Unilateral pallidotomy

Although thalamic targeting certainly helped with PD tremor, the pallidum became a very attractive target once it was realized that many of the levodopa-responsive symptoms, including bradykinesia, rigidity, tremor and gait changes, had the potential to improve if the pallidum was lesioned or stimulated. Unilateral pallidotomy has not been compared directly with unilateral pallidal DBS is in a head-to-head randomized study. However, Blomstedt et al. [30] analyzed the effect of unilateral pallidotomy and contralateral pallidal stimulation on symptoms within the same PD patients. Five consecutive patients received pallidotomy contralateral to the more symptomatic side of the body. At a mean of 14 months later, the same patients received pallidal DBS on the side contralateral to the site of the pallidotomy. Evaluations were performed at a mean duration of 37 months (range 22–60) after the pallidotomy and 22 months (range 12–33) after pallidal DBS. The scores of appendicular item numbers 20–26 of the UPDRS-III (motor severity) were compared between the side that received pallidotomy and the side that received pallidal DBS. The side contralateral to pallidotomy improved by 32.5%, whereas the side contralateral to the pallidal DBS improved by 25%. Dyskinesias and dystonia items 32–35 of the UPDRS-IV (motor complications) were rated separately for each side. The side contralateral to the pallidotomy improved by 91%, whereas the side contralateral to the pallidal DBS improved by 59%. With regard to safety, no pre- or postoperative complications occurred in connection with the initial pallidotomies. Concerning pallidal DBS procedures, two patients exhibited moderate dysarthria and one patient showed a severe worsening of dysphonia following DBS. The dysarthria was partly reversible with alteration of stimulation, but the dysphonia was irreversible regardless of stimulation changes. The DBS system was affected negatively by external factors in two cases. In one case, the neurostimulator was damaged during electrical welding and resulted in replacement. In the second case, the patient’s hearing device interacted with the DBS system creating a headache when this was used. This required cable revision without any improvement.

It should be noted that in this study patients always received pallidotomy contralateral to the more symptomatic side, which could explain the greater improvement. In other words, the first operated, most affected, side may have the greatest improvement regardless of the procedure chosen. The authors concluded that “the effect of pallidotomy to be at least as good and as safe as that of pallidal DBS.” To further address the veracity of this statement in the absence of head-to-head studies, we can separately examine trials of unilateral pallidotomy and those of unilateral pallidal DBS.

Two level I studies exist that compared unilateral pallidotomy with best medical care. In 1999, de Bie et al. [31] conducted a prospective, randomized, single-blinded, multicenter trial in which 37 patients were randomized to unilateral pallidotomy within 1 month of enrollment or to best medical care (with pallidotomy 6 months later). The primary outcome was the change in the motor UPDRS “off” score between baseline and 6 months. The pallidotomy patients median motor UPDRS “off” score improved by 31% (from 47 to 32.5), whereas the control group with best medical care slightly worsened by 8% (from 52.5 to 56.5). There was a statistically significant difference between the two groups (P<0.001). There were also statistically significant improvements in scores for the pallidotomy group on the Barthel index (a scale measuring activities of daily living [ADLs])(P=0.004), the ADL section of the UPDRS (UPDRS-II; P=0.002), and the Schwab and England Scale (P<0.001). During the “on”-phase assessment, median score in the dyskinesia rating scale improved compared with the control group by 50% (P=0.02). Patients’ symptom diaries in the pallidotomy group revealed that they had more “on” time without dyskinesias compared with the control group by an average of 2.8 h/day (P=0.02). Similarly, Vitek et al. [32] conducted a randomized trial with blinded ratings of 36 patients comparing the effects of unilateral pallidotomy (n=18) versus medical therapy (n=18) over a 6-month period. The primary outcome was change in the total UPDRS score, and it was found that pallidotomy induced a 32% improvement compared with a 5% decline in the control group (P<0.0001). Dyskinesias and motor fluctuations significantly improved only in the surgery group, although the study was not powered specifically to detect changes in these secondary indices. In general, the “off” UPDRS motor score has been reported to improve by 31–32% postoperatively [31, 32].

The efficacy of unilateral pallidotomy for improving motor symptoms has also been supported by a number of level III studies [33–42]. In addition to improvement in the symptoms of parkinsonism, pallidotomy has consistently been shown to improve dyskinesias, particularly contralateral to the side of the lesion [33–37, 40, 43]. Thus, unilateral pallidotomy seems to improve parkinsonism, dyskinesias and motor fluctuations, and several additional studies have reported on its safety.

A level II study by Perrine et al. [44] examined the impact on cognitive function. The group assessed change in a neuropsychological battery of tests over a 1-year period as the primary outcome. These tests included the Mini Mental Status Examination (MMSE), Controlled Oral Word Association Test, Trail Making Test (part A and B), Symbol Digit Modalities Test, Stroop test, Wisconsin Card Sorting Test, and the Beck Depression Inventory (BDI). There was no statistically significant difference between the two groups for any neuropsychological test at baseline or on retesting. A level III study by Trepanier et al. [41], on the other hand, reported an improvement in sustained attention but a decline in working memory and frontal executive functioning, along with behavioral changes of a “frontal” nature in about 25% of subjects.

Visual-field defects may be another side effect of pallidotomy, as described in a level III study by Biousse et al. [45]. Using microelectrode guidance, the group reported that 7.5% of patients (3/40) had visual defects, discovered with Goldmann visual-field testing, including contralateral homonymous superior quadrantanopias, likely related to pallidotomy. The authors noted that lesion locations in patients with visual-field defects significantly differed from those in patients with no visual-field defects in that they were more ventral, adjacent to or extending into the optic tract. They reported the mean distance from the ventral edge of the primary lesion to the optic tract was −1.67±0.4 mm in patients with visual-field defects and +1.3±0.5 mm in those without (P = 0.00015 by Student’s t-test). Thus, the authors changed their technique during the study and increased their lesioning threshold from 0.5 to ≥1.0 mA in order to place the lesion at a more distant target from the optic tract. By doing this, they were able to reduce the incidence of visual-field defects from 11% (2/18) to 4.5% (1/22).

In another level III study, Hariz and De Salles [46] analyzed complications of posteroventral pallidotomy in 138 consecutive patients who underwent 152 pallidotomies (12 patients had bilateral pallidotomies). Transient adverse reactions, typically lasting less than 3 months, appeared in 18% of the patients (n = 25) including subcortical hematoma (2), meningitis (1), facial paresis (3), leg paresis (2), dysarthria (5), limb dyspraxia (2), seizure (2), sialorrhea (4), fatigue (3) and confusion (5). Long-term complications that lasted more than 6 months were noted in 10% of the patients. Other complications described to have occurred alone or in various combinations in 14 patients included fatigue and sleepiness (2), worsening of memory (4), depression (1), aphonia (1), dysarthria (3), scotoma (1), slight facial and leg paresis (2) and delayed stroke (2).

Unilateral pallidal DBS

Many of the concerns described above that are associated with unilateral pallidotomy are not typically seen with unilateral pallidal DBS. Ideally, we would like to compare unilateral pallidotomy with unilateral pallidal DBS in a randomized fashion, but no such trial exists. In addition, there are no trials directly comparing unilateral pallidal DBS with best medical therapy. However, the efficacy of unilateral pallidal DBS has been well demonstrated through a randomized controlled trial by Zahodne et al. [47] in which pallidal DBS was compared with STN-DBS. In this trial, quality-of-life measures were compared in PD patients who received unilateral pallidal DBS (n = 22) and unilateral STN-DBS (n = 20). The primary outcome was change in the different domains of the quality of life before and 6 months after surgery in patients randomized to receive either therapy. The 39-Item Parkinson’s Disease Questionnaire (PDQ-39) summary index (PDQ-SI) score was used to assess quality of life, with higher scores indicating poorer quality of life. Patients with GPi-DBS had a significant 38.1% average reduction in their PDQ-SI scores postoperatively. Analysis of subscales revealed that GPi-DBS patients improved significantly with regard to mobility, ADLs, stigma and social support. On average, the GPi-DBS group exhibited a 28% improvement in the motor UPDRS “off” score compared with baseline. Scores on the Beck Depression Inventory 2nd edition (BDI-II) improved significantly after GPi-DBS. No significant safety concerns were reported in this trial. In a level III study by Rodrigues et al. [48], the change in quality of life following GPi-DBS was measured in 11 patients, in which four patients received unilateral DBS. Outcome measures included the PDQ-39 questionnaire and motor assessments. The PDQ-39 scores were reduced by 30% in the unilateral patients. There was a reduction in the motor UPDRS “off” scores by 19%.

Safety concerns that exist with unilateral pallidal DBS include concerns that occur with stereotactic surgery in general, risk of hemorrhage, infection and neurological deficits caused by local tissue damage [49]. Furthermore, it has been shown that patients undergoing unilateral DBS (GPi or STN) may have cognitive declines on psychomotor processing speed in PD patients compared with age-matched PD controls [50]. Other side effects reported include worsening of concentration and memory, but the sample sizes of those receiving unilateral procedures are too small to make general conclusions [48]. Although changes in cognition may be seen, these are typically isolated to subscales and are not declines in global cognition. Okun et al. [51] reported on 23 subjects who received unilateral GPi-DBS and found that there was no change of mood or cognitive outcomes from pre- to post-DBS in the optimal setting at a 7-month follow-up. Thus, overall, unilateral pallidal DBS is a well-tolerated procedure.

Bilateral pallidotomy

While unilateral procedures for asymmetric disease can be of benefit, PD is a progressive bilateral disease, requiring bilateral treatment. Bilateral pallidotomy and bilateral pallidal DBS have been performed to address this issue. There are no level I or level II studies comparing bilateral pallidotomy with best medical therapy or pallidal DBS. A number of level III studies have shown that bilateral pallidotomy may improve the cardinal features of PD and reduce motor fluctuations. Scott et al. [52] reported that bilateral posteroventral pallidotomy produced an improvement in the mean total UPDRS score of 53% (from 88 ± 22 to 41 ± 21). In five patients receiving staged bilateral pallidotomy, Intemann et al. [53] reported that the UPDRS total scores improved by 29% (P = 0.04) and motor UPDRS “off” scores improved by 36% in the “off” state (P = 0.01). The authors comment that the overall improvements in these scores were strongly influenced by specific improvements in tremor, bradykinesia and medication-induced dyskinesias. Counihan et al. [54] reported on the results of 14 patients undergoing staged bilateral pallidotomy and found that dyskinesias were greatly improved and there were significant reductions in “off” time (67%) and ADL “off” scores (24%), and a nonsignificant reduction in “off” motor score (39%). Additional level III studies also report similar findings [55–62].

However, bilateral pallidotomy is generally not considered to be an acceptable treatment for PD, due to its risk-to-benefit profile. According to the task force of the Movement Disorders Society, “serious concerns have been voiced on the risk of speech, balance, gait, and cognitive problems consequent to bilateral surgery” when discussing pallidotomy [63]. Merello et al. [64] conducted a level I prospective study of PD patients who were randomized either to simultaneous bilateral pallidotomy or simultaneous unilateral pallidotomy with contralateral GPi stimulation. However, the first three patients who received bilateral pallidotomy experienced severe side effects, resulting in termination of the trial. These adverse effects included depression and apathy. There was also a dramatic worsening of speech, drooling, swallowing, freezing, ambulation and falls. The reported side effects of bilateral pallidotomy that have been described in class III studies are summarized in Table 28.1. Studies after 1995 and those with a minimum of five patients are included in the table. A systematic review of the morbidity and mortality following pallidotomy reported the incidence of permanent adverse effects being as high as 60% in bilateral pallidotomy cases [67]. Due to these reported complications of bilateral lesioning procedures and the advent of DBS, bilateral pallidotomy has largely been abandoned.

Although the above studies and findings are certainly compelling, is it possible that bilateral pallidotomy received a worse reputation than it deserves? There certainly are factors that may have negatively influenced the outcome of bilateral pallidotomy. In the early years of pallidotomy, preoperative imaging was not sophisticated (CT rather than MRI, or, at best, low-resolution MRI). The low-resolution imaging combined with the permanent nature of lesioning contributed to the perceived need for extensive “mapping” of the GPi in many centers. Thus, multiple microelectrode tracks were made, sometimes exceeding six tracks per side [21]. In addition, most surgical reports do not mention details regarding the angles of the chosen trajectories from the burr hole to the target, and it thus remains unknown which cortical areas or subcortical nuclei such as the caudate were pierced by these multiple electrode tracts. Conceivably, many of the adverse events reported were at least in part attributable to excessive penetration of structures other than the pallidum along the trajectories through the frontal lobes.

Bilateral pallidal DBS

On the other hand, the rise in popularity of DBS has also led to the accumulation of a large body of evidence regarding its efficacy and safety that continues to grow and be refined. Regarding bilateral pallidal DBS, several studies have established its use for improving parkinsonism, reducing motor fluctuations and dyskinesias. The Deep-Brain Stimulation for Parkinson’s Disease Study Group [68] performed a prospective, double-blinded crossover study in patients with advanced PD in whom electrodes were implanted in the GPi (n = 38) and the STN (n = 96). Blinded evaluation of 35 patients implanted with bilateral pallidal stimulation at 3 months was associated with a mean improved of 32% and a median improvement of 37% in the motor UPDRS “off” score (P < 0.001). Unblinded evaluations of 36 patients at 6 months showed that bilateral GPi-DBS improved motor UPDRS “off” scores by 33.3%; in the “on” state, UPDRS motor scores were improved by 26.8%. Home diary assessments showed that between baseline and the 6-month time point, the percentage of time with poor mobility was reduced from 37 to 24% (P = 0.01). In addition, the percentage of time with good mobility and without dyskinesias during the waking day increased from 28 to 64% (P < 0.001).

The benefits of pallidal stimulation were also shown by Anderson et al. [69] in a randomized, blinded parallel-group study of patients receiving either bilateral pallidal (GPi) (n = 10) or bilateral subthalamic (n = 10) stimulation. In the GPi stimulation group, evaluation after 12 months showed that off-medication, on-stimulation scores improved by 39%, ADLs improved by 18% and dyskinesias improved by 89%. In another level I study, Follett et al. [70] randomized patients to pallidal stimulation (n = 152) or subthalamic stimulation (n = 147) and compared 24-month outcomes. The primary outcome measure was change in the motor UPDRS “off” scores through a blinded assessment. In the pallidal group, there was a 28.2% improvement in the motor UPDRS in the off-medication, on-stimulation state. There was only a slight improvement, 5.3%, in the motor UPDRS score in the on-medication, on-stimulation state. The efficacy of pallidal stimulation was further demonstrated in a randomized controlled trial by Odekerken et al. [71]. In this study, the investigators assessed the level of functional improvement provided by either bilateral GPi or STN stimulation using the Academic Medical Center Linear Disability Scale (ALDS) as the primary outcome. This scale quantifies functional status in performing basic and complex ADLs on a scale of 0–100, with lower scores indicating more disability. This score is interesting because it is weighted by time spent in the “off” phase and “on” phase, capturing disability throughout the day. In the GPi group (n = 44), the ALDS score improved by about 4% at 12 months. In the 62 patients who completed the secondary outcome measures, the off-medication, on-stimulation UPDRS motor score improved by 26% and UPDRS-ADL score improved by 21.7% between baseline and 12 months. In the on-medication, on-stimulation state, dyskinesias improved by about 56% as measured by the clinical dyskinesia rating scale.

Thus, a clear benefit for bilateral GPi stimulation has been established with several high-quality studies, as described above. However, as with any surgical therapy, the possibility of complications exists. A variety of adverse effects have been reported with bilateral pallidal DBS, but generally the rates of these events are significantly less compared with bilateral pallidotomy. Serious adverse events such as intracranial hemorrhage have been reported at a rate of between 1% [70] and 9% [68]. Altered mental status/confusion can be seen in as many as 20% of cases, but is often transient and resolves without sequelae [71]. Speech problems may occur in up to 29% of cases [71]. Other rare side effects include, but are not limited to, stroke [69], seizure [68, 71], falls [70, 71], stimulation-induced side effects such as dyskinesia [68, 70, 71], psychiatric symptoms [70], implantation site infections [70, 71] and death [70]. In summary, bilateral pallidal DBS is preferred to pallidotomy for several reasons. Speech changes, falls, and lack of sustained effect are more likely with pallidotomy compared with DBS based on the information available. This is likely because DBS may be more forgiving than pallidotomy. If a lesion is made in a suboptimal place, pallidotomy cannot be reversed, whereas DBS can be adjusted. It is important to note that studies of pallidotomy were conducted during a time where imaging and surgical techniques were not optimized, nor were detailed reports of the results documented or available. Thus, bilateral pallidotomy may have been doomed to fail since its inception. However, all things being equal, the point still remains that DBS is reversible while lesioning is not, making DBS the preferred therapy.