Article Text

Download PDFPDF

Research paper
Pallidal deep brain stimulation for DYT6 dystonia
  1. Fedor Panov1,
  2. Michele Tagliati2,
  3. Laurie J Ozelius3,4,
  4. Tania Fuchs4,
  5. Yakov Gologorsky1,
  6. Tyler Cheung3,
  7. Marat Avshalumov1,
  8. Susan B Bressman5,
  9. Rachel Saunders-Pullman5,
  10. Donald Weisz1,
  11. Ron L Alterman1
  1. 1Department of Neurosurgery, Mount Sinai School of Medicine, New York, New York, USA
  2. 2Department of Neurology, Cedars-Sinai Medical Center, Los Angeles, California, USA
  3. 3Department of Neurology, Mount Sinai School of Medicine, New York, New York, USA
  4. 4Department of Genetics and Genomic Sciences, Mount Sinai School of Medicine, New York, New York, USA
  5. 5Department of Neurology, Beth Israel Medical Center, New York, New York, USA
  1. Correspondence to Dr R L Alterman, Mount Sinai Medical Center, 1 Gustave L Levy Place, Box 1136, New York, NY 10029, USA; ron.alterman{at}


Background Mutations of the THAP1 gene were recently shown to underlie DYT6 torsion dystonia. Little is known about the response of this dystonia subtype to deep brain stimulation (DBS) at the internal globus pallidus (GPi).

Methods Retrospective analysis of the medical records of three DYT6 patients who underwent pallidal DBS by one surgical team. The Burke–Fahn–Marsden Dystonia Rating scale served as the primary outcome measure. Comparison is made to 23 patients with DYT1 dystonia also treated with GPi-DBS by the same team.

Results In contrast with the DYT1 patients who exhibited a robust and sustained clinical response to DBS, the DYT6 patients exhibited more modest gains during the first 2 years of therapy, and some symptom regression between years 2 and 3 despite adjustments to the stimulation parameters and repositioning of one stimulating lead. Microelectrode recordings made during the DBS procedures demonstrated no differences in the firing patterns of GPi neurons from DYT1 and DYT6 patients.

Discussion Discovery of the genetic mutations responsible for the DYT6 phenotype allows for screening and analysis of a new homogeneous group of dystonia patients. DYT6 patients appear to respond less robustly to GPi-DBS than their DYT1 counterparts, most likely reflecting differences in the underlying pathophysiology of these distinct genetic disorders.

Conclusions While early results of pallidal DBS for DYT6 dystonia are encouraging, further research and additional subjects are needed both to optimise stimulation parameters for this population and to elucidate more accurately their response to surgical treatment.

Statistics from


Primary generalised dystonia is a movement disorder characterised by patterned muscle contractions resulting in abnormal motion and postures in the absence of other neurological signs or structural brain abnormalities.1 2 The pathophysiology underlying primary dystonia is largely unknown. Seven primary monogenic dystonias have been classified2–4 and five gene loci have been mapped using genetic linkage studies; however, until recently only the gene responsible for the most common of these inherited dystonias (DYT1) was isolated.5 In 2009, Fuchs et al demonstrated that mutations of the gene encoding thanatos associated protein domain containing apoptosis associated protein 1 (THAP1) underlie the DYT6 phenotype, establishing a second group of genetically distinguishable primary dystonia patients.6 The THAP1 protein is known to interact with Par-4, a well characterised pro-apoptotic factor, previously linked to prostate cancer and neurodegenerative diseases.7 Mutations in the THAP1 gene may affect these interactions and alter apoptotic activity but it remains unclear if and how this may cause dystonia.

The DYT6 phenotype, which was first described in a large Mennonite family, is inherited in an autosomal dominant fashion with a penetrance approaching 60%.8 The average age at symptom onset is 16 years. Roughly half of patients initially experience cranial and cervical symptoms with secondary spread to the extremities. In the other half of patients the pattern is reversed, with symptoms starting in the arm.8 The DYT6 phenotype differs from DYT1 in its older average age at symptom onset, milder leg involvement and more severe craniocervical involvement.5

While medical therapy for primary dystonia is largely unsatisfactory,1 2 9 deep brain stimulation (DBS) at the internal globus pallidus (GPi) has been demonstrated to be a safe and effective intervention for this heterogeneous patient population.9–12 To date, the only clinical factors known to predict response to pallidal DBS among primary dystonia patients are: (1) age <20 years; (2) disease duration <15 years; and (3) absence of fixed skeletal deformities at the time of surgery.1 While the presence of the DYT1 mutation was initially thought to be an independent positive outcome predictor, recent studies suggest that the consistently positive response to pallidal DBS observed in DYT1 patients may be a reflection of their young age and short symptom duration at the time of surgery rather than their underlying genetic mutation per se.1

When genetic testing for the THAP1 mutation became available, we screened the stored DNA samples of the non-DYT1 primary dystonia patients on whom we had performed pallidal DBS in the past. Three patients were found to harbour a THAP1 mutation. The following is a retrospective analysis of the response of these three genetically proven DYT6 patients to pallidal DBS.



The three THAP1/DYT6 patients included in this study received bilateral pallidal DBS therapy by one clinical team (RLA and MT) at either the Beth Israel Medical Center (patient No 1) or the Mount Sinai Medical Center (patient Nos 2 and 3), both in New York City. Their respective THAP1 mutations were characterised as described in a previous report by Bressman and colleagues.5 All three DYT6 patients were followed for a minimum of 3 years after surgery. Clinical results in a consecutive cohort of 23 DYT1 positive patients treated with pallidal DBS by the same clinical team and also followed for a minimum of 3 years are presented for comparison. Demographic data for the DYT6 patients are provided in table 1.

Table 1

Baseline characteristics of the DYT6 patients

Indications for surgery

In accordance with the US Food and Drug Administration Humanitarian Device Exemption regulating the use of DBS in dystonia (HDE# H020007), the institutional review boards of either Beth Israel Medical Center (2003–2004) or the Mount Sinai School of Medicine (2005–2010) approved the requisite consent forms and monitored surgical results.

Each of the 26 patients included in this study were diagnosed with primary dystonia based on the following criteria: (1) dystonia was the sole neurological abnormality identified via clinical examination; (2) there was no known history of significant head trauma, birth injury, meningitis/encephalitis or other known causes of secondary dystonia; and (3) the patient's brain MRI revealed grossly normal anatomy.1 9 A neurologist specialising in the diagnosis and treatment of movement disorders (MT, SBB, RSP) evaluated each patient prior to surgery. The neurologist confirmed the diagnosis of primary dystonia and insured that the level of disability was appropriate for surgical intervention and that medical therapy had been optimised.

Surgical technique and device programming

Each patient underwent frame based, microelectrode guided, stereotactic implantation of DBS leads (model 3387; Medtronic Inc, Minneapolis, Minnesota, USA), as previously described.13 Proper lead location was confirmed with postoperative MRI in all cases. The anatomical position of the implanted leads relative to the mid-commissural point (MCP) was determined by merging the preoperative and postoperative MRIs for each implant procedure with the Stealth Framelink Software, V.5.0 (Medtronic Inc, Boulder, Colorado, USA). As per the senior author's routine, the preoperative images were already reformatted orthogonal to the intercommissural plane for the purposes of targeting. After ensuring an accurate merge of the postoperative MRI to the reformatted targeting images, the cursor was placed on the image of the deepest contact and the coordinates of that point relative to the MCP were recorded.

The pulse generators were implanted 7–14 days following the lead implantation procedure. Therapeutic stimulation commenced 1–2 weeks after that, allowing for proper wound healing. The stimulator settings were adjusted in order to maximise clinical benefit and minimise side effects, employing a previously established programming algorithm.14 15

Microelectrode recordings

The microelectrode recording data presented in this report are derived from a larger analysis of globus pallidus recordings in dystonia patients during DBS surgery that is being prepared for publication. Microelectrode recordings were performed with a platinum–iridium electrode (FHC Inc, Bowdoinham, Maine, USA) conditioned to ∼0.4–0.7 MΩ at 1000 Hz. The recordings were amplified with a gain of 6000–8000, bandpass filtered (300–6000 Hz), digitised at a 12 kHz sampling rate and stored for later offline analysis. The MER data were analysed employing Spike 2 software (Cambridge Electronic Design Ltd (CED), Cambridge, UK).

The recordings were processed initially with spike sorting tools in order to extract single unit recordings for analysis. We accepted for analysis only single unit recordings of consistent spike shape and width lasting longer than 10 s with a signal to noise ratio greater than 2:1. To evaluate the firing activity of GPi units, the median and mean firing rates were measured during a stable stationary period of recording. The assessment of general firing patterns included the following parameters: burst index (BI), coefficient of variation and auto-correlograms.16–20 A burst was defined as a period with a high number of discharges (more than 3 spikes) separated by periods of reduced or absent activity.21–23 The firing pattern of each cell was classified as regular, irregular or burster based on the BI as follows: BI <1.70 = regular; BI >1.7 and <9.9 = irregular; and BI >10 = burster.

Clinical data collection and statistical analysis

The presented data are derived via retrospective review of the office charts and/or electronic medical records of the 26 patients (three DYT6; 23 DYT1). The review was performed with the approval of the Mount Sinai IRB (protocol No 05-0589). In addition to basic demographic information (table 1), we recorded the motor and disability subscores of the Burke–Fahn–Marsden Dystonia Rating Scale (BFMDRS), as determined by the evaluating neurologist within 1 week of surgery (baseline) and 1, 2 and 3 years after commencement of stimulation. Stimulator settings, microelectrode recording data and dystonia related medications of the three DYT6 patients at baseline and at their last clinical follow-up were also recorded.

Statistical analyses were performed with a dedicated software package (SPSS V.17.0). Due to the small number of patients in the DYT6 group, we assumed the data not to be normally distributed and employed statistical tests for non-parametric data (the Mann–Whitney U test for continuous variables and the χ2 test for categorical data). Statistical tests were two tailed and p values <0.05 were considered to be statistically significant.


Clinical summaries

Patient No 1, a man, was referred for DBS at age 15 years after suffering with dystonia for 12 years. He was previously reported in Bressman and colleagues5 in family No 2 with a p.Phe45fs73X mutation. He began speaking with a ‘strangled’ voice at age 3 years. He later developed difficulty writing, a right lateral head tilt and anterocollis. The patient initially responded to trihexyphenidyl, baclofen, bromocriptine and carbidopa/levodopa. Botulinum toxin injections provided no benefit. On initial examination the patient exhibited slurred speech, dystonic movements of his left arm and leg, and severe torticollis with predominantly right lateral flexion and right rotational components.

Immediately after device activation, his left foot inversion improved, as did his cervical dystonia. His baseline BFMDRS motor score of 40 improved to 34 1 month postoperatively and was 13 at 6 months postoperatively. The score then increased to 20 by the third postoperative year. By the fourth postoperative year, his function had returned to preoperative levels, with a BFMDRS score of 41. His left-sided symptoms were particularly severe and monopolar screening of the right device activated the corticospinal tract at just 3.5 V. A repeat MRI revealed that the right DBS lead was positioned 2 mm anterior and slightly medial to what we then considered to be optimal so it was replaced with a new lead that was positioned 2 mm posterior and 1 mm lateral to the original. Control of the left side was restored and his gait improved. As of this writing, the patient's motor function has stabilised with decreased neck pulling and left foot inversion. He continues to have difficulties with speech, torticollis and dragging of the left foot during prolonged ambulation. His latest BFMDRS motor and disability scores are 27 and 6, respectively. There were no surgical, stimulation related or hardware related complications.

Patient No 2, a woman, was referred for DBS at age 36 years having suffered with dystonia from age 25 years. She harbours a 2 basepair deletion that removes the start codon of THAP1.5 Her symptoms began with cramping in the left arm and progressed to include dysarthria and action tremor in the right arm. Two years prior to presentation, she developed left foot inversion. A variety of medications yielded no significant improvement. Injections of botulinum toxin provided some relief but were stopped due to resultant hand weakness.

Examination revealed dystonic posturing predominantly affecting the left side. Right pallidal stimulation alleviated pain and spasms in the left foot and hand. Her BFMDRS motor score improved from 20.5 at baseline to 15, after which it worsened to 19. The patient requested implantation of a contralateral system, which was performed approximately 1 year after the first implant and yielded significant functional improvement. Currently, her symptoms are well controlled although she notes worsening of hand dystonia with heavy use. She also has frequent headaches and mildly slurred speech. Her BFMDRS motor scores were held below 10 for over 2 years following the second implant; however, at her 3 year follow-up visit, the patient complained of worsening fatigue, cramping, increased difficulty using her arms and greater stress at work. Her motor score was 16.5. The disability score, which improved from 5 at baseline to 3, has also increased to 8 during this recent exacerbation. There have been no surgical or therapy related complications.

Patient No 3, a man, presented at age 23 years after a 17 year history of dystonia. His symptoms began at age 6 with curling of the right index finger and progressed to include left foot inversion/plantar flexion, jaw tightness causing some speech difficulty and dystonic posturing of the trunk and neck. He could not ambulate more than 50–75 feet and employed a wheelchair outside his home. Carbidopa/levodopa and carbamazepine yielded minimal benefit. Botulinum toxin injections to the arm for finger flexion provided moderate relief. On examination, he exhibited a guttural, tremulous speech, right shoulder elevation and dysdiadochokinesia of the hands. No fixed skeletal deformities were appreciated. He has a novel mutation, c.161G>A, resulting in a p.C54Y substitution.

Subjective improvement was noted following the commencement of bilateral pallidal stimulation. His hands moved more freely, and his tremors dissipated. Involuntary movements of the torso decreased and left foot plantar flexion disappeared. His ambulation and voice improved modestly. His BFMDRS motor score improved from 77.5 at baseline to 52.5 after 1 month of stimulation and 32.5 at 6  months. Over the next 18 months the motor score trended down to 21. His disability score fell from 19 at baseline to 10 at 1 year and 7 at 3 years postoperatively. The patient was able to discontinue his preoperative medications but a recent mild worsening of his right hand flexion prompted the initiation of Klonopin (0.5 mg daily). There were no surgical or device related complications.

Clinical results

Figure 1 exhibits the individual BFMDRS motor and disability scores for the three DYT6 patients at baseline and at each annual follow-up. For the purposes of comparison, the mean scores (±SEM) for 23 consecutive DYT1 patients, also treated by our team with pallidal DBS for a minimum of 3 years, are provided. After 1 year of stimulation, the BFMDRS motor scores were improved to a similar degree in both the DYT1 (median 80% (range 44–97%)) and DYT6 groups (median 66% (range 60–72%); p=0.12, Mann–Whitney U test); however, after 2 years of stimulation the DYT1 cohort exhibited continued improvement (median improvement 87% (range 54–100%)) while the DYT6 group exhibited modest symptom regression with an overall median improvement of 61% (range 57–71%). The difference in outcomes between the two groups was statistically significant at this time point (p<0.05). After 3 years of stimulation, the difference in outcomes was even greater due to continued improvement in the DYT1 group (median improvement 90% (range 65%–100%)) and further regression in the DYT6 patients (median improvement 50% (range 20–74%); p<0.01).

Figure 1

Burke–Fahn–Marsden Dystonia Rating Scale (BFMDRS) scores in DYT6 dystonia. The BFMDRS motor (BFMDRS-M) (A) and disability (BFMDRS-D) (B) scores for the three DYT6 patients are presented at baseline and at each annual follow-up for their first 3 years of therapy. The mean (±SEM) BFMDRS scores of 23 DYT1 patients treated by the same clinical team and followed for a minimum of 3 years are presented for comparison. During the first year of therapy, the DYT6 patients exhibited improvements in both their motor and disability scores that were similar to those observed in the DYT1 cohort. However, in years 2 and 3, the DYT6 patients regressed while the DYT1 patients exhibited continued improvement. The presented data reflect the response of patient No 1 following his initial DBS surgery.

Changes in the disability scores demonstrated a similar pattern. After 1 year of stimulation the DYT1 group exhibited a median 65% improvement in their disability scores (range 33–94%) compared with a median 40% improvement (range 30–47%) in the DYT6 group (p<0.05). After 2 years of stimulation, the DYT1 group exhibited continued improvement (median improvement 79% (range 33–100%)) while the DYT6 group improved only marginally (median improvement 40% (range 30–63%); p<0.05). After 3 years of stimulation, the DYT1 group improved even further (median improvement 82% (range 33–100%)) while the DYT6 patients regressed, with one patient's scores worse than her preoperative scores (median improvement 30% (range −40 to 63%; p<0.05).

In order to determine if the response to DBS varied anatomically among the three DYT6 patients, we divided their respective BFMDRS motor scores into appendicular, truncal and craniocervical subscores (figure 2A,B,C). Examining the data in this manner, we noted that craniocervical symptoms improved for all three patients. Truncal symptoms emerged in the first patient during therapy, were stably absent in the second patient and were alleviated in the third patient. Extremity scores improved in the first and third patients and remained stable in the second patient.

Figure 2

Baseline motor scores compared with latest follow-up of patient Nos 1, 2 and 3 (A, B, and C, respectively) broken down by categories. The data for patient No 1 reflect his latest level of function following his revision surgery, which occurred in year 4. C/C, craniocaudal.

Medication reductions were realised in two of the three DYT6 patients. Only a minimal increase in the dosage of a single medication (clonazepam) was required in the third patient.

Lead placement and latest stimulator settings

The anatomical locations of the deepest contact (ie, contact 0) relative to the MCP are provided in table 2 and the patients' latest stimulator settings are shown in table 3. The lead tip locations, as measured on the immediate postoperative scans, indicate that the leads were all implanted within the GPi, as intended.

Table 2

Anatomical location of the implanted leads

Table 3

Stimulator settings

Microelectrode recording data

The firing characteristics of GPi cells recording during the DBS procedures for the DYT6 patients are shown in table 4. Seventeen cells met our criteria for analysis. With reference to the mean firing rates, burst indices and coefficients of variation, GPi cells in the DYT6 patients were statistically indistinguishable from GPi cells in DYT1 patients (unpublished results, D Weisz, 2011).

Table 4

Firing characteristics of GPi cells in DYT6 dystonia


Primary generalised dystonia is an uncommon disabling disorder with few effective medical therapies. The identification of specific genetic anomalies that are associated with the disorder allows it to be further categorised for the purposes of prognostication and treatment and may provide important clues regarding the molecular and cellular mechanisms that underlie the dystonia phenotype. As a group, patients with primary generalised dystonia respond robustly to pallidal DBS with larger series reporting improvements in motor function that range from 50% to 75%, as measured with the BFMDRS10 24 25; nevertheless, not all patients benefit to the same degree. Some reports have suggested that patients possessing the DYT1 mutation may respond best to DBS but analyses by Isaias and colleagues1 and Andrews and colleagues26 suggest that the response is best predicted by age and disease duration at the time of surgery, as well as a lack of fixed skeletal deformities.

The recent discovery that mutations in the THAP1 gene underlay the DYT6 phenotype presented the possibility of evaluating the effects of pallidal DBS in this dystonia subtype. Our limited experience suggests that DYT6 patients initially respond to pallidal stimulation much like the DYT1 patients; however, after 1–2 years of stimulation, the DYT6 patients seemed to regress, requiring adjustments to their stimulation parameters and in one case an adjustment in lead position. Examining the responses anatomically, it seems that the response to DBS in these patients is more variable with respect to the extremities, while the truncal and craniocervical responses appear more stable. Our results confirm three prior reports that pallidal DBS is mildly to moderately effective for DYT6 torsion dystonia.27–29 Djarmati and colleagues27 reported a modest but sustained response to DBS in one DYT6 patient. Groen and colleagues28 reported five DYT6 patients in whom bilateral pallidal DBS yielded improvements ranging from 16% to 55%. Similarly, Zittel and colleagues29 reported ‘only moderate improvement’ from bilateral pallidal DBS in two additional DYT6 patients.

The modest responses we have observed do not appear to be due to poorly positioned leads, as determined on the immediate postoperative scans (table 2). Since we have not scanned these patients more recently, we cannot definitively rule out the possibility that changes in lead position account for the clinical regression; however, we have not observed a similar regression in any of the 23 DYT1 patients and it seems highly improbable that this would occur only and in all three of the DYT6 patients. It is more likely that either the underlying pathophysiology of DYT6 dystonia and/or the stimulation parameters we are employing for therapy account for these more modest responses. Previously published positron emission tomography studies by Carbon and colleagues30 31 demonstrate both similarities and significant differences in the pathophysiologies of DYT1 and DYT6 torsion dystonia. Most significantly, these studies demonstrate decreases in striatal D2 receptor availability in both disorders that is more severe in the brains of DYT6 patients. These physiological differences between the two disorders may account for both the differing clinical presentations and the responses to pallidal DBS that we and others have observed.

Interestingly, we found no differences in the firing characteristics of DYT6 GPi neurons compared with DYT1, although only a small number of cells were analysed. Zittel and colleagues29 also reported MER data from the two DYT6 patients in whom they performed DBS implants. They also reported no difference in the firing characteristics of pallidal neurons in DYT6 compared with DYT1 patients. Of note, however, is that for both groups they reported significantly lower mean firing rates than we observed. We believe this difference is explained by the fact that Zittel et al performed their recordings with their patients under general anaesthesia while our patients were fully awake. Consequently, we believe our GPi recording data more accurately reflect the physiology of GPi neurons in DYT6 dystonia.

Finally, two of our three patients were being stimulated at a frequency of 80 Hz. In the past, we have reported on the successful use of both 60 Hz and 80 Hz stimulation in various forms of primary dystonia.15 32 It is possible, however, that due to differences in its underlying pathophysiology, THAP1/DYT6 dystonia patients may require the higher stimulation frequencies (ie, >130 Hz) commonly employed for the treatment of other movement disorders and utilised by the authors of the other three papers. It should also be noted that all three of our DYT6 patients had been symptomatic for a relatively long period of time (11–17 years) before undergoing DBS surgery, a clinical variable that we have previously demonstrated may negatively impact on the response of dystonia patients to pallidal DBS.1


This report brings to 11 the number of patients with THAP1/DYT6 torsion dystonia whose responses to pallidal DBS are reported in the literature. Thus far, the data suggest that DYT6 patients do improve with pallidal stimulation although the results may not be as robust as those observed in patients with DYT1 dystonia. Moreover, stimulation frequencies <130 Hz may not be as effective in this subgroup, as has been observed in DYT1. Longer follow-up in greater numbers of patients is required to understand the effects of pallidal DBS in these patients and to identify optimal stimulation parameters.


The authors would like to acknowledge Joseph Rudolph's assistance with the preparation of this manuscript.



  • Funding This study was supported in part by the Bachmann-Strauss Foundation for Dystonia and Parkinson's disease (RLA, LJO, SBB, RSP) as well as the Dystonia Medical Research Foundation (TF).

  • Competing interests RLA and MT receive occasional teaching honoraria from Medtronic Inc.

  • Patient consent Obtained.

  • Ethics approval This study was conducted with the approval of the institutional review boards of Beth Israel Medical Center and Mount Sinai School of Medicine.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.