Deep brain stimulation for dystonia: a meta-analysis.

Objective.  To use a meta-analysis on all reported cases of deep brain stimulation (DBS) for dystonia to determine which factors significantly influence outcome. The Burke-Fahn-Marsden (BFM) movement scale, the most reported measure, was chosen as the primary outcome measure for this analysis. Methods.  A MEDLINE search identified 137 patients who underwent DBS for dystonia in 24 studies that had individual BFM scores. Individual patient data, including age at onset of dystonia, age at surgery, gender, distribution of dystonia, etiology of dystonia, presence of associated features, abnormality of preoperative imaging, prior stereotactic surgeries, nucleus stimulated, type of anesthesia used, use of physiologic monitoring, type of imaging used for localization, stimulation parameters used, time of response to stimulation, and timing of outcome assessment were entered into an SPSS database for statistical analysis. Results.  The mean BFM percentage change (improvement in postoperative score from baseline) was 51.8% (range -34% to 100%). Significantly better outcomes were achieved with stimulation of the globus pallidus internus (GPi) than with stimulation of the posterior portion of the ventral lateral (VLp) nucleus of the thalamus (p = 0.0001). The etiology of the dystonia also had a significant effect on outcomes. Statistically significant improvements in outcomes were seen for all etiologic categories, except encephalitis. Dystonia due to birth injury and encephalitis had significantly worse outcomes when compared to other etiologies. However, there were no significant differences in the outcomes of patients who were DYT1 (DYT1 is the gene associated with the disorder Dystonia Musculorum Deformans) gene positive, DYT1 gene negative, or had pantothenate kinase-associated neurodegeneration (PKAN), tardive dyskinesia, and idiopathic and posttraumatic dystonias. Longer duration of dystonia symptoms correlated negatively with surgical outcome. A regression model using the three variables-stimulation site, etiology of dystonia, and duration of dystonia symptoms-explained 51% of the variance in outcomes. Conclusion.  Deep brain stimulation of the GPi provides significant improvement in BFM scores in a variety of dystonic conditions.