CHAPTER 81 Subthalamotomy in Parkinson’s Disease
Indications and Outcome
Subthalamic Nucleus as Target
In a classic paper, Whittier and Mettler1 described the consequences of an electrolytic lesion in the STN in monkeys. They induced hemiballism (which they termed hyperkinesia choreica) contralateral to the STN lesion and showed that two elements must be present for hyperkinetic movement to occur: minimal (>20%) destruction of the structure, and integrity of the globus pallidus (GP) and pallidothalamic pathway. One year later, Carpenter and coworkers2 showed that lesions of the GP or pallidal efferent eliminated the dyskinesias induced by STN lesions. These studies demonstrated the importance of this small structure in the origin of hyperkinesias. This correlation was already known in clinical practice, because patients with vascular insults in the STN developed a hemichorea-ballism (HCB) contralateral to the side of the injury.3 Accordingly, for a long time the STN was thought to exert an inhibitory effect on target nuclei because their lesioning induced a movement disorder. Subsequently in 1988, Smith and Parent4 first reported that the STN efferent projection shows glutamate inmunoreactivity, and Albin and associates5 confirmed the glutamate subthalamic pathway 1 year later. This finding was relevant because the STN is the only excitatory structure of the basal ganglia (BG) and consequently is capable of exciting BG output nuclei, GPi, and substantia nigra pars reticularis to thalamus.
Mitchell and coworkers6 from Manchester University described metabolic studies in the STN of monkeys using 2-deoxyglucose (2-DG) autoradiography with carbon 14 to map the metabolic changes associated with dyskinesia. Uptake of 2-DG is believed to represent an index of regional afferent synaptic activity. Their study demonstrated that the external globus pallidus (GPe)-STN pathway is overactive on the side contralateral to dyskinesia, causing inhibition of the STN and leading to reduced pallidal output to the thalamus.6 The same group found that monkeys previously treated with 1-methyl, 4-phenyl, 1,2,3,6-tetrahydropyridine (MPTP) showed a major decrease in 2-DG uptake, indicating reduced inhibition from the GPe to the STN in parkinsonian primates’ brains.7 The authors concluded that the STN plays a key role in both experimental parkinsonism and dyskinesias, operating in opposite functional directions.6,7 At the same time, Miller and DeLong8 used neuronal recording to demonstrate that the STN exhibits increased neuronal activity in MPTP-treated monkeys. Subthalamic hyperactivity was associated with increased GPi neuronal activity, which led to excessive inhibition of the motor thalamus and thalamocortical projections. Overall, these studies provided the basis for the pathophysiologic model of the BG, where STN activity plays a pivotal and opposite role in the origin of the major BG motor disorders.
The next strategy was to create an STN lesion in MPTP-treated monkeys to alleviate the parkinsonian condition. This was first reported by Bergman and colleagues,9 who found considerable alleviation of akinesia, rigidity, and tremor on the contralateral side of a chemically (ibotenic acid) induced lesion in two parkinsonian monkeys. One year later, Aziz and associates10 reported similar findings in six parkinsonian monkeys with STN thermolesions. Finally, Guridi and coworkers11 showed that subthalamic chemical lesions in parkinsonian monkeys improved their motor features and reduced the increased BG output activity, as shown by in situ hybridization for glutamic acid decarboxylase messenger RNA, in the GPi and substantia nigra pars reticularis. All these studies determined that ablation of the STN to reduce or eliminate hyperactivity in the parkinsonian state led to behavioral motor improvement (Fig. 81-1). Thanks to experimental research, a new target had been discovered, and it was paradoxical, given that the STN had been the most feared anatomic structure during early surgery for PD, before the introduction of levodopa.
Benazzouz and associates12 implanted electrodes in the STN for high-frequency stimulation in two parkinsonian monkeys and found a marked clinical improvement. Soon after, the Benabid group13 reported the first parkinsonian patient treated with STN-DBS, and in 1995 they described three cases showing a great improvement in motor symptoms.14 This report was the beginning of the STN-DBS era, which is where functional neurosurgery for PD still stands today. Subthalamic lesions and STN-DBS generally produce the same clinical response in terms of motor disability improvement. However, stimulation procedures do not result in lesions in the brain, and the reversibility and adjustability of such procedures are potential advantages. In contrast, ablative surgery does not require a second operation for battery implantation, and device-related complications, such as infection or adverse stimulation effects, are not a factor. In some patients, however, the ablative lesion may cause contralateral dyskinesia, which means that subthalamic nucleotomy is performed in relatively few patients compared with DBS procedures, owing to the fear of inducing HCB.
All these research findings led to a revitalization of surgery for PD and paved the way for a reevaluation of the STN as a potential surgical target.15 However, pallidotomy was the preferred approach when surgery for PD was revitalized in the early 1990s. This was based mainly on the surgical dogma that STN lesions are always associated with severe dyskinesias. Thus, early surgery during the 1960s relied on subthalamic area lesions, termed campotomies or subthalamotomy, as an alternative to thalamotomy for alleviating tremor and rigidity in parkinsonian surgical candidates.16–18 The surgical targets were Forel’s field, the zona incerta, and the prerubral field, but never the STN itself, for fear of inducing HCB. Indeed, early stereotactic surgery for PD (before the introduction of levodopa as medical treatment) established that surgeons should not perform any lesion below the intercommissural line because of the high risk of inducing dyskinetic movement. Postoperative HCB was a severe complication in some cases, and surgeons were reluctant to touch the STN.19,20 In patients with parkinsonism after encephalitis or after bilateral thalamotomies, the risk of developing dyskinesia was increased.20
Laitinen and coworkers’21,22 experience with pallidal lesions using Leksell’s target reestablished the GPi as a surgical target for advanced PD. All the cardinal features, such as tremor, rigidity, and bradykinesia, as well as levodopa-induced dyskinesias (LIDs), were improved on the side contralateral to the lesion. Pallidotomy was performed in the posteroventral and lateral portions of the GP, close to the putamen.22 However, it soon became apparent that the improvement conveyed by unilateral pallidotomy fell short of the needs of patients with advanced PD. This coincided with the early development of DBS, summarized earlier.
Surgical Procedure
Imaging and Target Selection
Magnetic resonance imaging (MRI) is performed 24 to 48 hours before surgery. T1- and T2-weighted sequences are used, with 2-mm-thick slices and no interspaces. On the day of surgery, the stereotactic frame is positioned with the use of local anesthesia, and a computed tomography (CT) scan is obtained (slices 2 mm thick with no interspaces); this information is sent to the neuronavigation workstation.23 The CT scan is fused with the T2-weighted image, and the target is selected by indirect and direct procedures. With indirect targeting, the anterior commissure–posterior commissure (AC-PC) line is measured, and the mid-intercommissural point (ICP) is located. The target located is 12 mm lateral (X coordinate) and 2 to 3 mm posterior to the ICP (Y coordinate) and 4 mm below the AC-PC line (Z coordinate). The STN can be identified as a hypointense almond-shaped nucleus located lateral to the red nucleus (RN) in the mesencephalon on axial T2-weighted images. This image target may be moved to a line crossing the anterior border of the RN approximately 3 mm from its lateral wall. The RN can be used as a landmark for the anteroposterior coordinate of the STN on axial T2-weighted images (direct targeting).24 After target selection, the anteroposterior and coronal arc is chosen, avoiding the ventricle wall.
Access and Electrophysiology
After the STN target has been chosen, the patient is placed in a semisitting position and administered local anesthesia. A bur hole (14 mm) is created 15 to 20 mm from the midline. Dural opening and cortical coagulation are performed while the surgical field is continuously sealed to prevent air embolism. The first electrophysiologic recording track is advanced until it reaches the nucleus. The beginning of the STN is defined by an abrupt and significant increase in electrical activity in the recording. During this part of the procedure, we look for the sensorimotor portion of the STN. Kinesthetic neurons with a response to contralateral limb movements and tremor cells (in patients with tremor) are located in the dorsolateral portion of the anatomic structure. Somatotopically, lower limb units are more medial than upper limb units, which are in the most lateral part of the STN, near the internal capsule.25 After recording with microelectrodes or semi-microelectrodes, a mapping of the nucleus is obtained.23 Nuclear length is an important parameter for lesion placement. Currently, there is no consensus on the best target, and different groups use different criteria to determine lesion location. The lesion must be placed in the sensorimotor portion of the STN (dorsolateral region), which is where the kinesthetic units are recorded; this is the same place used for electrode introduction in stimulation surgery.
