Efficacy and Safety of Deep Brain Stimulation (DBS) of the Pallidal (GPi) in Patients With Tardive Dystonia

• Improvement of the motor scale of Burke-Fahn-Marsden-Dystonia Rating Scale via blinded video assessment 3 months after starting DBS in comparison to sham-stimulated patients [ Time Frame: 3 months ] [ Designated as safety issue: No ]
  

• AIMS [ Time Frame: 3 months ] [ Designated as safety issue: No ]
  
• Non-motor subscores of BMFDRS [ Time Frame: 3 months ] [ Designated as safety issue: No ]
  
• Visual analogue scales for both patients and treating physicians [ Time Frame: 3 months ] [ Designated as safety issue: No ]
  
• Quality of life (SF-36) [ Time Frame: 3 months ] [ Designated as safety issue: No ]
  
• Psychiatric assessment (HADS-D and PANSS) [ Time Frame: 3 months ] [ Designated as safety issue: No ]
  

Deep brain stimulation (DBS) has been established as a new reversible, neurosurgical therapeutic option for patients suffering from disabling neurological movement disorders such as essential tremor and Parkinson´s disease. Recently, deep brain stimulation has been successfully applied in patients with primary generalized and segmental dystonia. Additionally, a number of case reports suggest that pallidal deep brain stimulation may also improve tardive dystonia, which may for instance result from the intake of neuroleptics and which is notoriously difficult to treat medically. The present study will investigate the effects of pallidal DBS using a double blind, randomized design (sham- versus verum-stimulation within a 3-months interval post implantation of the electrodes).

Initially 60 patients had been calculated in a power analysis to assess significant results based on an average improvement of dystonic symptoms of 30%. However, in a recent study (Damier et al., Archives of General Psychiatry, 2007), 10 out of 10 showed a successful outcome of approximately 50% decrease on the extrapyramidal symptoms rating scale score. The exact one- sided lower 95% confidence limit would be 0.794 for this result. If such an approach is chosen for sample size estimation with 18 verum and 18 placebo patients one would obtain a power of 82% against a placebo effect of 30% success rate. For a placebo effect of 25% one needs 16+16 patients and for the placebo effect of 20% one needs 12+12 patients. We thus decided to reduce the sample size to 36- 32- 24 patients. It is expected that the continuous primary outcome measure will preserve even higher power than the binary one used in the study mentioned above. The local ethical committee has approved this.