identified abnormal activity in motor cortex, basal ganglia and thalamus. These structures have already been linked to the pathophysiology of human dystonia. EP-HFS (i) increased striatal c-Fos expression in controls and dystonic hamsters and (ii) reduced thalamic c-Fos expression in dt(sz)-hamsters. EP-HFS had no effect on COI expression. The present results suggest that EP-HFS induces a new network activity state which may improve information processing and finally reduces the severity of dystonic attacks in dt(sz)-hamsters.