Comparison of subthalamic nucleus and globus pallidus deep brain stimulation in patients with Parkinson's disease: A systematic review of literature

Background: Deep brain stimulation (DBS) is regarded as a viable therapeutic choice for Parkinson's disease (PD). The two most common sites for DBS are the subthalamic nucleus (STN) and globus pallidus (GPi). In this study, the clinical effectiveness of these two targets was compared. Methods: A systematic literature search in electronic databases were restricted to English language publications 2010 to 2021. Specied MeSH terms were searched in all databases. Studies that evaluated the Unied Parkinson's Disease Rating Scale (UPDRS) III were selected by meeting the following criteria: (1) had at least three months follow-up period; (2) compared both GPi and STN DBS; (3)at least ve participants in each group; (4)conducted after 2010. Study quality assessment was performed using the Modied Jadad Scale. Results: 3577 potentially relevant articles were identied,3569 were excluded based on title and abstract, duplicate and unsuitable article removal. Eight articles satised the inclusion criteria and were scrutinized (458 PD patients). Majority of studies reported no statistically signicant between-group difference for improvements in UPDRS ш scores. Conclusions: Although there were some results in terms of action tremor, rigidity, and urinary symptoms, which indicated that STN DBS might be a better choice or regarding the adverse effects, GPi seemed better; but it cannot be concluded that one target is superior. Other larger randomized clinical trials with longer follow-up periods and control groups are needed to decide which target is more ecient for stimulation and imposes fewer adverse effects on the patients. ± 7.02 years old. The duration of follow-up varied from six months to 36 months. Two studies were conducted in the USA (34, 35), two studies in Europe (36, 37), and four studies in Asia (38–41).


Introduction
Parkinson's disease is the most common movement disorder, the second common progressive, disabling, and neurodegenerative disease, which is expected to be as prevalent as 12.9million cases by 2040 (1)(2)(3)(4). Although acute diseases such as stroke are decreased in developed countries, neurodegenerative diseases are increasing and affecting most of their population (5). PD prevalence is estimated at 160 per 100000 in individuals over 65 years old (6). Parkinson's disease is characterized by slow movements, rigidity, and low amplitude movements without the previous automaticity. Gait problems are among the most prominent and disabling signs of this disorder which progress as time passes (7). Various genetic, environmental, lifestylerelated factors and aging have been proposed as the riggers for Parkinson's disease initiation (8,9). Classically, Parkinson's disease is attributed to the progressive death of dopaminergic neurons of basal ganglia and hyperactivity of striatopallidal pathway in the dorsal striatum due to loss of dopamine signaling and presence of Lewy bodies and Lewy neuritis. PD patients suffer from various motor and non-motor symptoms that negatively impact their daily lives (10). Levodopa, a dopamine precursor, was developed for the rst time in 1911 (11). It has been used as an e cacious drug for PD treatment for over 50 years, but it has been shown that its e cacy declines as the disease progresses (6,12) . Surgical treatment of movement disorders started in 1987 by targeting the ventral intermediate nucleus of the thalamus (13,14) by the DBS technique. DBS is a common and effective surgical treatment option that alleviates motor symptoms. It was introduced about three decades ago and is recently performed on several new targets in the brain, including the Subthalamic nucleus (STN) and Globus Pallidus internus (GPi) (10, 15 ,16). Performing this technique on the STN and Gpi has proven to be highly effective and safe, but several adverse effects like verbal uency problems are reported. This is attributed to the stimulation site in some studies (17). This systematic review investigated the e cacy of STN and GPi deep brain stimulation on UPDRS score outcomes in Parkinson's disease and its related adverse effects.

Search plan
Potentially relevant English-language articles, published from 2010 to 2021, were recognized by searching in Embase, Cochrane Library, and PubMed databases. We searched for studies including both targets (GPi and STN) and their related aspects. Search terms were deep brain stimulation (DBS) [

Study selection criteria
For this research, we included the clinical studies that (1) contained evaluated uni ed PD rating scale (UPDRS) III before and after deep brain stimulation ;(2) compared GPi-DBS and STN-DBS for PD; (3) recruiting more than ve subjects in GPi and STN groups ;(5) had a description of adverse events; (5) had more than three months of follow-up period; (6) Availability of English full text (7). The selection process took place in two phases: 1. title and the abstract selection; 2. full-text selection. These processes were done by all researchers independently. The full texts of the articles were ordered and compared according to the eligibility criteria. Any disagreements were resolved by discussion.

Data collection
The database was created by gathering the data about study design, e cacy, symptoms, materials, and population. Additionally, different adverse events and UPDRS III scores were extracted from articles.

Results
In total, 3577 potentially related articles were identi ed from the following databases: 834 studies from Embase, 676 studies from Cochrane Library, and 2067 studies from PubMed. After the primary evaluation, studies with unsuitable titles and abstracts were excluded (3461), duplicate articles were removed (102), 14 articles remained for further assessment. Two systematic reviews (18,19), two meta-analysis (20,21), and two letters (22,23) were also removed. The full texts of the remaining eight articles were scrutinized (Fig. 1).

Study characteristics
The features of the Eight studies evaluated in this meta-analysis are shown in Table 1. This study included 485 PD patients (226 in the GPi group and 259 in the STN group). The mean and SD age of GPi group participants was 54 ± 4.51 to 64.26 ± 8.79 years old, and STN group participants were 56.16 ± 9.6 to66.5 ± 7.02 years old. The duration of follow-up varied from six months to 36 months. Two studies were conducted in the USA (34, 35), two studies in Europe (36, 37), and four studies in Asia (38-41).   is also true about comparing UPDRS on medication before surgery to UPDRS on medication and on stimulation after surgery ( p = 0.038). (28) In another study conducted by Odekerken et al., more improvements were reported the off-drug phase UPDRS-ME score after three years in the STN group (p = 0.04 while no between-group differences were shown in the on-drug phase (27 events for the two groups; only minor events were reported (37). Wong et al. found that the most common adverse events were problems with DBS lead hardware and hemorrhage, which were minor and controlled easily (34).

Discussion
As a surgical technique, DBS has gained wide popularity in treating patients with advanced Parkinson's disease. In this review, the UPDRS ш score and adverse events were evaluated to measure the outcomes after GPi DBS or STN DBS as common targets.
Tremor is an important sign of PD and its pathogenesis is attributed to the disruption of the cerebello-thalamo-cortical pathway.It has been shown that GPi has a role in tremor pathogenesis too, and its stimulation could even trigger tremor through its efferent bers because of stimulation spread into the pallido-thalamic out ow bers (24).the practical difference of STN and GPi in terms of tremor is mainly because of their distinct connectivities in the tremor circuit, where STN has afferent and efferent connections with cerebello-thalamo-cortical network and GPi has only efferent connections (24). The better control of resting tremor in the dorsal STN DBS can be because of dentato-rubro-thalamic bers modulation in the posterior sub-thalamic region or relation of bers to motor and premotor cortex, while assessment of action tremor did not reveal any relation with speci c region (24). Direct anti-dyskinesia effects of STN DBS were exerted by stimulation of the area above STN where pallidothalamic, pallidosubthalamic and subthalamopallidal bers are present (30). Wong et al. found that in the short term, rst six months after surgery, for rigidity, tremor and bradykinesia, STN-DBS had much more positive effects than GPi-DBS. They evaluated tremor according to UPDRS3( item 21), which combines postural and kinetic tremor, and it cannot differentiate between re-emergent postural tremor and pure kinds of them (24). Additionally, the GPi dyskinesia reduction effect was better than STN in the onphase, but this result was thought to be an artefact of the study design since patients in the STN DBS group had less severe dyskinesias than the standard assessments, and they took less medication than the GPi group in their regular life.Meanwhile all patients received the same amount of Levodopa as the baseline amount; (26,27). STN-DBS decreases detrusor hyperre exia and increases bladder capacity via modulation of bladder afferents and central sensory processing, while the impact of GPi on urinary problems is not well understood (31). Additionally, sleep-related problems such as insomnia, daytime sleepiness, and restless legs are common in Parkinson's disease patients. It has been shown that STN-DBS has a role in improving the objective polysomnographic features of sleep quality, which may be due to the resolution of motor symptoms and not the effect on the sleep center (31). Regarding the adverse events, the reason for fewer complications of Globus pallidus in swallowing motor function is not clearly known, but it can be attributed rstly to reciprocal connections between pedunculopontine nucleus(PPN) and Gpi or STN. GPI has an inhibitory effect on it, and STN has excitatory effects ,and secondly the theory that patients who received STN DBS did not respond well like those who underwent GPi DBS. (28). Several studies had short follow-up periods, so determining the long-term e ciency of DBS in PD patients needs studies with more follow-up duration.
In addition, to assess the effect of medications and DBS on pain symptoms, a control group should be considered in futurestudies. The study conducted by Fan et al. was retrospective, and it is clear that randomized controlled trials are much better for these kinds of research (30). In longitudinal studies, like the studies in this review, the follow-up period is important because there will be dropouts due to different reasons. In the study conducted by Odekerken et al., at the third year of their follow-up, the sample size was 70 percent of the baseline cohort ( there were withdrawals in the follow-up perio). However, the sample size of the study was considered suitable for statistical analysis (27). The ndings of this study must be interpreted cautiously. It can be can generally hypothesized, but not de nitely stated, that STN-DBS and GPi-DBS are both suitable stimulation sites. However, STN had better outcomes regarding urinary symptoms, pain, dyskinesia, and action tremor. The main limitation of this systematic review study was moderate to low quality of the included studies (according to the Modi ed Jadad scale).

Conclusion
According to the results of this systematic review, although STN was better in terms of improving action tremor, and GPi was related to less adverse events in the studies, it cannot prove the superiority of any of the two targets. It is clear that more and more clinical trials with large sample size, longer follow-up periods and more speci c outcome assessments are needed to conclude which target is de nitely better.

Declarations
The author declares no competing interests.