Microelectrode mapping of the globus pallidus. Characteristic firing patterns seen in the striatum (1), globus pallidus externus (GPe) (2), border cells (3) and globus pallidus internus (GPi) (4).
Initial programming occurs several weeks after implantation, in order to give time for resolution of any postoperative “lesion effect,” which can result in improved symptoms (and therefore more difficult initial programming) for 1–2 weeks after the macroelectode is placed. Stimulation is started at a low amplitude and increased gradually over time. There are four contacts on the electrode, and stimulation can be delivered in a monopolar or bipolar fashion. The most effective contact configuration and frequency settings provide the greatest benefit with the least number of side effects due to the spread of current to surrounding structures. Generally, programming may require multiple programming sessions and can take up to 4–6 months before full optimization of function is reached [25].
Globus pallidus internus deep-brain stimulation outcomes
The use of GPi-DBS produces a significant improvement in the off-medication Unified Parkinson’s Disease Rating Scale (UPDRS) III scores with respect to baseline, by about 40%. It also improves cardinal features and activities of daily living (ADLs) while prolonging the “on” time with good mobility and without dyskinesias [26]. There is improvement in the ADL score of about 40%, and 66% improvement in levodopa-induced dyskinesias [9, 27]. In advanced PD, pallidal DBS can also ameliorate “off”-period dystonia, cramps and sensory symptoms [28]. The duration of benefit from GPi-DBS is sustained, with retention of beneficial effects demonstrated in long-term follow-up [26, 29].
The effect of GPi-DBS is predominantly stimulation related. However, in addition to the aforementioned “lesion effect” that can last for 1–2 weeks after macroelectrode implantation, there are some benefits found when off stimulation even after this time, with some studies reporting an improvement in levodopa-induced dyskinesias at the 6-month follow-up during on-medication periods despite the stimulator being turned off [9].
Comparing subthalamic nucleus and globus pallidus internus deep-brain stimulation
Multiple studies have compared the outcomes of GPi and STN stimulation to determine which stimulation target is more effective for the treatment of PD symptoms. The first double-blinded crossover study comparing GPi-DBS and STN-DBS for PD was published in 2001. The results showed an improvement in dyskinesias in both STN-DBS and GPi-DBS of 58 and 66%, respectively, with an improvement in motor function of 49 and 37%, respectively, and an improvement in “on” time [30]. Daily levodopa dose equivalents were significantly reduced in the STN group (from 1218.8 mg at baseline to 764.0 mg at 6 months), and were unchanged in the GPi group (1090.9 mg at baseline and 1120 mg at 6 months). With DBS at either target, there is an increase in “on” time of about 4.6 h a day [31]. In 2005, Anderson et al. [32] found that there was insufficient evidence to support the superiority of one target over the other. Twelve months of stimulation at either location improved baseline rigidity, bradykinesia, tremor and axial symptoms (speech, gait, posture and postural stability), along with ADLs. In 2010, the Veterans Affairs Cooperative Studies Program (CSP 468) – a multicenter, randomized, blinded study – found similar improvements in motor function, quality of life and adverse events in both groups [33]. Another comparative trial in 2013 found that, during the on-medication state, there was a greater antidyskinetic effect in the GPi group, whereas in the STN group there were better off-medication motor symptoms and improvements in disability [34]. However, the results of this trial may have been confounded by some of the methods. The on-medication dosages were similar in both the GPi and STN groups, which may have increased the dyskinesia burden in the STN group. In addition, the DBS programmers were not blinded to subject assignment and could have differentially managed the settings and medications. Table 18.1 looks at seven randomized controlled trials comparing STN-DBS and GPi-DBS to date.
Blinded, randomized controlled trials comparing globus pallidus internus (GPi) and subthalamic nucleus (STN) deep-brain stimulation (DBS)
Author (country of origin) | Year | No. in sample (GPi/STN) | Primary outcome measure(s) | Primary outcomes | Other outcomes |
---|---|---|---|---|---|
Anderson et al. [32] (USA) | 2005 | 11/12 | Change in off- medication UPDRS-III | GPi=STN | Trend toward greater improvement in bradykinesia in STN |
Rothlind et al. [35] (USA) | 2007 | 23/19 | Neuropsychological testing | Unilateral and bilateral: WAIS-R digit symbol GPi > STN; backward digit span GPi < STN; otherwise GPi=STN | Decrease in verbal fluency and working memory for both GPi and STN; trend toward decline in arithmetic and Stroop with STN |
Okun et al. [36] (USA) | 2009 | 23/22 | Visual Analog Mood Scale; verbal fluency | Unilateral: GPi=STN in optimal DBS state | Motor UPDRS GPi=STN; Adverse mood effects with ventral stimulation both targets; worsening of verbal fluency in “off” STN-DBS |
Zahodne et al. [37] (USA) | 2009 | 22/20 | Quality of life, measured by PDQ-39 summary index | GPi > STN | STN group: reduction in letter fluency |
Follett et al. [33] (USA) | 2010 | 152/147 | Change in motor function measured by UPDRS-III | Bilateral: GPi=STN | STN required lower dopaminergic medication; visuomotor processing speed declined more after STN; depression worsened after STN and improved after GPi |
Rocchi et al. [38] (Italy) | 2012 | 14/15; 22 of the 29 patients were part of the VA Coop study Follett [33] | Step initiation measured by APAs, first-step length and velocity | Bilateral: postoperative APA same for GPi and STN; first-step velocity worsened in STN but not GPi | Although equal, APA for both GPi and STN were worse than non-DBS PD controls |
Odekerken et al. [34] (The Netherlands) | 2013 | 65/63 | Functional health as measured by the weighted ALDS; composite score for cognitive, mood and behavioral effects | Bilateral GPi=STN | STN associated with more improvement in “off”-phase motor disability |
ALDS, Academic Medical Center Linear Disability Scale, a functional health measure comparing time spent in the “off” phase compared with the “on” phase; PDQ-39, Parkinson’s Disease Questionnaire; APA, Anticipatory postural adjustments – feed-forward postural preparation preceding voluntary step initiation; UPDRS, Unified Parkinson’s Disease Rating Scale.
Dyskinesias improve significantly in both STN-DBS and GPi-DBS. This occurs by a direct dyskinesia suppression effect of GPi stimulation [39] or a reduction in levodopa dosage by STN stimulation. Levodopa and levodopa-equivalent dosages in general are significantly reduced after STN-DBS but not after GPi-DBS. The improvements in motor function after GPi-DBS have been shown to be sustained for up to 5–6 years after surgery [40].
Some studies endorse a trend toward increased improvement in bradykinesia and rigidity with STN-DBS compared with GPi-DBS [3, 32, 36]. However, GPi-DBS may be more effective against axial deficits resistant to levodopa [41]. The higher DBS settings required to obtain optimal benefit at the GPi may lead to more frequent replacements of pulse generators [33, 34]. However, GPi-DBS may be associated with a greater reduction in nonmotor features, including an improvement in depression [42].
Alternatively, unilateral GPi-DBS can be an effective option for PD patients with baseline asymmetry in symptoms. In a study with a 3.5 year follow-up, patients implanted with unilateral GPi-DBS were less likely to require a future second lead implantation than patients with unilateral STN-DBS [43]. The most likely cause for implantation in the contralateral side was insufficient relief of motor symptoms. The odds of proceeding to bilateral DBS were 5.2 times higher in the STN-DBS group than in the GPi-DBS group [43].
Complications
Complications from GPi-DBS include perioperative infection, hardware fracture and premature battery failure [9]. Transient adverse effects include paresthesias and tonic muscle contractions contralateral to the side of stimulation. Dysarthria can also be observed, likely due to stimulation of the internal capsule. Stimulation fields too close to the optic tract (ventrally) can cause photopsias or nausea. These adverse effects are relieved upon turning off the stimulation and can be managed in the long term by reducing the voltage or considering an alternative stimulating electrode to maintain the antiparkinsonian benefit with minimal side effects [9].
Cognitive changes following GPi-DBS, if present, appear to be mild, limited to specific areas and of debatable clinical significance. Early research suggested that older patients may be at a higher risk for significant cognitive or behavioral decline after bilateral STN-DBS, while GPi-DBS was thought to be safer [44]. More recent comparative trials have shown that the cognitive side effects are actually similar between the two groups. Subscores of neurocognitive function in the VA Cooperative study revealed similar slight decrements in all measures for both targets, except for a greater decline in visuomotor processing speed after STN-DBS [33]. Declines in letter verbal fluency, a consistent finding after STN-DBS, are present to a lesser extent with GPi-DBS [36]. Compared with best medical therapy, GPi-DBS was associated with lower performance on one measure of learning and memory that requires mental control and cognitive flexibility [45].
Regarding behavioral complications, postoperative delirium and perioperative anxiety may also be less common with GPi-DBS when compared with STN-DBS [32]. As noted above, GPi-DBS may result in a slight improvement in depressive symptoms compared with a decline following STN-DBS, although the difference is small and the clinical meaningfulness is debatable [33].
Impulse control disorders may improve following medication reduction after DBS surgery; however, there is emerging evidence that impulse control disorders could also be precipitated by DBS [41]. It does not appear that the target (STN vs GPi) influences the extent to which this complication may occur, but the numbers studied so far are small.
Conclusion
The use of GPi-DBS is an effective treatment option for the management of PD-related motor fluctuations and dyskinesias. Selection of the optimal surgical target must take into account the full spectrum of both motor and nonmotor symptoms that affect quality of life in individual patients. The evidence suggests an overall similar efficacy for both STN-DBS and GPi-DBS but with differential cognitive side effect profiles, although these are of uncertain clinical relevance. Individual factors may make one target more attractive. Table 18.2 summarizes the various characteristics that may lead to the decision to choose STN over GPi as the stimulation target. Ultimately, the decision between targets also depends on the neurosurgeon’s preference and expertise.
Decision making: patient characteristics and goals favoring globus pallidus internus (GPi) and subthalamic nucleus (STN) deep-brain stimulation

Stay updated, free articles. Join our Telegram channel

Full access? Get Clinical Tree

