Unilateral pallidotomy for Parkinson's disease: Comparison of outcome in younger versus elderly patients

Medial pallidotomy represents a surgical option for improving motor functioning in patients with Parkinson's disease (PD) in whom medical therapy is problematic. The few published series^1-6 report pallidotomy as beneficial for PD patients (the majority being <60 years at time of surgery in these series; see table 1). The surgery is thought ideal for relatively young patients with intractable drug-related side effects including dyskinesias and "on/off" fluctuations.⁷ However, the majority of PD patients have onset of symptoms >60 years and typically have a good response to pharmacologic therapy for at least 5 years. Consequently, the majority of PD patients experiencing difficulties in treatment of their symptoms are ≥65 years. We evaluated the effects of pallidotomy in the first 20 patients undergoing this surgery at our institution, allowing the opportunity to reflect on the outcome of this surgery in elderly (≥65 years) and younger patients with PD.

Patients and methods. One neurologist (R.J.U.) evaluated all patients, determining a diagnosis of PD by virtue of patients having at least two of the following features: resting tremor, bradykinesia, and rigidity. Each patient enjoyed benefit with levodopa therapy but had advancing symptoms including development of disabling motor fluctuations. R.J.U. evaluated all patients on multiple occasions, and found significant disability despite optimal medical therapy (including use of at least one dopamine agonist in conjunction with levodopa therapy). Patient assessments occurred within 1 month prior to unilateral pallidotomy using a modified "Core Assessment Program for Intracerebral Transplantation" (CAPIT) protocol.⁸ Patient evaluations took place in the morning, at least 2 hours after patients awakened, having not taken any antiparkinsonian medications or food since midnight the previous evening. We obtained the following data: Unified Parkinson's Disease Rating Scale⁹ (UPDRS) motor score in an "off" state, Goetz¹⁰ (severity range 0-4) and Mayo dyskinesia rating scale (severity scores 0-4 for dyskinesia of neck, trunk, and each limb), Purdue pegboard testing¹¹ (total pegs placed for 30" intervals using each hand alone and simultaneously) and counter tapping(total counters tapped in 1" using a single hand alternating between 2 counters mounted 20 cm apart on a Plexiglas board) measures. We recorded the same ratings initially at 30" intervals (for the first 2 hours) and then at 60" intervals following the administration of the patient's typical first dose of antiparkinsonian therapy. Scoring continued for the ensuring 5 hours or until patients returned to their baseline functioning (within 10%) for two consecutive ratings, providing a measure of the best "on" state (minimal UPDRS motor scores, optimal pegboard scores, and tapping scores). Videotapes recorded sequential UPDRS testing. We scored UPDRS activity of daily living(ADL), mood, mentation, and behavior (MMB), and complications of therapy(CoT) assessments prior to surgery. We re-evaluated patients approximately 3 months following surgery, collecting data identical to those listed above.

We performed neuropsychological evaluation within 2 weeks prior to and approximately 3 months following pallidotomy in the last nine patients in this series. One patient received a limited assessment because of deafness. This testing occurred on a day other than the aforementioned drug profile motor function testing day, in an "on" state as much as possible, and included the following tests: Dementia Rating Scale (DRS), subtests of the Wechsler Adult Intelligence Scale-Revised (WAIS-R), Boston Naming Test (BNT), Controlled Oral Word Association, Verbal Fluency for Semantic Categories, Benton Judgment of Line Orientation (JLO), Logical Memory and Visual Reproduction subtests of the Wechsler Memory Scale-Revised (WMS-R), California Verbal Learning Test (CVLT), Trail Making Test, Wisconsin Card Sorting Test, and Geriatric Depression Scale.^12,13 These patients also completed 4-day diary accounts, recording their motor functioning (as "on," "off," or "on with dyskinesia") for consecutive days (including 2 weekdays and a weekend) within 2 weeks prior to and 3 months following pallidotomy.

Pallidotomy. R.E.W. performed each pallidotomy. Determination of an anatomic target employed an MRI with a stereotactic frame (targets selected were within the following ranges: 2-3 mm anterior, 3-5 mm inferior, and 21-23 mm lateral to the midpoint of the anterior commissure-posterior commissure [AC-PC] line). We determined the coordinates and trajectory of the target in part on the basis of computerized stereotactic software calculations employing each patient's own MRI information. We employed a constant trajectory for reaching targets, utilizing a 30° angle to a perpendicular from the midpoint of the AC-PC line in the parasagittal plane. Through a standard burr hole, a microelectrode (or semimicroelectrode in some instances) was inserted to the target depth and electrophysiologic recordings made of the tonic high frequency electrical activity in the internal segment of globus pallidus (GPi). We recorded somatotopic representations within GPi on the basis of electrical activity elicited with passive extremity movement. We delineated the superior and inferior boundaries of the GPi and the location of the optic tract on the basis of electrophysiologic recordings, correlating these findings with the patient's own MRI. Following determination of stimulation thresholds for the internal capsule and optic tract, radiofrequency lesions (70° for 60 seconds) were placed in the somatotopic region of GPi. We made the initial lesion in the somatotopic region of the GPi corresponding to the contralateral arm. We subsequently placed lesions based on somatotopic information and clinical response following lesioning (more medial for leg and more lateral for arm). Ablation of tremor often required a more medial lesion. The number of lesions ranged from one to six (average 2.5) and the number of trajectories ranged from one to four (average 1.5). The average operating time was 270 minutes.

We performed unilateral pallidotomy in 20 patients (14 men and six women). The mean age at onset of symptoms was 51.0 ± 10.4 years. The mean age at surgery was 65.5 ± 9.0 years. Eleven patients had pallidotomy involving the dominant hemisphere (19 patients were right-handed; eight underwent right medial pallidotomies; the left-handed patient underwant a left medial pallidotomy). Eleven patients (eight men, three women) were≥65 at the time of surgery (mean = 71.4 ± 4.4; median = 72; range: 65-78). Nine patients (six men, three women) were <65 years at the time of surgery (mean = 56.3 ± 5.1; range: 49-62).

Statistical methods. A two-sided paired t test compared preoperative and 3-month values. The signed rank test (not included in tables) confirmed these results. The Wilcoxon rank sum test compared preoperative, 3-month, and change (3-month preoperative) values between those patients age <65 and ≥65. Mean ± standard deviation and the median values summarized pre- and postoperative measurements.

Results. UPDRS scoring. Table 2 summarizes pre- and postoperative UPDRS scoring. The mean improvement in total UPDRS scores was a reduction by 22%. The mean improvement in UPDRS "on" motor score was 24% and "off" motor score 20%. Subgroup comparisons between<65 and ≥65 age pre- and postoperative scores did not vary significantly(see table 2). Eleven patients had severe disturbances of gait preoperatively in an "off" state (UPDRS motor examination #29 score 4: unable to walk at all, even with assistance, n = 6; score 3: severe disturbances of gait, requiring assistance, n = 5. Postoperatively five patients scored 3 or 4 in an "off" state (score 4, n = 2; score 3, n = 3). Six of 11 patients had improvement to scores of 2 or less, requiring little or no assistance to walk.

UPDRS ADL, MMB, and CoT scores. The mean ADL score preoperatively was 20.9 ± 8.6 (n = 19). The postoperative score was significantly improved: 16.7 ± 7.1 (p = 0.01). The mean MMB scores did not vary significantly (p = 0.37; n = 19): preoperative: 2.8 ± 1.7, postoperative: 2.4 +/- 2.2. Mean CoT score improved significantly following surgery (p 0.001; n = 19): preoperative: 8.3 ± 3.3, postoperative: 5.6 ± 2.9.

Time motoer tasks. Objective measures of hand speed with Purdue pegboard testing and tapping counter testing showed some significant improvements in speed contralateral to surgery (table 3).

"On" time. The mean daily preoperative "on" time, as recorded by patient diary (recorded for last nine consecutive patients ony), was 4.1± 2.9 hours. Postoperative "on" time increased to 8.8 ± 5.2 hours (p = 0.015). The mean daily preoperative "on" time in the≥65 yr subgroup increased from 4.9 ± 2.4 to 9.7 ± 4.0 hr(p = 0.03). The mean daily preoperative "on" time in the <65 yr subgroup increased from 3.5 ± 2.9 to 5.9 hrs (p = 0.02). There were no significant differences between pre- and postoperative "on" time between the ≥65 and <65 age subgroups.

Goetz and Mayo dyskinesia scores. The mean maximal Goetz severity scale score was 1.4 ± 1.2 preoperatively and 1.2 ± 1.2 postoperatively. The mean maximal total Mayo dyskinesia score was 11.6± 5.6 before surgery and 7.6 ± 3.2 after surgery (p< 0.01).

Neuropsychological data. We collected neuropsychological data in the last nine consecutive patients of the series; all of these patients had dominant hemisphere pallidotomy. Seven of the nine patients were≤65 years of age while only two were <65 years. Consequently, there were not sufficient numbers to compare age subgroups directly. We analyzed data using paired t tests with alpha set at a more stringent level of 0.01 in order to control for experimentwise error rates. We found no changes postoperatively in general cognitive functioning, visuospatial abilities, attention span, memory, concept formation/cognitive flexibility, or depression. There was a reduction in semantic fluency and controlled oral word association by approximately one-quarter of the mean values. There was a trend toward improved visuomotor integration/sequencing. The results of neuropsychological studies are presented in table 4.

Levodopa intake. There was no significant change in daily levodopa intake following surgery. Preoperative combined standard +[controlled release × 0.7 (to account for difference in bioavailability)] = 1,023 mg/d (1,112 mg/d for <65 yr, 965 mg/d for >65 yr); 3 months postoperative = 1,127 mg/d (1,036 mg/d for <65 yr, 1,187 for ≥65 yr). (Levodopa intake used for the UPDRS protocol [first dose] was similarly unchanged: mean preoperative = 174 mg, 3 months postoperative = 168 mg.)

Complications of surgery. Three patients experienced transient confusion (disorientation) lasting 1 to 3 days postoperatively (all three were ≥65 years). One patient had urinary incontinence for several weeks before this complication resolved spontaneously. No patients had clinically detectable motor weakness or visual field deficits following surgery.

Discussion. In the present series, PD patients were selected for pallidotomy by the neurologist using inclusionary criteria consisting of failure of optimal medical therapy and exclusion criteria consisting of significant dementia or lack of levodopa responsiveness. There was no systematic attempt to select only younger patients, those with highly asymmetric signs, or those without severe disability. Many patients had severe gait difficulties (11 of 20 could not walk without aid in an "off" state). The improvements in motor function found in our series were similar to those reported regarding pallidotomy in younger PD patients with similar data collection protocols. Dogali et al.¹ reported an improvement in UPDRS "off" scores by a mean of 57% (from approximately 35 to 15 postoperatively as from figure 3). Lozano et al.³ reported an improvement of 30% in the UPDRS "off" motor score 6 months postoperation. Baron et al.⁶ reported finding an improvement in the UPDRS "off" motor score of 24.9% at 3 months and improvement in the total UPDRS score of 30.1% at 3 months. We found an improvement of 20% in the UPDRS "off" motor score and an improvement in the total UPDRS score of 22% at 3 months postoperatively. Others have found minimal changes in best "on" motor scores.^1,3, Our patients did have mild improvement in "on" motor scores (improvement of 24%). Baron et al.⁶ also reported improvement in "on" motor scores at 6 months follow-up, which waned slightly at 1 year. In this series, six of 11 patients who were unable to walk without assistance in an"off" state preoperatively were able to do so postoperatively. Finally, we also found improvements in objective, timed tasks evaluating individual hand speed and dexterity, as well as tests of simultaneous hand speed, in contrast to others.¹⁴

Dyskinesia scores using a CAPIT protocol with a single antiparkinsonian dose profile undoubtedly underestimated the severity of dyskinesia in many of our patients. Nevertheless, we measured significant reductions in dyskinesia scores using the Mayo dyskinesia score. Perhaps the most significant improvement following pallidotomy, in terms of patient functioning, is reflected by the impressive gains (by >100%) in hours of daily functional"on" time reported postoperatively. This improvement stems mainly from reductions in time spent with intolerable levels of dyskinesia. These improvements were confirmed by reductions in the UPDRS activities of daily living and complications of therapy scores. There were no changes in the geriatric depression scale, suggesting that the aforementioned improvements likely cannot be explained solely on the basis of improved mood. Other studies have indicated that follow-up evaluations at 3 months are likely to remain unchanged at 12 month intervals.⁶ Longer follow-up is required to determine whether this is the case in the present series.

More studies are needed to confirm the value of pallidotomy for Parkinson's disease.⁷ The few series published to date generally consider small numbers of somewhat younger PD patients with disabling complications secondary to medical therapy (see table 1). However, the largest contingent of disabled PD patients are of elderly age. Elderly PD patients are also subject to disabling drug-related side effects. While these side effects may relate to motor complications, they are also more likely to include cognitive dysfunction in elderly persons. Neuropsychological studies in this series indicated no significant changes in general cognitive/intellectual abilities, confrontation naming, visuospatial ability, attention, memory, executive functions, or depression following pallidotomy of the dominant hemisphere. These findings are consistent with those reported in other recent studies.^6,15 Significant postsurgical declines were evident, however, on two measures of generative naming: controlled oral word association and verbal fluency for semantic categories. Rilling et al.¹⁵ also found a decline in semantic fluency; however, they observed no reduction in controlled oral word association. Baron et al.⁶ did not measure semantic fluency, but like Rilling et al.,¹⁵ reported no significant change in controlled oral word association. The discrepancy of results between the current report and those of previous studies warrants further investigation and may reflect differences in sample sizes, disease-related parameters(e.g., age of onset), and subject characteristics (e.g., age, gender, lateral dominance, side of surgery).

The present series permits evaluation of patients with an average age of over 65 years. We divided the series into subgroups less than or greater than 65 years. These groups were similar preoperatively with respect to motor scoring, levodopa-responsiveness, and neuropsychological baseline measurements. Medial pallidotomy appeared equally beneficial between groups, on the basis of improvement in motor function scores, timed tests, activity of daily living scores, duration of daily functional "on" time, and reduction in complications of therapy. Importantly, measures of neuropsychological function were also similarly preserved regardless of age sub-group. Although Dogali et al.¹ indicated their impression that younger PD patients showed greater improvement, they acknowledged their study did not show any statistical support for this impression. Others have suggested that pallidotomy be restricted to those younger than 70 years of age without hypertension.¹⁴ Our series also confirms the lack of any pronounced difference in benefit/side effects merely on account of advanced age. In contrast, another report, based on six patients, indicated that pallidotomy did not appear particularly useful in ameliorating overall function.⁵ This previous study did include the most elderly group of patients reported to date. In contrast to the present series, the other study used a CT-guided technique without additional electrophysiologic mapping. Others experienced with this surgery¹⁶ have suggested that the lack of significant improvement in the former series may be related to the target localization process.

Despite the advanced age of the patients in this series, there were no significant complications. None had any motor weakness or visual field deficits. Transient adverse effects included disorientation and urinary incontinence. We dismissed 18 of 20 patients from the hospital on the first postoperative day, the balance leaving within 3 days.

We believe this study provides evidence that unilateral pallidotomy is safe and beneficial for PD patients with motor fluctuations resistant to medical therapy. Additionally, the safety and benefits with pallidotomy may occur regardless of age, as our first 11 patients ≥65 (median = 72; mean = 71.4 years) experienced benefit of similar magnitude to their younger counterparts. We believe these conclusions are conservative as this study reports the experience of the first 20 consecutive patients at our institution.