Bilateral globus pallidus stimulation for severe Tourette's syndrome: a double-blind, randomised crossover trial

Summary Background Deep brain stimulation (DBS) has been proposed as a treatment option for severe Tourette's syndrome on the basis of findings from open-label series and small double-blind trials. We aimed to further assess the safety and efficacy of bilateral globus pallidus internus (GPi) DBS in patient's with severe Tourette's syndrome. Methods In a randomised, double-blind, crossover trial, we recruited eligible patients (severe medically refractory Tourette's syndrome, age ≥20 years) from two clinics for tertiary movement disorders in the UK. Enrolled patients received surgery for GPi DBS and then were randomly assigned in a 1:1 ratio (computer-generated pairwise randomisation according to order of enrolment) to receive either stimulation on-first or stimulation off-first for 3 months, followed by a switch to the opposite condition for a further 3 month period. Patients and rating clinicians were masked to treatment allocation; an unmasked clinician was responsible for programming the stimulation. The primary endpoint was difference in Yale Global Tic Severity Scale (YGTSS) total score between the two blinded conditions, assessed with repeated measures ANOVA, in all patients who completed assessments during both blinded periods. After the end of the blinded crossover phase, all patients were offered continued DBS and continued to have open-label stimulation adjustments and objective assessments of tic severity until database lock 1 month after the final patient's final trial-related visit. This trial is registered with ClinicalTrials.gov, number NCT01647269. Findings Between Nov 5, 2009, and Oct 16, 2013, we enrolled 15 patients (11 men, four women; mean age 34·7 years [SD 10·0]). 14 patients were randomly assigned and 13 completed assessments in both blinded periods (seven in the on-first group, six in the off-first group). Mean YGTSS total score in these 13 patients was 87·9 (SD 9·2) at baseline, 80·7 (SD 12·0) for the off-stimulation period, and 68·3 (SD 18·6) for the on-stimulation period. Pairwise comparisons in YGTSS total scores after Bonferroni correction were significantly lower at the end of the on-stimulation period compared with the off-stimulation period, with a mean improvement of 12·4 points (95% CI 0·1–24·7, p=0·048), equivalent to a difference of 15·3% (95% CI 5·3–25·3). All 15 patients received stimulation in the open-label phase. Overall, three serious adverse events occurred (two infections in DBS hardware at 2 and 7 weeks postoperatively, and one episode of deep-brain-stimulation-induced hypomania during the blinded on-stimulation period); all three resolved with treatment. Interpretation GPi stimulation led to a significant improvement in tic severity, with an overall acceptable safety profile. Future research should concentrate on identifying the most effective target for DBS to control both tics and associated comorbidities, and further clarify factors that predict individual patient response. Funding UK National Health Service.