Article Text

Download PDFPDF

Deep brain stimulation in 18 patients with severe Gilles de la Tourette syndrome refractory to treatment: the surgery and stimulation
  1. D Servello1,
  2. M Porta2,
  3. M Sassi1,
  4. A Brambilla2,
  5. M M Robertson3
  1. 1
    Neurosurgical Division, Functional Neurosurgery Unit, Istituto Galeazzi IRCCS, Milano, Italy
  2. 2
    Department of Neurology, Tourette Centre, Istituto Galeazzi IRCCS, Milan, Italy
  3. 3
    Institute of Neurology, University College, London, UK
  1. Dr D Servello, Neurosurgical Division, Istituto Galeazzi IRCCS, via Galeazzi 4, 20161 Milano, Italy; servello{at}


Background: There have been several reports of successful deep brain stimulation (DBS) for the treatment of severe Gilles de la Tourette syndrome (GTS).

Method: 18 cases of GTS who were resistant to at least 6 months of standard and innovative treatments, as well as to psychobehavioural techniques, underwent DBS. DBS was placed bilaterally in the centromedian–parafascicular (CM–Pfc) and ventralis oralis complex of the thalamus. Patients were evaluated after surgery, with immediate and formal assessments at least every 3 months, including “on–off” and “sham off” in the first nine patients.

Results: All patients responded well to DBS, although to differing degrees. The duration of follow-up assessments ranged from 3 to 18 months. The comorbid symptoms of obsessive–compulsive behaviour, obsessive–compulsive disorder, self-injurious behaviours, anxiety and premonitory sensations decreased after treatment with DBS. There were no serious permanent adverse effects.

Conclusions: DBS is a useful and safe treatment for severe GTS. The results of ours and previous DBS reports suggest that the CM–Pfc and ventralis oralis complex of the thalamus may be a good DBS target for GTS.

Statistics from

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.

Gilles de la Tourette syndrome (GTS) is characterised by multiple motor tics and one or more phonic (vocal) tics lasting longer than 1 year.1 2 Premonitory sensations occur in the majority of cases, typically tics wax and wane, are suppressible and suggestible, and approximately 90% of patients have comorbid disorders, including attention deficit hyperactivity disorder, obsessive–compulsive disorder (OCD) and behaviours (OCB), self-injurious behaviours, depression, anxiety and non-obscene socially inappropriate behaviours. The pharmacological treatment of GTS usually includes typical and atypical neuroleptics, clonidine, botulinum toxin injections3 and more recently behavioural methods.

The basal ganglia are involved in models of the pathophysiology of this disorder. The basal ganglia interconnect the network of cortico-striato-pallido-thalamic structures, and there are suggestions that the pathophysiology of GTS is within these areas.410 Pathology at any one level of the circuitry alters activity at all other levels of the circuit “loop”. However, no absolute areas have been identified as consistently abnormal in GTS.1113 One study used positron emission tomography techniques, combined with time synchronised audio and videotaping of tics to demonstrate that simple motor tics appear to be associated with the sensorimotor cortex, while more complex tics (eg, coprolalia, vocal tics) were associated with activity in pre-rolandic and post-rolandic language regions, insula, caudate, thalamus and cerebellum.14 The thalamus has also been suggested to be involved in the pathophysiology of GTS from neuroimaging15 and experimental evidence,1618 stereotactic lesions1921 and deep brain stimulation (DBS) surgery,22 23 which have targeted the thalamus.

DBS has relatively recently been reported to be successful in GTS and its associated disorders. The Dutch–Flemish Group treated four patients (n = 1, internal globus pallidus (GPi); n = 3, bilateral thalamic22 2426) while others have conducted DBS as follows: Flaherty et al, n = 1, anterior limb of the internal capsule, terminating in the area of the nucleus accumbens27; Diederich et al, n = 1, GPi28; Houeto et al, n = 1, GPi and centromedian–parafascicular complex (CM–Pfc)23; Priori et al, n = 1, CM–Pf bilateral thalamic29; and Sturm, n = 3, nucleus accumbens.30 DBS has also been used in disorders occurring frequently in association with GTS (eg, OCD,31 anxiety32 and depression33).

We treated 18 GTS patients with DBS who were severely affected and refractory to treatment. Based on the fact that the presumed pathophysiology of GTS is within the cortico-striato-pallido-thalamic circuits and that alterations at any level of the circuitry would be predicted to modify activity throughout the circuit, the CM–Pfc and the ventral oral thalamic nuclei (Voa) were the chosen targets (CM–Pfc–Voa).


Patients were recruited in our Tourette Clinic. Eighteen patients (aged 17–47 years; mean 28.4 (SD 8.9) years; 15 males) satisfying DSM-IV-TR1 and World Health Organisation2 criteria for GTS, with chronic marked to severe tics refractory to treatment, were included in the DBS protocol. The duration of their GTS symptomatology at DBS was 9–39 years (table 1).

Table 1 Demographic data of 18 patients with Gilles de la Tourette syndrome who had deep brain stimulation in the centromedian–parafascicular and ventralis oralis complex of the thalamus

All patients were targeted at the CM–Pfc–Voa nuclei (follow-up 3–19 months). The demographic data and diagnostic confidence index (DCI) scores34 of the patients are shown in table 1. DCI was calculated by two investigators separately (MP and AB). The scores of the more senior (MP) were used in table 1. The range of DCI scores of all patients (AB) was 52–100% (mean 77.8 (SD 15.26)%); the range of scores calculated by MP was 53–100% (mean 80.2 (SD 15.4)%). The correlation was highly significant (Pearson’s r = 0.90).

Neuropsychiatric assessment methods and criteria

Inclusion criteria were: (1) satisfying diagnostic criteria for GTS; (2) resistance to at least 6 months of conservative treatment with standard and innovative treatments (for medication details see table 1); (3) GTS severity resulting in disability considered to substantially reduce the patient’s quality of life (QOL); (4) it was established that the tic symptomatology was not caused by any other condition; and (5) all patients were deemed medically fit to withstand the operation and possible sequelae.

Assessment included full history and examination, blood analysis and evaluations using standardised assessment schedules, including the Yale Global Tic Severity Rating Scale (YGTSS)35 and a battery of neuropsychological and neuropsychiatric schedules and cognitive tests before and after DBS.

Patient’s impairments were also considered. Two patients had cervical myelopathy as a direct consequence of violent tics of the head and neck. One patient (No 12) developed cervical disc herniation at C4–C5 requiring spinal surgery 6 months after successful DBS. Another patient (No 15) developed incapacitating cervical myelopathy at the C4–C5 level, and was operated on 6 years prior to DBS. Two patients (Nos 8, 17) were unable to eat unaided because of severe tics in the arms and head. Prior to DBS only 7/18 patients were able to work independently.

Exclusion criteria included psychosis, cognitive impairment, suicide attempts, drug addiction and poor compliance with medication. Among eight patients selected for DBS, three were excluded (poor understanding of surgery, suicide attempt, poor compliance, addiction) and a further five were offered DBS but they or their families declined.


All patients were videotaped before and after DBS. Additional recordings were performed during follow-up if the patients’ symptoms increased or decreased markedly.

Surgical procedural techniques

Preoperatively, an MRI scan of the brain (3 mm thickness, with T1 weighed axial and sagittal slices and T2 weighed coronal slices) and a CT scan of the brain (after positioning of the stereotactic frame) were processed by a neuronavigator (Treon; Medtronic, Minneapolis, Minnesota, USA) to obtain the coordinates of the nuclei according to the Schaltenbrand–Wahren atlas.36

Standard antibiotic therapy and benzodiazepines were given prior to surgery.

A pre-coronal drill hole of 14 mm in diameter was made under local anaesthetic. Three channel simultaneous micro recording (high impedance, tungsten bipolar—Inomed MER system) started from 6 mm above the target point and ended 1.5 mm below the target. Micro recording was performed every 0.5 mm, 1–2 min each, with the following parameters: 500–5000 MHz filters, cut-off 200 μV/div, 100 ms/div sweep, obtaining a “map” of neuronal activity, on the base of which macro stimulation was made (100 Hz frequency and a 60 μs interval, starting at 1 mAmp and gradually increasing until the threshold of 5 mAmp was reached). All subjective impressions reported by the patients were carefully recorded. The choice of the trajectory of the definitive stimulating electrode (model 3387; Medtronic) was then made on the basis of these results.

Ten patients were treated under general anaesthetic either because of the severity of tics or because they were unable to remain immobile. In these patients, both surgical interventions were performed under a single general anaesthetic. An MRI scan of the brain (coronal slices, 3 mm thickness, T2 weighted) demonstrated the correct positioning of the electrodes. All patients maintained their routine drug therapeutic regimen throughout surgery.

The pulse generator (Kinetra; Medtronic) was positioned in a subcutaneous premuscular subclavicular fashion for the first three patients and in the abdominal subcutaneous layers for subsequent patients because of aesthetic concerns.

Follow-up protocol

Patients were admitted to the Tourette Clinic for post-procedural management after 1 day of clinical observation following activation of the Kinetra stimulator. Patients were checked 1 week after discharge, and then monthly, provided no worsening of the clinical picture occurred. Follow-up included a visit to the clinical multidisciplinary team, care giver interview and adjustment of drug therapy and neuroelectrical parameters if and when required. Part of the follow-up evaluation included a “switch on–off” pulse generator, performed after 6 months of follow-up. The “switch off” period lasted 48–72 h during which the patients were admitted to the department.

Statistical analysis

Data were examined using SPSS (SPSS inc. V.11.5 Microsoft Windows); t tests were used to determine significant differences between the means of the patient scores (YGTSS) before and after DBS. A p value <0.05 was considered statistically significant.


Duration of follow-up

The patients were followed-up clinically for 3–18 months (table 1); the last assessments were completed at 17 months (table 2).

Table 2 Yale Global Tic Severity Rating Scale scores before, and the last score after, deep brain stimulation (including follow-up, ranging from 2 to 18 months)

Tic severity

All 18 patients responded well to DBS. The effects on tics and social impairment are summarised in table 2 and fig 1. All four components of the YGTSS scores improved significantly. In six patients the improvement in tic symptomatology was progressive, with only a few modifications of the stimulation parameters (Nos 3, 4, 7, 9, 11, 12) and sustained improvement at follow-up. The remaining 12 patients showed recurrent motor and phonic tics after DBS, requiring several adjustments of the stimulation parameters. In three cases (the younger patients (Nos 5, 8, 14) and one other (No 16)), the results were less satisfactory after 3–6 months of follow-up, with significant spontaneous “waxing and waning” of symptoms. ANOVA statistical analysis was performed. No difference between motor and phonic tics was observed (p = 0.339).

Figure 1 Effects of deep brain stimulation on tics, as measured by Yale Global Tic Severity Rating Scale (YGTSS) scores.

Intraoperative results and findings

For all patients an irregular (20–40 Hz) discharge was recorded by the three probes, with a bi-tri-phasic shape, occurring mainly between 2 and 7 mm above the target, and then becoming progressively less evident.

Eight patients who underwent local anaesthesia had macro stimulation: these latter eight reported a sensation of “well being” produced between 5 and 2 mm above the target. All eight patients experienced different sensations during macro stimulation at the three different tracks, between 5 mm and the target, ranging from no side effects to significant drowsiness and/or agitation. We used a three track recording, and chose the central one because of minimal side effects. Activity patterns of that target were recorded intraoperatively.

Neuroradiological results

We verified lead positioning with 3 mm section inversion recovery axial MRI, superimposing on and comparing with the Shaltenbrand–Wahren atlas.36 The first 14 patients were studied with 1.5 T control inversion recovery MRI, demonstrating the correct positioning of the electrodes (figs 2, 3).

Figure 2 Postoperative inversion recovery axial MRI, superimposed on the Schaltenbrand–Wahren Atlas.36 Electrode tips at the intercommissural plane are shown (white circles). H1+H2, campus forelii; Vo, nucleus ventrooralis medialis.
Figure 3 Postoperative inversion recovery axial MRI, superimposed on the Schaltenbrand–Wahren Atlas.36 Slice taken 2 mm above the intercommissural plane. White circles, electrodes; VO, nucleus ventrooralis; CM, nucleus centralis magnocellularis; Pf, nucleus parafascicularis.

Stimulation parameters

The protocol stipulated that clinic visits occurred every month in the post-DBS period. However, patients were seen more often according to individual needs. Stimulation parameters were modified on the basis of fluctuations in the tic symptomatology of GTS. The frequency of the stimuli was 130 Hz in all patients, and pulse width was maintained within the range 60–120 μs. The pattern of stimulation and amplitude varied (2.5–4 V) according to the symptomatology. The initial electrode setting was 0–2+, 4–6+, amplitude 2.5 V. Seven of the 13 patients (53%) who were followed-up for 6 months or more (Nos 1, 2, 5, 8, 9,10, 13) required more frequent modifications: two (Nos 5, 13) were stimulated with a monopolar setting (1- case+) and required an increase in amplitude reaching up to 4 V at the last follow-up.

Minor side effects occurred with the parameter regulations when the intensity was >4 V. Transient subjective vertigo occurred in almost all patients. Transient blurring of vision occurred in four patients (Nos 1, 5, 8, 16) and abdominal discomfort in two patients. One patient developed an upward ocular deviation at 3 V (No 17); of note is that this patient had similar side effects with neuroleptics. Table 3 indicates the stimulation parameters at the last follow-up visit.

Table 3 Stimulation parameters at the last follow-up visit

Surgical adverse effects

Two patients developed adverse side effects after DBS. The first, a 34-year-old male (No 9), developed poor healing of his scalp scar as he felt compelled to touch it repetitively. He underwent plastic surgery to repair his scalp skin; also, his arms and chest were placed in a cast to prevent him from touching the scar. This occurred again and he required a second similar cast, and again he improved and subsequently made a good recovery. A 46-year-old male (No 12) developed an abdominal haematoma where the pulse generator was located. The haematoma was evacuated, and he improved within a few days.

Post-procedure pharmacological treatment

Three patients (Nos 4, 10, 12) required no medication after DBS. The remaining 16 patients required a reduced amount and type (25–50%) of their pre-DBS medications. After DBS, two patients required vocal cord infiltration with botulinum toxin for severe phonic tics.


Videotapes of 17/18 patients (excluding No 14) showed a marked improvement in tic severity. In many cases, on video, patients volunteered that their OCB and anxiety were reduced. In one patient (No 1), walking tics remained.

“On–off” and “sham off” findings

For motor and phonic tics and psychobehavioural symptoms, after 9 months a short “on–off” evaluation was performed with a blinded rater in nine patients (Nos 1–9). Eight patients (except No 4) showed varying degrees of deterioration in the “off” condition, including tic reappearance, development of severe anxiety and return of premonitory sensations and obsessional symptoms; with “sham off” the patients became anxious.


DBS was performed in 18 patients with GTS who were severely affected and refractory to medication and psychobehavioural techniques, and in whom their quality of life was substantially reduced. All 18 responded well to DBS although to differing degrees and all had bilateral DBS targeted at the Vo–CM–Pfc. The safety profile was good (only 2/18 with “surgical” adverse effects). The majority had some minor side effects with the parameter regulations when the intensity was >4 V (transient subjective vertigo, transient blurring of vision, abdominal discomfort and in one an upward ocular deviation at even 3 V). The duration of follow-up with assessments ranged from 3 to 17 months. The comorbid symptoms of OCB, OCD, anxiety, self-injurious behaviours and premonitory sensations also decreased after DBS. At least half of the patients required more frequent evaluations before achieving a good result. Motor tics responded only apparently better than phonic tics (ANOVA p = 0.339).

Tics did not disappear entirely. After DBS, only 3/18 patients (ie, 16%; Nos 4, 7, 12) required no medication. All patients received substantial support in person, on the telephone and by email. After DBS, medication requirements in most patients declined by half.

We chose to stimulate the thalamus (Vo–CM–Pfc complex) in our patients on the basis that the basal ganglia appear to play a major role in the “timing” and “sequencing” of motor and behavioural programmes, are implicated in the pathophysiology of GTS by neuroimaging and other experimental evidence and have been targeted successfully in previous stereotactic and DBS procedures.

In our patients, we did not observe any significant clinical variations during intraoperative stimulation. We demonstrated, by employing short “on–off” and “sham” conditions, that DBS works well: the tics and other symptoms were obvious and distressing when DBS was “off” but improved when DBS was switched “on”. In all but one patient, after a few seconds of turning “off” the DBS, their tics and behavioural symptoms reappeared dramatically. As with “sham off”, all patients developed anxiety symptoms; we felt it inappropriate to continue “sham off” because of the distress and anxiety (not increase in tics) that it caused. We therefore did no further “sham off”. One other DBS study reported “on–off” DBS testing in GTS26 and tics increased when the stimulator was “off”.

One patient (No 1) continues to consider that GTS has a severe impact on his life, despite a decrease in tics, well documented by both videotape and caregiver interview. A further patient had self-assessment visual analogue scale scores that also indicated that he (No 14) rated his QOL at the same level as that before DBS, despite the fact that his tics had reduced.

We suggest that DBS in GTS patients is safe and that there are few serious adverse effects either from surgery or the stimulation. Surgical complications with DBS in general can exceed 25%, and permanent neurological sequelae result in 4–6% of cases; in Parkinson’s disease, the incidence of stroke and death in 1–5%.3739 The low morbidity in the present patient sample may reflect the fact that our patients were young (age range 17–47 years) and were carefully selected. With regard to stimulation parameters and regulation, most of our patients experienced vertigo, some experienced abdominal discomfort, some transient blurring of vision (at >4 V) and one had an upward deviation of the eye (at >3 V). Transient side effects have been reported in most previous studies. Thus Visser-Vandewalle and colleagues22 reported decreased energy (3/3), “well being” (1/3) and altered sex drive (2/3) during stimulation as well as traction pain (2/3) from the pulse generator. Houeto and colleagues23 reported loss of weight in their patient, despite eating normally and having a normal endocrine evaluation. Flaherty and colleagues27 reported that their patient had significant mood swings (apathy, depression, hypomania) during stimulation, but was never suicidal; these mood swings disappeared when the stimulator was turned off. Ackermans and colleagues26 described reduced energy at the stimulus voltage necessary for the best effect on their tics. One patient experienced a short sudden dystonic jerk of the whole body each time the stimulators were switched “on” (during on–off trials).26 All nine documentations to date (n = 11 patients) suggest that DBS in GTS patients is a safe and well tolerated procedure. Our results are in agreement with these previous findings.

In contrast with previous reports of relatively easy DBS regulation, we found that we had to frequently regulate stimulation parameters for patients (7/13, 53%) who were followed for 6 months or more (Nos 1, 2, 5, 8, 9, 10, 13), and who required frequent modifications because of the reappearance of symptoms. This feature was more evident in the younger patients. This fluctuation in symptoms may be partially explained by more recent hypotheses concerning the long term neuromodulation induced by electrical stimulation.37 40

The procedure should thus be considered a symptomatic treatment, reserved only for a few patients with marked to severe GTS symptomatology.

At the present time, no consensus for DBS exists. Training of surgeons and the need for a multidisciplinary team are mandatory.39 Only carefully selected patients with severe GTS who are refractory to medical and psychobehavioural treatments and who have a markedly reduced QOL may benefit from DBS. DBS should be undertaken only in centres that are well acquainted with the treatment of GTS patients. Ad hoc protocols should be constructed based on evidence.

Limitations of our study

  1. We have not reported our video assessments as we did not follow standardised procedures, as suggested by Goetz and colleagues.41 We do have pre- and post-DBS videos, and all demonstrate the improvement.

  2. We also do not have longitudinal YGTSS scores for our patients (ie, for 6 months prior to the study); this would have been useful in further assessing treatment response.


Our results (n = 18) suggest that DBS of the Vo–CM–Pfc complex of the thalamus and other targets can be used successfully and safely in carefully selected patients with GTS who are refractory to medications and psychobehavioural treatments and who have severe symptomatology. We also demonstrated, using both “on–off” and “sham” conditions, that DBS is efficacious.


We would like to thank the Italian Tourette Syndrome Association for its continuing support. MMR was supported by the National Hospital Research Development Fund. We would also like to thank Drs Jeremy Stern and Neal Swerdlow for their invaluable comments on the drafts of the manuscript.


View Abstract


  • Competing interests: None.

Linked Articles