Dystonia is characterized by sustained muscle

This article has been updated from its originally published version to replace Figure 2 and update the legend. See the corresponding erratum notice, DOI: 10.3171/10.3171/ 2013.11.JNS13844a, for full details. J Neurosurg 119:1537 1545, 2013 AANS, 2013 A randomized double-blind crossover trial comparing subthalamic and pallidal deep brain stimulation for dystonia Clinical article Lisbeth Schjerling, M.D., 1,2 Lena E. Hjermind, M.D., Ph.D., 3,4 Bo Jespersen, M.D., 1 Flemming F. Madsen, M.D., D.Sc., 1 Jannick Brennum, M.D., D.Sc., 1 Steen R. Jensen, R.N., 5 Annemette Løkkegaard, M.D., Ph.D., 5 and Merete Karlsborg, M.D. 5 1 Department of Neurosurgery, Rigshospitalet, Copenhagen University Hospital; 2 Department of Emergency, Hilleroed Hospital, Copenhagen University Hospital; 3 Section of Neurogenetics, Memory Disorders Research Group, Department of Neurology, Rigshospitalet, Copenhagen University Hospital; 5 Department of Neurology, Bispebjerg, Copenhagen University Hospital, Copenhagen, Denmark; and 4 Department of Cellular and Molecular Medicine, Section of Neurogenetics, Faculty of Health Sciences, University of Copenhagen, Copenhagen, Denmark Object. The authors aim was to compare the subthalamic nucleus (STN) with the globus pallidus internus (GPi) as a stimulation target for deep brain stimulation (DBS) for medically refractory dystonia. Methods. In a prospective double-blind crossover study, electrodes were bilaterally implanted in the STN and GPi of 12 patients with focal, multifocal, or generalized dystonia. Each patient was randomly selected to undergo initial bilateral stimulation of either the STN or the GPi for 6 months, followed by bilateral stimulation of the other nucleus for another 6 months. Preoperative and postoperative ratings were assessed by using the Burke-Fahn-Marsden Dystonia Rating Scale (BFMDRS) and video recordings. Quality of life was evaluated by using questionnaires (36-item Short Form Health Survey). Supplemental Toronto Western Spasmodic Torticollis Rating Scale (TWSTRS) scores were assessed for patients with focal dystonia (torticollis) by examining the video recordings. Results. On average for all patients, DBS improved the BFMDRS movement scores (p < 0.05) and quality of life physical scores (p < 0.01). After stimulation of the STN, the mean 6-month improvement in BFMDRS movement score was 13.8 points; after stimulation of the GPi, this improvement was 9.1 points (p = 0.08). Quality of life did not differ significantly regardless of which nucleus was stimulated. All 12 patients accepted 6 months of stimulation of the STN, but only 7 accepted 6 months of stimulation of the GPi. Among those who rejected stimulation of the GPi, 3 accepted concomitant stimulation of both the STN and GPi for 6 months, resulting in improved quality of life physical and mental scores and BFMDRS movement scores. Among the 4 patients who were rated according to TWSTRS, after 6 months of stimulation of both the STN and GPi, TWSTRS scores improved by 4.7% after stimulation of the GPi and 50.8% after stimulation of the STN (p = 0.08). Conclusions. The STN seems to be a well-accepted, safe, and promising stimulation target in the treatment of dystonia, but further studies are necessary before the optimal target can be concluded. Simultaneous stimulation of the STN and GPi should be further investigated. Clinical trial registration no.: KF 01-110/01 (Committees on Biomedical Research Ethics of the Capital Region of Denmark). (http://thejns.org/doi/abs/10.3171/2013.8.jns13844) Key Words subthalamic pallidal DBS deep brain stimulation randomized dystonia quality of life functional neurosurgery Dystonia is characterized by sustained muscle contractions that cause twisting and repetitive movements or abnormal postures. It is classified according to age at onset, localization of symptoms, and supposed etiology. 9 Although assumed to be a dysfunction in the processing of sensorimotor signals, the exact pathophysiology of dystonia is still unclear; it might differ for Abbreviations used in this paper: BFMDRS = Burke-Fahn-Marsden Dystonia Rating Scale; DBS = deep brain stimulation; GPi = globus pallidus internus; PD = Parkinson s disease; SF-36 = 36-item Short Form Health Survey; STN = subthalamic nucleus; TWSTRS = Toronto Western Spasmodic Torticollis Rating Scale. primary and secondary dystonia. 5 Medical treatment can be troublesome, leading to unacceptable side effects or insufficient or unsustained symptom relief. Well known as a treatment for Parkinson s disease (PD), deep brain stimulation (DBS) has become an effective alternative or supplement to medical treatment of dystonia. 6,16 20,23,24,28,30,31 It is based on implantation of electrodes into specific target areas in the brain, which deliver low-voltage stimulation with high frequency. Deep brain stimulation is assumed to disrupt pathological signals and perhaps also influence the release of neurotransmitters. 13 At most centers, the preferred DBS target for pa- 1537

L. Schjerling et al. tients with dystonia has been the globus pallidus internus (GPi). 16,19,20,30,31 However, for patients with PD, stimulation of the STN has also been effective for controlling dystonic symptoms. 21 We expected that stimulation of the STN might also be effective for controlling dystonic symptoms in patients without PD, perhaps even more effective than stimulation of the GPi. This expectation is supported by theoretical assumptions describing the functional organization of the basal-ganglia-thalamocortical circuits, 1,13 including the proposed direct, indirect, and hyperdirect pathways. 1,13,22,35 The STN has been suggested to be a central nucleus in the indirect and hyperdirect pathways and to have a substantial effect on the control of basal ganglia output. Recent studies among 1 12 patients have shown that the STN can be used as a target in the treatment of dystonia; results have been promising, supporting the assumption that the STN might be a better target than the GPi. 6,18,21,24,28 To further investigate this assumption, we conducted this double-blind randomized study comparing the STN and the GPi as DBS targets in the treatment of dystonia; we used alternating stimulation through electrodes placed in the STN and the GPi. Methods Patients During the period from 2002 to 2009, patients consecutively admitted to the Department of Neurology, Bispebjerg Hospital, Copenhagen, were included in the study. The inclusion criteria were medically intractable dystonia and age 18 65 years. The exclusion criteria were inability to speak and understand Danish fluently, short life expectancy, alcohol or drug abuse, psychiatric disease, followup impossibility, or presence of structural brain lesions. This clinical trial was approved in 2001 and monitored by the Committees on Biomedical Research Ethics of the Capital Region of Denmark (no. KF 01-110/01). Molecular genetic tests were performed for the most frequent mutations in the dystonia genes: DYT1 (TOR1A), DYT5a (GCH1), and DYT11 (SGCE). Preoperative brain MRI was performed for surgical planning and exclusion of structural brain lesions. The patients were evaluated by a neuropsychologist to detect cognitive dysfunctions and, if clinically relevant, by a psychiatrist to exclude psychiatric diseases. Surgical Procedure Stereotactic framed-based CT scans of the brain were fused with preoperative MR images using Stereoplan (Radionics) or I-plan (Brainlab). The target point was marked on T2-weighted MR images by visual comparison with the Schaltenbrand-Wahren stereotactic atlas, and an entry point was selected on the top of the middle frontal gyrus, avoiding sulci and lateral ventricles. Target points were placed central in the STN and posteroventral in the GPi. Stereotactic placement of electrodes was performed, and the optimal placement was confirmed through microrecording. All operations were performed by the same 2 experienced neurosurgeons at the Department of Neurosurgery, Rigshospitalet, Copenhagen. Postoperative MRI or CT was performed to confirm the correct position of the electrodes. Patients were transferred to the Department of Neurology at Bispebjerg Hospital for adjustment of stimulation parameters. Postoperative Assessment The day after surgery, the patients were randomly selected by the research nurse, using a statistically validated randomization table, to receive DBS of the STN or the GPi for the first 6 months. The nuclei were activated bilaterally. A double monopolar setting was chosen. The pulse width and frequency were held constant at 60 msec and 130 Hz, respectively; the amplitude was increased in increments of 0.1 V to 0.5 V per day, with a goal of reaching 2.5 V in the nucleus within the first 2 weeks. The contacts, amplitude, and pulse width were adjusted during the next 5.5 months until the optimal stimulation parameters were established, that is, by adjusting the settings after each month of stable settings. After 6 months, the site of stimulation was changed to the alternate target and the procedure was repeated. If a satisfactory effect was not achieved when either the STN or the GPi was stimulated, concomitant stimulation in both the STN and the GPi for 6 months was offered. All adjustments were performed by the same experienced nurse throughout the study. Clinical Evaluation For each period of stimulation, the severity of dystonia was rated preoperatively and after 1 week, 3 months, and 6 months. All ratings were performed by 2 experienced neurologists using the Burke-Fahn-Marsden Dystonia Rating Scales (BFMDRSs) for movement (BFMDRS movement) and disability (BFMDRS disability). 9 The ratings were documented by video recordings. Video 1 shows the evaluation of Patient 1. Video 1. Clip demonstrating part of the clinical evaluation of Patient 1. Evaluation is shown before the procedure, during DBS in the STN and then during DBS in the STN and the GPi simultaneously. Copyright Lisbeth Schjerling. Published with per mission. Click here to view with Media Player. Click here to view with Quicktime. Raters and patients were blinded as to which site was stimulated. Quality of life was evaluated using the 36- item Short Form General Health Survey (SF-36), 34 which was completed by the patients before the procedure and at 3 and 6 months after the procedure, for each period of stimulation. The questions were related to either physical health (physical function, physical role, body pain, and general physical health [quality of life physical score]) or to mental health (vitality, social function, emotion, and general mental health [quality of life mental score]). Post hoc analyses of the videos of patients with torticollis were performed according to the Toronto Western Spasmodic Torticollis Rating Scale (TWSTRS). 7 Ethics The patients gave written consent to participate in the study, and the study protocol was approved by the Committee of Science and Ethics, Copenhagen. Separate written consent was obtained from the patient shown in Video 1 (Patient 1). 1538

Subthalamic or pallidal DBS for dystonia Statistical Analysis Burke-Fahn-Marsden Dystonia Rating Scale data were analyzed by using hypothesis testing based on an intention-to-treat statistical analysis with repeated measures and correlation for time. The Wilcoxon signed-rank test was used to compare TWSTRS data. A descriptive statistical method was used in figures and for comparing means. Standard errors of means were calculated when appropriate. A p value 0.05 was considered significant; a trend was considered for p values between 0.05 and 0.10; and a p value > 0.10 was considered not significant. Results Patients Of the 13 patients who agreed to participate in the study, 6 had generalized dystonia, 1 had multifocal dystonia, and 6 had focal dystonia (all torticollis). The mean patient age at disease onset was 35.3 years (range 12 57 years), and the mean age at time of surgery was 51 years (range 32 68 years). The mean duration of disease was 16.2 years (range 3 30 years). Eleven patients were negative for the DYT1, DYT5, and DYT11 mutations; the other 2 (Patients 11 and 12) did not agree to undergo molecular genetic testing. For 11 patients, the dystonia was diagnosed as primary by medical history, biochemical analyses, and neuroimaging. For 2 patients (Patients 2 and 8), the dystonia was believed to be caused by previous medication, either neuroleptics (clozapine or haloperidol) or metoclopramide; each of these 2 patients experienced focal dystonia. Preoperative data are shown in Table 1. Preoperatively, 11 patients received botulinum toxin injections, 2 of whom had discontinued botulinum toxin because of lack of effectiveness. Postoperatively, only 2 patients continued treatment with botulinum toxin. In addition to adjustment of the botulinum treatment, other changes in the medical treatment throughout the study period were insignificant. Eight patients (Patients 2, 3, 5, 6, and 8 11) completed the stimulation protocol as planned and received stimulation of each target for 6 months. However, 1 patient (Patient 10) could not complete the final ratings because of concomitant disease unrelated to the study (a fractured leg). Some patients did not follow the protocol. Patient 13 was excluded from the study early because of an infection that necessitated removal of the electrodes. For Patient 7, signs of depression developed after 3 months of stimulation of the first target (GPi), and dystonic symptoms showed no improvement. Further stimulation of this target was therefore not acceptable, and at the 3-month ratings it was decided to change to the second target (STN), which was subsequently stimulated for 6 months as planned. The patient and raters were still blinded; for that reason, the patient was not excluded despite the interrupted protocol. Patients 1, 4, and 12 did not accept the change of stimulation target from the first (STN) to the second (GPi) because of worsening dystonia. It was consequently decided to stimulate both targets simultaneously (unblinded) for 6 months, although this procedure was not originally described in the protocol. Placement of Electrodes Single-channel microrecording was applied. Alternative trajectories were tested, but for only 1 patient did the alternative trajectory produce more optimal recordings, and, consequently, placement of the electrode was changed. The localization of the electrodes was investigated with postoperative MRI and/or CT (Fig. 1); 1 patient (Patient 12) refused to undergo postoperative imaging. For optimal localization, postoperative CT scans were merged with preoperative T2-weighted MR images. The position of the electrodes was plotted on the Schaltenbrand atlas by visual comparison with the nuclei TABLE 1: Clinical data and preoperative ratings* BFMDRS Score QOL Score Patient No. Age at Onset (yrs), Sex Duration (yrs) Phenotype Etiology Mov Dis Phys Ment Primary DBS Site 1 39, M 17 generalized primary 48.0 15 29 56 STN 2 52, F 7 focal secondary 12.5 7 31 40 GPi 3 34, F 4 generalized primary 44.0 8 23 51 STN 4 16, F 22 multifocal primary 43.0 7 39 25 STN 5 45, F 21 focal primary 7.5 2 28 29 GPi 6 25, M 26 generalized primary 30.0 7 36 44 STN 7 40, F 20 focal primary 8.0 1 59 83 GPi 8 57, F 3 focal secondary 8.0 8 21 22 Gpi 9 12, F 21 generalized primary 71.0 13 43 57 GPi 10 52, F 16 focal primary 12.0 10 24 44 STN 11 47, F 9 focal primary 9.0 5 46 52 STN 12 28, F 15 focal primary 56.0 12 39 39 STN 13 12, M 30 generalized primary 78.5 28 37 53 STN * Dis = disability; Ment = mental; Mov = movement; Phys = physical; QOL = quality of life. Patient 13 was excluded from the study (because of electrode removal). 1539

L. Schjerling et al. Fig. 1. Postoperative MR image demonstrating the position of the implanted electrodes in the axial plane. as seen on T2-weighted MR images (Fig. 2). Most electrodes were positioned near the intended location (central in the STN and posteroventral in the GPi), but for 3 patients (Patients 4, 5, and 7), the electrodes in the GPi were placed more anteroventrally than posteroventrally. These 3 patients had either multifocal or generalized dystonia. BFMDRS Movement and Disability Scores Intention-to-treat analysis with repeated measures of all included patients showed a significant effect of DBS on BFMDRS movement scores (p < 0.05). Improvements were seen after the 1st week of STN or GPi stimulation (Fig. 3). The 6-month BFMDRS movement scores for the first period were improved by 10.8 ± 4.4 points compared with an improvement of 13.8 ± 6.2 points for the second period. When ratings from the first and second periods were combined, the mean 6-month improvement of BFM- DRS movement scores was 13.8 ± 4.2 points (44%) after STN stimulation compared with 9.1 ± 6.7 points (8%) after GPi stimulation. Testing the hypothesis that for patients with dystonia the STN could be a better target than the GPi, intention-to-treat analysis was correlated for time and data as referred to the baseline to compensate for the possible difference between nuclei being stimulated in the first or second period. Intention-to-treat analysis then showed a trend for better improvement during the STNstimulated period compared with the GPi-stimulated period (p = 0.08). It is notable that in following the intention-to-treat concept, the scores from the 3 patients who received stimulation to both targets instead of to the second target alone (GPi) in the second period were included in the statistical analysis as GPi results. This inclusion of these scores might explain why we found only a trend for the STN being superior despite finding an apparently marked difference in mean BFMDRS movement scores in the descriptive analysis. With respect to BFMDRS disability score, intentionto-treat analysis with repeated measurements showed no significant effect of time or nucleus stimulated. The mean 6-month improvement in BFMDRS disability score was 1.8 ± 0.6 for STN stimulation and 1.3 ± 1.2 for GPi stimulation. The mean improvements in BFMDRS movement scores are shown in Fig. 3. For the 6 patients with multifocal or generalized dystonia, BFMDRS movement scores improved by 23.8 ± 4.8 points (46%) after 6 months of either STN or GPi stimulation. For the 6 patients with torticollis (primary or secondary), BFMDRS movement scores improved by 1.5 ± 1.8 points (17%). For the 2 patients with secondary dystonia (Patients 2 and 8), scores showed negative or very little effects of GPi stimulation. For 2 patients who initially received GPi stimulation (Patients 2 and 5), the averaged BFMDRS movement and BFMDRS disability scores suggested worsening at 6 months (BFMDRS movement scores increased from 12 to 24.5 and from 7.5 to 12, respectively). The electrodes were located posteroventrally in the GPi, as intended, in Patient 2 but were located quite anteriorly in Patient 5. Changing the target to the STN resulted in improvement for both patients. For 2 other patients (Patients 4 and 7), electrodes were placed anteriorly and centrally in the GPi according to postoperative MR and CT images. Patient 4 initially received STN stimulation, and the BFMDRS movement score improved from 43 to 19 after 6 months. When the target was changed to the GPi, 1-week ratings still showed marked improvement (BFMDRS movement score 24), despite the frontally located electrodes. However, this patient did not accept the worsening in symptoms when the stimulated target changed from the STN to the GPi and refused further stimulation of the GPi. Instead, concomitant stimulation of the STN and the GPi was performed. For Patient 7, for whom electrodes were suboptimally positioned, the BFMDRS movement score did not change after GPi stimulation but did improve from 8 (preoperative rating) to 2 (6-month rating) with STN stimulation. For 3 patients (Patients 1, 4, and 12) who did not accept change of the target, simultaneous stimulation of the STN and the GPi was performed, as illustrated in Fig. 3C. For these patients, BFMDRS movement scores improved more when receiving stimulation to both targets than when receiving stimulation to the STN alone. For the BFMDRS movement scores, when both targets were stimulated, mean score improvement from preoperative to 6 months was 38.0 (77%); when only the STN was stimulated, mean improvement was 23.5 (48%). For the BFMDRS disability score, when both targets were stimulated, mean improvement was 4.0 (30%); when only the STN was stimulated, mean improvement was 2.0 (14%). TWSTRS Scores The 6 patients with focal dystonia were also rated according to TWSTRS (Table 2). Because TWSTRS scores were estimated retrospectively from video recordings, it was not possible to determine subscore C (the effect of sensory trick). These subscores are therefore not included in the total TWSTRS scores. Only 4 of the 6 patients with focal dystonia complet- 1540

Subthalamic or pallidal DBS for dystonia Fig. 2. Eleven patients had postoperative documentation of the electrode placement: 4 by MRI, 3 by CT, and 4 by both. The center of each electrode was marked on axial slides of postoperative CT or FLAIR MR images, which were fused to preoperative T2-weighted MR images to ensure that anatomical structures were not influenced by artifacts. The positions are marked on a standard anatomical atlas. AC = anterior commissure; Fx = fornix; GPe = globus pallidus externus; GPi = globus pallidus internus; Int. Cap. = internal capsule; MMT = mamillothalamic tract; Ruber = nucleus ruber (red nucleus); STN = subthalamic nucleus; III = third ventricle. A: GPi. B: STN. ed 6 months of stimulation of both the STN and the GPi, respectively. Stimulation effects differed; the median improvement in TWSTRS score with GPi stimulation was 1 point, whereas that with STN stimulation was 13 points. For those 4 patients, the mean improvement in TWSTRS score with GPi stimulation was 4.8%, and that with STN stimulation was 50.9%. The difference between stimulation of the STN and the GPi did not, however, reach statistical significance (p = 0.08). Quality of Life Scores Intention-to-treat statistical analysis of repeated measurements in all included patients showed that DBS had a significant effect on the quality of life physical score (p < 0.01) but not on the quality of life mental score. After 6 months of DBS, the mean increase in the quality of life physical score was 17.9 ± 4.5 points (58%), and the mean increase in the quality of life mental score was 13.1 ± 4.7 points (37%). A comparison of the effects of stimulation of the STN and the GPi showed that after 6 months, the mean improvements in quality of life physical scores were 22.5 ± 6.1 points (68%) and 10.1 ± 5.5 points (41%), respectively. The mean improvement in quality of life mental scores Fig. 3. Ratings in BFMDRS (BFM) movement score. A: The 5 patients who initially received GPi stimulation followed by STN stimulation. B: The 4 patients who initially received STN stimulation followed by GPi stimulation. C: The 3 patients who initially received STN stimulation but did not accept stimulation of the GPi and instead received 6 months of simultaneous STN and GPi stimulation. 1w = 1 week; 3m = 3 months; 6m = 6 months. 1541

L. Schjerling et al. TABLE 2: Summary of TWSTRS ratings for patients with torticollis* Time Patient No. 1 2 8 9 11 12 Significance preop score 14 29 28 23 24 23 1-wk score STN 6 ND 21 13 25 12 GPi 21 26 24 22 ND 12 NS 3-mo score STN ND 8 19 ND 23 15 GPi 20 22 26 23 22 23 NS 6-mo score STN 1 7 15 14 26 24 GPi 10 28 ND 23 ND 26 NS * ND = ratings not done; NS = not significant. Subscore C could not be determined from video recordings and therefore is not included in the TWSTRS scores. p = 0.08. was 20.1 ± 6.3 points (55%) with STN stimulation but 1.1 ± 4.4 points (4%) with GPi stimulation (Fig. 4). However, using intention-to-treat statistical analysis for testing the study hypothesis, we found that quality of life-physical (p = 0.14) or quality of life-mental scores did not differ significantly with stimulation of the STN or the GPi (p = 0.13). For the 3 patients who received simultaneous stimulation of the STN and the GPi, the mean quality of life physical scores improved by 43.2 points (126%) after 6 months of stimulation of both targets and by 39.9 points (111%) after 6 months of STN stimulation only. For these same 3 patients, the quality of life mental score improved 37.4 points (115%) after 6 months of simultaneous STN and GPi stimulation compared with 39.3 points (123%) after 6 months of STN-only stimulation. Stimulation Parameters All patients received stimulation at 130 Hz. In the respective 6-month study periods, maximum amplitudes in individual patients were 4.0 V for the GPi and 3.6 V for the STN. A maximum pulse width of 240 msec was attempted for the GPi and 90 msec for the STN. A pulse width below 60 msec was not attempted for either target. The mean voltage settings were 2.8 ± 0.2 V for the GPi and 2.6 ± 0.12 V for the STN. The optimal pulse width was 116.3 ± 18.0 msec for the GPi and 60.0 msec for the STN. Postoperative Complications In 2 patients (Patients 6 and 13), infection developed and resulted in unilateral removal of both electrodes and stimulator. For Patient 6, the stimulator was later reimplanted and the protocol was resumed without further complications. Patient 13 was excluded because of the permanent removal of the electrodes. Two months after surgery, Patient 11 experienced lipothymia and was evaluated with blood tests and electrocardiography; results were within reference range. Epilepsy was not suspected. The patient continued stimulation as planned with no further complications. Adverse Events For 3 patients who received GPi stimulation (Patients 5, 7, and 11), signs of depression developed despite stimulator setting adjustments. Patient 11 responded well to citalopram during continued GPi stimulation, and the other 2 patients improved after the target was changed to the STN. Two patients complained of reduced neck mobility, thought to be a possible consequence of the electrical wires. Reoperation was not necessary. Hyperkinesia was seen, especially with stimulation of the STN, but it was generally mild and responsive to adjustment. Microlesion effects were suspected in 3 patients with initial hyperkinesia and spontaneous improvement within weeks of the operation. Sustained hyperkinesia despite optimal adjustments was seen in 3 patients (2 with STN stimulation and 1 with GPi stimulation). Other adverse effects were temporary and insignificant, such as mild hypomania, dysphagia, and nausea, and resolved without treatment when optimal stimulation settings were accomplished. Discussion To our knowledge, ours is the first randomized double-blind study to compare DBS of the GPi with DBS of the STN as target nuclei in patients with dystonia. Thirteen patients agreed to participate; 1 was excluded. Both STN and the GPi stimulation exhibited a significant clinical effect. We found no statistically significant difference in BFMDRS scores when comparing the targets, but the results suggest that stimulation of the STN produced a greater improvement in BFMDRS movement scores than did stimulation of the GPi. Stimulation of the STN resulted in better compliance. Six months of STN stimulation was completed by 12 pa- 1542

Subthalamic or pallidal DBS for dystonia Fig. 4. Quality of life (QOL) scores were calculated for all patients preoperatively and after 6 months of stimulation of the GPi, the STN, or both (GPi+STN). Scores are presented as mean ± SEM (error bars). tients, but 6 months of GPi stimulation was completed by only 8 patients (including the patient who was not rated because of concomitant disease). Moreover, we found that STN stimulation resulted in no more adverse effects than did GPi stimulation. We chose to implant electrodes in both the STN and the GPi and randomly assign the patients to initial stimulation of the STN or the GPi, followed by stimulation of the other nucleus. In this way the patients were their own controls and statistical variations were minimized. Because of the low number of patients in this study, an alternative design would be more sensitive to the preoperative severity of disease and the response to DBS. In our study, 3 patients who left the blinded full program after the first period of STN stimulation received simultaneous stimulation of the STN and the GPi. To our knowledge, double stimulation for pure dystonia has not been previously described. It has, however, been reported for 3 patients with dystonia-parkinsonism; the effect was reported to be either equal or additional to that of STN or GPi stimulation alone. 4,36 Our study showed an additional effect for 2 of the 3 patients stimulated at both the STN and the GPi, suggesting that simultaneous stimulation of the STN and the GPi might produce an additional effect in patients with dystonia. Further studies are warranted to clarify this assumption. We found an overall improvement of the BFMDRS movement score of 42% after DBS. This finding is comparable with those of previous studies that found 33% 89% improvements. 20,28,30 The BFMDRS movement scores in patients with torticollis improved 17%. Torticollis patients are reported to respond well to DBS when the STN is stimulated. 3,16,24,26 However, in most of these studies, the patients were rated with TWSTRS. For torticollis symptoms, this scale is more selective than the BFMDRS. When our study was designed, we expected to mainly include patients with generalized, multifocal, or segmental dystonia, and we therefore chose to use the BFMDRS as the rating scale. However, we decided to perform supplementary evaluation with TWSTRS when the study was ongoing to complement the BFMDRS results. The resulting TWSTRS ratings based on video recordings support the conclusion that STN stimulation is at least as effective as GPi stimulation also in the subgroup of patients with torticollis. The improvement in TWSTRS scores in our study was 5% when patients were stimulated at the GPi compared with 51% when the same patients were stimulated at the STN. However, only 4 patients completed the 6 months of GPi stimulation, and there was no significant difference (p = 0.08). A different range (2% 70%) of improvements in TWSTRS scores has been reported for treating torticollis with DBS. 24,26 In our study, the limited TWSTRS score improvement with GPi stimulation could have various explanations. The range of amplitude tested when adjusting the settings might have been too low; maximum amplitude was 4.0 V. The optimal setting can be as high as 4.6 V for some patients when stimulating the GPi for dystonia, 29 but high settings were not tested in our study. Suboptimal location of electrodes could also influence the results. When treating primary dystonia with DBS, the best effects are reported when electrodes are located posteroventrally in the GPi. 29 Although postoperative MR and CT images seem to confirm a good position of the GPi electrodes in torticollis patients, determining the exact location of electrodes by MRI can be difficult because of artifacts from the electrodes. Consequently, suboptimal position of electrodes cannot be excluded. In 3 patients with generalized and multifocal dystonia, the electrodes in the GPi were located more anteriorly than intended. Consistent with this positioning, for 2 of these 3 patients, GPi stimulation had little or negative effect. Only 2 patients in our study had secondary dystonia resulting from earlier treatment (tardive dystonia). The effect of DBS is thought to depend on etiology, and patients with primary dystonia as well as tardive dystonia are reported by most studies to respond well to DBS. 3,28 However, these results are based on stimulation of the GPi, and specific types of dystonia probably respond differently to stimulation of the STN and the GPi. In our study, STN stimulation produced the same or a better effect than GPi stimulation in the 2 patients with tardive dystonia (both with torticollis). However, no conclusions regarding effects on different kinds of dystonia can be drawn from this study because of the small number of patients and the heterogeneity of etiology and phenotype. Further studies with selected groups of dystonia patients are needed to address this question. With respect to the BFMDRS ratings, the effect of DBS was observed in the first week but did not improve between the first week and 6 months of stimulation. A delayed effect of DBS in patients with dystonia has been described. Additional improvement from DBS of the GPi has been reported for up to 1 year postoperatively for patients with generalized dystonia and up to 2 years postoperatively for patients with torticollis. 38 However, the main improvement seems to be during the first 6 months. 8,37 In our study, we therefore considered 6 months of observation to be sufficient. The study design required that stimulation of the new target during the second period be initiated immediately after stimulation of the first target. It has been reported 1543

L. Schjerling et al. that pausing DBS in dystonic patients results in a very fast return to preoperative symptoms. 12 However, it cannot be ruled out that stimulation in the first period may change the effect of stimulation in the second period by modulating the pathological pathways involved in dystonia. The BFMDRS movement score was improved by 10.8 points after 6 months of stimulation in the first period and by 13.8 points after an additional 6 months of stimulation in the second period. However, the variation in improvement can be partly accounted for by different occurrence of dropouts. In the first period, all patients completed the DBS regardless of worsened symptoms. In the second period, 3 patients dropped out before the 6-month ratings were performed, possibly affecting the mean 6-month improvement in a positive way. In the intention-to-treat analysis correlation for time was included. In this study, we encountered only a few severe surgical complications of DBS (that is, 2 instances of infection). Others have found increased rates of mortality (< 0.4%), intracranial hemorrhage (< 2%), and infections (< 5%). 19,25,33 Adverse effects such as hyperkinesia were typically resolved with optimal resetting of the stimulation parameters. For 3 patients, hyperkinesia persisted despite adjustment: 2 with STN stimulation and 1 with GPi stimulation. For patients with PD, stimulation of the GPi or STN is used for fluctuations and hyperkinesia. Although GPi stimulation is known to have a direct effect on dyskinesia, STN stimulation is thought to be effective in these patients because of an effect on parkinsonian symptoms causing substantial reductions in medication, thereby avoiding hyperkinesia. 32 For 3 patients, signs of depression developed during GPi stimulation but not during STN stimulation. In patients wth dystonia, depression has been reported with stimulation of the STN as well as the GPi. 15,24 Depression in patients with PD is reported more often when the STN rather than the GPi is the DBS target. 2,10 However, this finding might be a consequence of the abrupt and significant reduction of dopaminergic medication in patients when the STN, in contrast to the GPi, is stimulated and may not be related to the stimulation of the different targets. 11 The effect of DBS on the STN might therefore differ in patients not receiving levodopa. Because all patients in our study had pure dystonia, none received levodopa. Our quality of life analysis of dystonia patients did not suggest a negative effect on mental health scores with STN stimulation. Several studies have reported that quality of life is significantly reduced for patients with both focal and generalized dystonia. 15 The most common scale used to measure quality of life is the SF-36 15 ; this scale has been validated in comparison with other determinants of quality of life. 27 The SF-36 displays physical and mental scores of the dystonic patient, including vitality, social functioning, and mental health. It has been discussed whether the mental scores are just manifestations of a chronic disabling disease or whether there might be a more direct effect on emotional symptoms. 27 DBS of the GPi improves quality of life in patients with dystonia. Improvement has been shown in physical and mental scores and in physical scores only. 15,16,20,30,31 Few researchers have investigated quality of life for patients with dystonia undergoing STN stimulation. 18,23,24 The largest study, to our knowledge, included 9 patients with torticollis who were treated with STN stimulation and showed improvement of physical health but not of mental health. 24 In our study, stimulation of the STN or the GPi improved physical health, but mental health was not significantly improved. We detected no difference between stimulation of the STN and the GPi. Subthalamic nucleus stimulation needed lower pulse widths than did GPi stimulation. Lower pulse width results in longer battery life. 14 Battery changes carry an inherent risk for infection, which often results in repetitive hospital stays and reoperations for removal of the battery or even the electrodes. Prolonging battery life therefore reduces costs and improves the life of the dystonic patient. Conclusions We have not been able to conclude which is the optimal target of stimulation: the GPi or the STN. This inconclusiveness is because of the heterogeneity of patients, the relatively small number of patients, and the rather high percentage of patients with protocol interruptions. However, in this study, the STN was a safe and promising DBS target, which warrants further evaluation. Simultaneous stimulation of the STN and the GPi also seems safe and might add to the effect of STN stimulation alone. Disclosure Dr. Løkkegaard reports receiving fees for lecturing from Lundbech Pharma, UCB, and Medtronic within the last 3 years. Dr. Karlsborg reports a nonfinancial relationship with Medtronic, which supplies the equipment used in the author s DBS clinic. Author contributions to the study and manuscript preparation include the following. Conception and design: Hjermind, Madsen, Brennum, Løkkegaard, Karlsborg. Acquisition of data: Hjermind, Jensen, Karlsborg. Analysis and interpretation of data: Schjerling. Drafting the article: Schjerling. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Schjerling. Statistical analysis: Schjerling. Analyzing CT/MR images for localization of electrodes: Jespersen. Illustrations: Jespersen. Preoperative information and surgery: Jespersen, Madsen. Drafting the trial protocol: Brennum. References 1. Alexander GE, DeLong MR, Strick PL: Parallel organization of functionally segregated circuits linking basal ganglia and cortex. Annu Rev Neurosci 9:357 381, 1986 2. Anderson VC, Burchiel KJ, Hogarth P, Favre J, Hammerstad JP: Pallidal vs subthalamic nucleus deep brain stimulation in Parkinson disease. Arch Neurol 62:554 560, 2005 3. Andrews C, Aviles-Olmos I, Hariz M, Foltynie T: Which patients with dystonia benefit from deep brain stimulation? A metaregression of individual patient outcomes. J Neurol Neurosurg Psychiatry 81:1383 1389, 2010 4. Baizabal-Carvallo JF, Roze E, Aya-Kombo M, Romito L, Navarro S, Flamand-Rouvière C, et al: Combined pallidal and subthalamic nucleus deep brain stimulation in secondary dystonia-parkinsonism. Parkinsonism Relat Disord 19:566 568, 2013 (Letter) 5. Breakefield XO, Blood AJ, Li Y, Hallett M, Hanson PI, Stan- 1544

Subthalamic or pallidal DBS for dystonia daert DG: The pathophysiological basis of dystonias. Nat Rev Neurosci 9:222 234, 2008 6. Chou KL, Hurtig HI, Jaggi JL, Baltuch GH: Bilateral subthalamic nucleus deep brain stimulation in a patient with cervical dystonia and essential tremor. Mov Disord 20:377 380, 2005 7. Comella CL, Stebbins GT, Goetz CG, Chmura TA, Bressman SB, Lang AE: Teaching tape for the motor section of the Toronto Western Spasmodic Torticollis Scale. Mov Disord 12: 570 575, 1997 8. Coubes P, Cif L, El Fertit H, Hemm S, Vayssiere N, Serrat S, et al: Electrical stimulation of the globus pallidus internus in patients with primary generalized dystonia: long-term results. J Neurosurg 101:189 194, 2004 9. Fahn S, Bressman SB, Marsden CD: Classification of dystonia. Adv Neurol 78:1 10, 1998 10. Follett KA, Weaver FM, Stern M, Hur K, Harris CL, Luo P, et al: Pallidal versus subthalamic deep-brain stimulation for Parkinson s disease. N Engl J Med 362:2077 2091, 2010 11. Funkiewiez A, Ardouin C, Krack P, Fraix V, Van Blercom N, Xie J, et al: Acute psychotropic effects of bilateral subthalamic nucleus stimulation and levodopa in Parkinson s disease. Mov Disord 18:524 530, 2003 12. Grabli D, Ewenczyk C, Coelho-Braga MC, Lagrange C, Fraix V, Cornu P, et al: Interruption of deep brain stimulation of the globus pallidus in primary generalized dystonia. Mov Disord 24:2363 2369, 2009 13. Hammond C, Ammari R, Bioulac B, Garcia L: Latest view on the mechanism of action of deep brain stimulation. Mov Disord 23:2111 2121, 2008 14. Isaias IU, Alterman RL, Tagliati M: Deep brain stimulation for primary generalized dystonia: long-term outcomes. Arch Neurol 66:465 470, 2009 15. Jahanshahi M, Czernecki V, Zurowski AM: Neuropsychological, neuropsychiatric, and quality of life issues in DBS for dystonia. Mov Disord 26 (Suppl 1):S63 S78, 2011 16. Kiss ZH, Doig-Beyaert K, Eliasziw M, Tsui J, Haffenden A, Suchowersky O: The Canadian multicentre study of deep brain stimulation for cervical dystonia. Brain 130:2879 2886, 2007 17. Kleiner-Fisman G, Herzog J, Fisman DN, Tamma F, Lyons KE, Pahwa R, et al: Subthalamic nucleus deep brain stimulation: summary and meta-analysis of outcomes. Mov Disord 21 Sup pl 14:S290 S304, 2006 18. Kleiner-Fisman G, Liang GS, Moberg PJ, Ruocco AC, Hurtig HI, Baltuch GH, et al: Subthalamic nucleus deep brain stimulation for severe idiopathic dystonia: impact on severity, neuropsychological status, and quality of life. J Neurosurg 107: 29 36, 2007 19. Krause M, Fogel W, Kloss M, Rasche D, Volkmann J, Tronnier V: Pallidal stimulation for dystonia. Neurosurgery 55:1361 1370, 2004 20. Kupsch A, Benecke R, Müller J, Trottenberg T, Schneider GH, Poewe W, et al: Pallidal deep-brain stimulation in primary generalized or segmental dystonia. N Engl J Med 355:1978 1990, 2006 21. Lyons KE, Pahwa R: Effects of bilateral subthalamic nucleus stimulation on sleep, daytime sleepiness, and early morning dystonia in patients with Parkinson disease. J Neurosurg 104: 502 505, 2006 22. Nambu A, Tokuno H, Takada M: Functional significance of the cortico-subthalamo-pallidal hyperdirect pathway. Neurosci Res 43:111 117, 2002 23. Novak KE, Nenonene EK, Bernstein LP, Vergenz S, Cozzens JW, Rezak M: Successful bilateral subthalamic nucleus stimulation for segmental dystonia after unilateral pallidotomy. Stereotact Funct Neurosurg 86:80 86, 2008 24. Ostrem JL, Racine CA, Glass GA, Grace JK, Volz MM, Heath SL, et al: Subthalamic nucleus deep brain stimulation in primary cervical dystonia. Neurology 76:870 878, 2011 25. Paluzzi A, Belli A, Bain P, Liu X, Aziz TM: Operative and hardware complications of deep brain stimulation for movement disorders. Br J Neurosurg 20:290 295, 2006 26. Skogseid IM, Ramm-Pettersen J, Volkmann J, Kerty E, Dietrichs E, Røste GK: Good long-term efficacy of pallidal stimulation in cervical dystonia: a prospective, observer-blinded study. Eur J Neurol 19:610 615, 2012 27. Soeder A, Kluger BM, Okun MS, Garvan CW, Soeder T, Jacobson CE, et al: Mood and energy determinants of quality of life in dystonia. J Neurol 256:996 1001, 2009 28. Sun B, Chen S, Zhan S, Le W, Krahl SE: Subthalamic nucleus stimulation for primary dystonia and tardive dystonia. Acta Neurochir Suppl 97:207 214, 2007 29. Tisch S, Zrinzo L, Limousin P, Bhatia KP, Quinn N, Ashkan K, et al: Effect of electrode contact location on clinical efficacy of pallidal deep brain stimulation in primary generalised dystonia. J Neurol Neurosurg Psychiatry 78:1314 1319, 2007 30. Valldeoriola F, Regidor I, Mínguez-Castellanos A, Lezcano E, García-Ruiz P, Rojo A, et al: Efficacy and safety of pallidal stimulation in primary dystonia: results of the Spanish multicentric study. J Neurol Neurosurg Psychiatry 81:65 69, 2010 31. Vidailhet M, Vercueil L, Houeto JL, Krystkowiak P, Benabid AL, Cornu P, et al: Bilateral deep-brain stimulation of the globus pallidus in primary generalized dystonia. N Engl J Med 352:459 467, 2005 32. Vitek JL: Deep brain stimulation for Parkinson s disease. A critical re-evaluation of the STN versus the GPi DBS. Stereotact Funct Neurosurg 78:119 131, 2002 33. Voges J, Hilker R, Bötzel K, Kiening KL, Kloss M, Kupsch A, et al: Thirty days complication rate following surgery performed for deep-brain-stimulation. Mov Disord 22:1486 1489, 2007 34. Ware JE Jr, Sherbourne CD: The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care 30:473 483, 1992 35. Wichmann T, DeLong MR, Guridi J, Obeso JA: Milestones in research on the pathophysiology of Parkinson s disease. Mov D i s ord 26:1032 1041, 2011 36. Wöhrle JC, Blahak C, Capelle HH, Fogel W, Bäzner H, Krauss JK: Combined pallidal and subthalamic nucleus stimulation in sporadic dystonia-parkinsonism. Case report. J Neurosurg 116:95 98, 2012 37. Woehrle JC, Blahak C, Kekelia K, Capelle HH, Baezner H, Grips E, et al: Chronic deep brain stimulation for segmental dystonia. Stereotact Funct Neurosurg 87:379 384, 2009 38. Yianni J, Bain PG, Gregory RP, Nandi D, Joint C, Scott RB, et al: Post-operative progress of dystonia patients following globus pallidus internus deep brain stimulation. Eur J Neurol 10: 239 247, 2003 Manuscript submitted April 27, 2013. Accepted August 27, 2013. Please include this information when citing this paper: published online October 11, 2013; DOI: 10.3171/2013.8.JNS13844. Part of this work was presented in poster form at the Congress of the European Society for Stereotactic and Functional Neurosurgery, September 22 25, 2010, in Athens, Greece. Supplemental online information: Video: http://mfile.akamai.com/21490/wmv/digitalwbc.download. akamai.com/21492/wm.digitalsource-na-regional/jns13-844_ video_1.asx (Media Player). http://mfile.akamai.com/21488/mov/digitalwbc.download. akamai.com/21492/qt.digitalsource-global/jns13-844_video_1. mov (Quicktime). Address correspondence to: Lisbeth Schjerling, M.D., Department of Emergency, Hilleroed University Hospital, Dyrehavej 29, 3400 Hilleroed, Denmark. email: lisbeth.schjerling@regionh.dk. 1545