Background: The subthalamic nucleus is the preferred target for deep brain stimulation in patients with advanced Parkinson's disease. The site of permanent stimulation is the subject of ongoing debate, as stimulation both within and adjacent to the subthalamic nucleus may be effective.
Objective: To assess the position of active electrode contacts in relation to the dorsal margin of the subthalamic nucleus as determined by intraoperative microrecordings and magnetic resonance imaging (MRI).
Methods: In 25 patients suffering from severe levodopa sensitive parkinsonism, deep brain stimulating electrodes (n = 49) were implanted following mapping of the subthalamic nucleus by microrecording and microstimulation along five parallel tracks. Postoperative stereotactic radiography and fusion of pre- and postoperative MRI studies were used to determine the stereotactic position relative to the midcommissural point of the most effective electrode contacts selected for permanent stimulation (n = 49). Intraoperative microrecordings were analysed retrospectively to define the dorsal margin of the subthalamic nucleus. In cases where the dorsal margin could be defined in at least three microrecording tracks (n = 37) it was correlated with the position of the active contact using an algorithm developed for direct three dimensional comparisons.
Results: Stimulation of the subthalamic nucleus resulted in marked improvement in levodopa sensitive parkinsonian symptoms and levodopa induced dyskinesias, with significant improvement in UPDRS III scores. In several instances, projection of the electrode artefacts onto the T2 weighted MRI visualised subthalamic nucleus of individual patients suggested that the electrodes had passed through the subthalamic nucleus. When the actual position of active electrode contacts (n = 35) was correlated with the dorsal margin of the subthalamic nucleus as defined neurophysiologically, most contacts were located either in proximity (+/- 1.0 mm) to the dorsal border of the subthalamic nucleus (32.4%) or further dorsal within the subthalamic region (37.8%). The other active contacts (29.7%) were detected within the dorsal (sensorimotor) subthalamic nucleus. The average position of all active contacts (n = 49) was 12.8 mm (+/- 1.0) lateral, 1.9 mm (+/- 1.4) posterior, and 1.6 mm (+/- 2.1) ventral to the midcommissural point.
Conclusions: Subthalamic nucleus stimulation appears to be most effective in the border area between the upper subthalamic nucleus (sensorimotor part) and the subthalamic area containing the zona incerta, fields of Forel, and subthalamic nucleus projections.