High-frequency stimulation of the globus pallidus for the treatment of Parkinson's disease

Surgical treatment for Parkinson's disease (PD) was introduced in the 1950s after the serendipitous discovery that lesions of the globus pallidus ameliorated parkinsonian features.¹ Enthusiasm for surgical treatment waned after the discovery of levodopa for the treatment of PD. Long-term treatment with levodopa is complicated by motor fluctuations, dyskinesias, loss of efficacy, and mental status changes.² The addition of other antiparkinsonian medications like dopamine agonists often results in improvement of motor fluctuations and dyskinesias. However, this improvement is only temporary, and as the disease progresses the symptoms worsen. The long-term complications of levodopa treatment, together with advances in stereotactic and neuroimaging techniques that allow for greater accuracy and less surgical morbidity, have made surgical interventions appealing.

Pallidotomy has been reported to be effective for the treatment of tremor, rigidity, bradykinesia, and drug-induced dyskinesia in PD.^3-6 Laitinen et al.³ reported significant improvement in all symptoms in 38 cases. Iacono et al.⁴ performed 50 unilateral and 68 bilateral pallidotomies with a mean follow-up of 4.5 months and reported the reversal of akinetic symptoms, elimination of dyskinesias, and reduction in "off" periods. Similarly, Dogali et al.⁵ reported that all symptoms of PD improved with pallidotomy. Unilateral pallidotomy is generally safe, but homonymous visual field defects, motor paresis, or intracranial hemorrhage can occur.^3,6 Bilateral pallidotomy can cause cognition and memory problems.⁷

An early observation with surgery was that electrical stimulation of anatomic targets resulted in the same clinical effects as lesioning. Siegfried and Lippitz⁸ reported improvement in Webster rating scale in three patients with pallidal stimulation. All the patients underwent bilateral stimulation, and improvement was reported in Hoehn and Yahr (H&Y) stage, speech, gait, and dyskinesias. They reported no complications of the procedure and did not report any adverse effects related to stimulation. We performed this study to further investigate the efficacy and safety of chronic high-frequency stimulation of the globus pallidus in patients with PD.

Methods. Subjects. Inclusion criteria included diagnosis of idiopathic PD (presence of two of three cardinal symptoms-tremor, bradykinesia, rigidity; and levodopa responsiveness) with levodopa-related motor fluctuations and dyskinesias as well as a stable antiparkinsonian drug regimen 1 month before surgery. Patients were excluded if they were poor surgical candidates due to unstable medical problems, concurrent use of a cardiac pacemaker, or had medical conditions that would require repeated MR imaging. Patients with hallucinations, dementia, major depression, or prior neurosurgical procedures were also excluded. Five patients (four men, one woman) were recruited from the PD clinic of the University of Kansas Medical Center (table 1). The mean age was 55.8 years (range, 48 to 65); mean disease duration was 14.6 years (range, 11 to 20). All patients signed informed consent before entry into the study.

Experimental protocol. At baseline, patients presented to the clinic in the morning without having taken AM antiparkinsonian medications. They underwent the following evaluations (off meds): Unified Parkinson's Disease Rating Scale (UPDRS),⁹ H&Y, Schwab and England Activities of Daily Living Scale (S&E), Tapping Total (TT), and Walking Time (WT). TT was the sum of the score obtained when the patient taps two counters 9 inches apart back and forth with each hand individually as many times as possible in 60 seconds. WT was the time required by the patient to walk 15 feet from a standing start, turn around, and return to the starting point. After baseline evaluations, the patients took their usual AM dose of antiparkinsonian medications. Sixty minutes later, the following evaluations were repeated (on meds): UPDRS (motor section, Part III), TT, and WT. Global disability ratings by the patient and the examiner were also performed. All patients were video-taped. All patients kept diaries of their global motor functions for 2 days before their visits. For each hour, they marked whether they were asleep, had good motor function ("on"), poor motor function ("off"), or good motor function accompanied by involuntary movements("on" with dyskinesia).

Three patients underwent bilateral implantation, the second of which was performed 2 to 3 months following the initial procedure. Follow-up evaluations were performed 3 months after the initial implantation in patients who underwent unilateral placement and 3 months after the second implantation in patients who underwent bilateral placements (3-month evaluations). For follow-up evaluations, patients presented to the clinic in the morning without taking their AM antiparkinsonian medications and with stimulators turned off. Evaluations were performed without PD medications and with stimulator off (off meds/off stim), without PD medications and with stimulator on (off meds/on stim), 60 minutes after taking medications and with stimulator off (on meds/off stim), and 90 minutes after taking medications and with stimulator on (on meds/on stim). Assessments included the UPDRS, TT, and WT, global disability ratings, patient diaries, and videotaping. The patients also rated themselves on a global scale compared with their previous visit (+3 markedly improved to -3 markedly worse).

Surgical procedure. A Cosman-Roberts-Wells (Radionics, Inc.) stereotactic head ring was placed under local anesthesia. CT of the head was performed with the gantry angled to approximate the anterior commissure-posterior commissure (AC-PC) plane. The target was defined as 3 mm anterior to the AC-PC midpoint and 20 mm lateral. The vertical coordinate was chosen between 3 and 6 mm inferior to the AC-PC plane. A burr hole was placed under local anesthesia 3 cm from the midline, anterior to the coronal suture. Microelectrode recording and stimulation techniques were then used to further refine the target. Once an optimum position was selected, macrostimulation with a 1.1 × 3 mm electrode was used to determine whether there were any side effects to the stimulation. If no side effects were demonstrated, the DBS electrode (Medtronics, Inc.) was then implanted in the same location. Stimulation was again used to confirm there were no side effects. The DBS electrode was then secured at the burr hole site, and the rest of the system hardware was implanted.

Initial Implantable Pulse Generator (IPG) programming was performed 12 to 48 hours postoperatively. For programming, the ITREL 7432 Console Programmer with S10344 MemoryMod software cartridge was used to select the stimulation settings that provided the best relief of parkinsonian symptoms and dyskinesias with least side effects. Adjustable settings included electrode selection, stimulation pulse rate, width, and amplitude. Patients were provided with an ITREL II 7452 control magnet and instructed on its use; they could turn the stimulator on and off by placing the hand-held control magnet over the IPG for several seconds. To conserve battery, the patients were instructed to turn the stimulator off at night while asleep. Readjustments of the IPG were made as needed if there was loss of benefit or intolerable side effects developed.

For efficacy analysis, the change in percentage of time "on," "off," and"on" with dyskinesias calculated from patient diaries at baseline and at 3-month post-implant were analyzed using t tests. Similarly, change in scores from baseline and 3-month follow-up evaluations were also analyzed with respect to UPDRS Part I (mentation), Part II (activities of daily living [ADL]), and Part III (motor examination) using t tests. The change in motor scores on meds/off stim at baseline and on meds/on stim at 3-month follow-up as well as off meds/off stim and off meds/on stim at 3-month follow-up were also analyzed using t tests. The level of significance was p < 0.05.

Results. Four patients were markedly improved (51 to 100%) and one patient was moderately improved (26 to 50%) at 3-month follow-up compared with baseline. There was a statistically significant improvement with stimulator on in the mentation (UPDRS Part I) and the ADL (UPDRS Part II) scores from baseline to 3-month follow-up (p < 0.05)(table 2). The mentation scores improved by 25%, and the ADL scores in the "off" state improved 19% and in the "on" state, 42%.

All motor assessments performed with the patient in the "off" state (off meds/off stim at baseline and off meds/off stim at 3-month follow-up) were also improved at 3-month evaluations. UPDRS (motor section) improved 24%(p = 0.07), WT improved 28% (p < 0.05), and TT improved 50% (p = 0.05). UPDRS (motor scores), WT, and TT on meds/off stim at baseline and on meds/on stim at 3-month follow-up also improved significantly (p < 0.05) (seetable 2). UPDRS (motor scores) improved by 60%, WT by 72%, and TT by 68%.

Motor assessments performed at follow-up with the patient in the "off" state (off meds/off stim) were compared with assessments performed when the stimulator was turned on (off meds/on stim). UPDRS (motor scores) improved 21% (p < 0.05), WT improved 27% (p = ns), and TT improved 13% (p = ns).

At baseline, four patients were rated as H&Y stage III, and one patient as H&Y stage IV. These ratings improved at 3-month follow-up, when one patient was rated as H&Y stage III, two patients as H&Y stage 2.5, and two patients as H&Y stage 2. Mean S&E scores improved from 60 to 76% (100% being normal).

As measured by patient diaries, the amount of time "on" increased significantly at 3 months compared with baseline (p < 0.05). Similarly, there was a significant decrease in the "off" time and "on" time with dyskinesias (figure). Mean "off" time decreased from 39 to 23%, and "on" time improved from 21 to 65%. "On" with dyskinesias decreased from 40 to 12%. All patients reported that the "off" periods were not as severe as before surgery, and the dyskinesias were not disabling. Patients required an average of 4.8 adjustments over time. The majority of the adjustments were done during the first month after surgery.

• Download figure
• Open in new tab
• Download powerpoint

Figure. Data from patient diaries at baseline and follow-up.

Surgical complications were minor and resolved completely. One patient had an asymptomatic intracranial hemorrhage that was discovered during a routine postoperative CT. One patient had transient speech difficulty and hemiparesis in the operating room during stimulation, which resolved in 5 minutes. One patient had facial dystonia and paresthesia that required surgical repositioning of the lead about 6 weeks after initial implantation.

Adverse effects related to stimulation were minimal. One patient reported transient visual disturbances, and another patient had choreiform movements in the foot related to stimulation.

At baseline, global assessments by the examiner reported two patients with severe disability (75 to 100%) and three patients with marked disability (50 to 74%). At 3 months, the examiner rated four patients with mild disability(1 to 24%) and one patient with moderate disability (25 to 49%). Similarly, at baseline three patients rated themselves as severely disabled, and two patients rated themselves as markedly disabled. At follow-up, three patients reported mild disability, and two patients reported moderate disability.

There was a slight increase in antiparkinsonian medication at 3 months compared with baseline. Patients were receiving 47.4 units of antiparkinsonian medication at baseline (1 unit = 100 mg of standard levodopa= 125 mg of sustained-release levodopa = 100 mL of liquid levodopa = 1 mg of pergolide = 10 mg of bromocriptine) compared with 49.9 units at follow-up(p = not statistically significant).

Discussion. Chronic pallidal stimulation is a safe and effective treatment for PD complicated by motor fluctuations and dyskinesias. The majority of objective and subjective rating scales improved significantly after pallidal stimulation. At the follow-up evaluations, two patients had only unilateral stimulators implanted. One patient felt there was a marked improvement from baseline and did not want to have the second side implanted, and the second patient has recently undergone surgery on the second side. According to the S&E, four patients required assistance in daily living before the procedure compared with only one patient after surgery. This patient subsequently had a second implant.

We report similar outcomes to those of Siegfried and Lippitz⁸ in our patients with pallidal stimulation. Patients who underwent bilateral implants did better on all UPDRS subscales compared with patients who underwent unilateral implants. An interesting observation was that all objective scores (WT, TT, UPDRS-III) improved even when the medications were "off" and the stimulator was off. Although this could be due to the slight increase in antiparkinsonian medications, the patients presented to the clinic without taking medication. Another possibility is a placebo effect related to the overall feeling of improvement following a major surgical procedure; however, these evaluations were performed 3 months after the last surgery. We do not believe that this improvement is due to a micropallidotomy due to the surgical procedure, as these improvements were not present immediately after surgery. Chronic stimulation might induce changes that are evident even when the stimulator is turned off.

One of the concerns regarding bilateral pallidotomy is the risk of cognitive dysfunction and speech abnormalities.⁷ None of our patients who underwent bilateral pallidal stimulation had worsening of mental status. Although there were subtle changes on detailed neuropsychological testing (Tröster, personal communication), these changes also occurred with unilateral implants.

One of the disadvantages of pallidal stimulation is the cost of the system and the presence of a foreign body. Another additional risk is the general anesthesia required for the implantation of the stimulator. However, there are many advantages: low surgical complications, safety of bilateral stimulator implants, the stimulator settings can be adjusted to decrease the adverse effects or to increase the efficacy, and reversibility of the procedure.

Limousin et al.¹⁰ reported that subthalamic nucleus(STN) stimulation improved parkinsonian symptoms. Because only a small number of patients have undergone STN and pallidal stimulation, it is not possible to determine which is the better site. Pallidal stimulation has the possible advantage of lessening drug-induced dyskinesias compared with STN stimulation. However, STN stimulation may result in a need for less antiparkinsonian medication, thereby lessening dyskinesias. The effect of these two procedures on various parkinsonian symptoms remains to be further evaluated.