How is cognition in subthalamic nucleus deep brain stimulation Parkinson’s disease patients?

Abstract The impairments in cognitive functions such as memory, executive function, visuospatial skills and language in Parkinson’s disease (PD) are drawing increasing attention in the current literature. Studies dedicated to investigating the relationship between subthalamic nucleus deep brain stimulation (STN-DBS) and cognitive functioning are contradictory. This systematic review aims to analyze the impact on the cognitive functioning of patients with PD and STN-DBS. Articles published in the 2007-2017 period were retrieved from the Medline/Pubmed databases using PRISMA criteria. The analysis of 27 articles revealed many conflicting results, precluding a consensus on a cognitive functioning standard and hampering the establishment of a neuropsychological profile for PD patients who underwent STN-DBS surgery. Further studies investigating this relationship are needed.

T he diagnosis of PD is performed using clinical criteria by trained professionals, such as neurologists. These criteria are based on the identification of clinical manifestations and pure motor symptoms. Patients with PD present, in addition to motor impairments, non-motor impairments manifesting as a variety of neuropsychiatric symptoms, 1,2 changes in sleep, behavior and cognition, 3,4 and which may lead to dementia. 5,6 The impairments in cognitive functions, such as memory, executive function, visuospatial skills and language in PD, are drawing increasing attention in the current literature. 6 One in three patients with PD presents cognitive impairment at the time of (or soon after) diagnosis, which progressively worsens and may even cause dementia in the later stages of the disease. 7 Since 1940, surgical treatment of PD has been performed. More recently, since 1998, ablation has given rise to deep brain stimulation (DBS) targeting the subthalamic nucleus (STN) or globus pallidus internus (GPi). 5,8 The target most chosen by centers performing the surgery is the STN due to the possibility of decreasing drug doses and, consequently, reducing adverse effects.
The literature points to evident motor and QoL improvement after DBS in patients with PD. However, studies dedicated to investigating the relationship between STN-DBS and cognitive functioning are controversial, and further studies investigating this relationship are needed.
In this context, the investigation of the cognitive effects of STN-DBS in PD becomes paramount. The objective of this study is to analyze the effects of subthalamic nucleus (STN) DBS on the cognition of PD patients through a systematic review. The Preferred Reporting Items for Systematic Review and Meta-Analyzes (PRISMA) Checklist was employed.

METHODS
The systematic review is a type of scientific research that aims to gather, critically evaluate and conduct a synthesis of multiple primary studies. 10

Bibliographic survey
We designed a systematic review of the literature according to the Preferred Reporting Items for Systematic Review and Meta-Analyzes (PRISMA) criteria. The following terms were used: "Deep Brain Stimulation", "DBS", "Cognitive Functions" and "Parkinson Disease" with the Boolean operator "AND". We selected scientific papers published in English between January 2007 and January 2017, with comparative clinical trials in humans, on the Medline/Pubmed databases. Articles published before 2007, systematic reviews, case studies, books chapters and studies using animals were excluded.

Studies selection
Initially, this method retrieved 345 papers ( Figure 1). To refine the research, the following topics were selected: "Parkinson's Disease", "Subthalamic Nucleus", "Deep Brain Stimulation", "DBS", "Cognition" (263), published on the Medline/Pubmed databases (223) between 2007 and 2017 (195). For the papers selected, a title and abstract analysis was performed manually to consider only studies with human clinical trials (66). Literature reviews and case studies were excluded, as were articles containing problems in the methodology, such as absence of: (a) inclusion and exclusion criteria; (b) complete assessment protocol; and (c) pre or post-surgery assessment (27). The researchers selected the articles independently: they considered suitable studies that: (a) evaluated cognition of PD patients with STN-DBS; (b) presented the instruments and domains evaluated; and (c) reported pre and post-surgical results of articles.

RESULTS
The final list of included articles that met the study criteria, in ascending order of year, together with objectives and results, is given in Table 1. A list of studies, grouped according to the effects of DBS on specific cognitive domains, with neuropsychological tasks (carried out in each study assessed) is given in Table 2.
There were 27 studies involving a total of 832 patients with STN-DBS and 458 patients with DBS and/or healthy subjects in the control group who did not undergo surgery. Age ranged from 51 to 67 years, disease duration ranged from 9.7 to 15.75 years, education (when reported) ranged from 1.9 to 14.5 years, while pre-surgical evaluation occurred 2 weeks before surgery and postoperative up to 132 months after surgery (11 years).

2007
To assess changes on evaluation after DBS-STN and their possible correlation with the cognitive result related to the frontal lobe. Patients with STN-DBS improved motor symptoms and reduced medications, but selectively declined in category fluency.
[2] Klempírová et al. Deep brain stimulation of the subthalamic nucleus and cognitive functions in Parkinson's disease.

2007
To evaluate how STN-DBS affects cognitive functions.
Patients treated by STN-DBS tend to worsen in executive functions and logical memory.

2008
To research STN-DBS effects on cognition and mood.
Bilateral STN-DBS did not lead to a significant overall deterioration in cognitive function. However, it has small, long-term detrimental impacts on memory and frontal lobe function.
[5] Witt et al. Neuropsychological and psychiatric changes after deep brain stimulation for Parkinson's disease: a randomised, multicentre study.

2008
To evaluate DBS neuropsychiatric consequences in patients with PD. STN-DBS does not reduce overall cognition or affectivity, although there is a selective decrease in frontal cognitive functions and an improvement in anxiety in patients after treatment, changes not affecting improvements in quality of life.

2008
To determine the effects of unilateral and bilateral STN-DBS on upper extremity motor function and cognitive performance under single and double-task conditions in patients with advanced PD.
Significant declines in cognitive and motor function under modest dualtask conditions with bilateral , but not unilateral STN-DBS.
[7] Lueken et al. Impaired performance on the Wisconsin Card Sorting Test under left-when compared to right-sided deep brain stimulation of the subthalamic nucleus in patients with Parkinson's disease.

2008
To evaluate whether changes in performance on executive tasks after chronic DBS may be predominantly associated with stimulation of only one hemisphere. The STN is not only involved in motor control, but also participates in functions of the cognitive domain. All patients had a significant improvement in motor symptoms postoperatively. Selected aspects of executive task performance were compromised under left -when compared to right-sided stimulation. To evaluate DBS cognitive and behavioral effects. Verbal fluency worsening after DBS, but relatively safe surgery from a cognitive point of view, since short-term worsening of front-executive functions was transient.
[9] Williams et al. Deep brain stimulation plus best medical therapy versus best medical therapy alone for advanced Parkinson's disease 2010 To evaluate whether surgery and best medical therapy improved self-reported quality of life more than best medical therapy alone After 1 year, surgery and best medical therapy improved patient selfreported quality of life more than best medical therapy alone in patients with advanced PD, constituting clinically meaningful differences. To evaluate baseline parameters that contribute to deterioration of cognitive functioning after DBS. Surgical procedure, exact placement of electrode or postoperative management might be more relevant for a decline in executive functioning after STN-DBS, in addition to factors such as age, high levodopa dosages and high scores on the UPDRS III axial subscore in OFF state. To assess long-term PD patients undergoing STN-DBS for 8 years: long-term motor outcome of symptoms that improve in the short and medium-term with STN-DBS; identification of predictors of long-term motor outcome; and long-term cognitive and behavioral outcome. STN-DBS is a safe procedure regarding cognitive and behavioral morbidity over long-term follow-up. However, the global benefit decreases later in the course of the disease due to the progression of PD and to the appearance of stimulant-resistant medications and symptoms. Clinically-measured "low-level" motor function responds to STN-DBS, but cognitive and "high-level" motor functions related to VMC may not respond to STN-DBS. To examine whether the rate of change in global cognitive functioning during the initial 6 months after STN-DBS differed from the mean 6-month change that occurred between 6 and 36 months after surgery. The decline in global cognitive function was faster in the first 6 months after surgery, compared to a 6-month period between 6 and 36 months post-surgery. To specifically establish a detailed neuropsychological profile before and after STN-DBS and identify any pre-surgical cognitive profile that can predict cognitive outcomes after stimulation. Non-dementia patients with mild impairment in both general intellectual functions and list learning, may be at a greater risk of decline in other aspects of verbal memory after STN-DBS.  To determine the effects of STN-DBS on the comprehension of metaphor and linguistic abilities such as lexical and semantic abilities. STN-DBS had a significant beneficial effect on motor symptoms in PD, but this stimulation had no effect on metaphor comprehension or any other cognitive ability assessed in this study. To outline the nature of verbal fluency dysfunction. The STN-DBS group task performance was lower than that of healthy controls. In addition to affecting motor symptoms, surgery seems to influence the dynamics of cognitive procedures. To investigate and compare results of treatment with dopamine versus DBS in the ability of PD patients to acquire and maintain over the successive days their performance in visual working memory While STN-DBS patients demonstrate more accurate and faster responses in the ON stage than in the OFF stage, regardless of the day of testing, patients using dopamine replacement therapy had more accurate and faster ON response compared to OFF during the first day of learning and then maintained or even improved their performance on the second day after consolidation in both the OFF and ON stages.  Global cognitive functioning Most studies 9-21 evaluated global cognitive functioning with 3 different instruments and observed no significant change in subject performance. Only 3 articles [22][23][24] reported impairment in the overall cognitive functioning of their sample.

Language
Five articles 13,20,21,27,33 showed better or stable performance in language, production of words, 12 semantic and phonemic verbal fluency tasks, 19 and vocabulary subtest of the Wechsler Abbreviated Scale 24 postoperatively.
Visuoconstructive and visuospatial skills Two articles 16,19 reported decline in Visuoconstructive and visuospatial skills, while 3 articles 10,12,20 showed no difference pre and post-operatively.

Motor and sensory coordination
There was no decline in coordination. 9,13,26

DISCUSSION
This systematic review sought to investigate the cognitive functions most affected by STN-DBS according to studies published in the last 10 years.
Analysis of the results of all 27 articles revealed no consensus among studies on the effect of this surgery on patients. In most articles that evaluated global cognitive functioning, cognition either improved or did not worsen, a good finding since the technique does not target non-motor symptoms. However, STN-DBS can promote an improvement in cognition indirectly in that, once the subject has reduced or eliminated motor symptoms, their quality of life (QoL) improves, allowing them to return to previously discontinued tasks and habits. This behavioral change can yield both cognitive and mood benefits. To confirm this hypothesis would require studies comparing mood (anxiety, depression) before and after stimulation.
In general, this heterogeneity of results can be due to several factors, as discussed below. The aggravation of cognitive disorders can be strongly predicted by neuropsychological tests in the early stage of the disease, with or without timely medical treatment. On average, 25-50% of PD patients develop MCI or dementia or progress from MCI to dementia within 5 years of diagnosis. 36 Thus, the selection of instruments is of paramount importance and needs to be accompanied by certain precautions. There is no specific protocol defining the most appropriate instruments for this evaluation, but knowing which functions are influenced by PD makes choosing the tests easier. Establishing a protocol to be used by studies and research centers would render it easier to access, understand and compare results, leading to further investigation of the impaired aspects. 37 Any change indicated by the tests is subtle, as cognitive impairment detected in specialized tests is not commonly reported by patients, caregivers or health professionals. As stated above, QoL assessments in these patients show improvement, even when cognitive impairment is detected. With regard to memory impairment, for example, there are several associated factors, such as the subject's age, duration of illness, and even executive functioning. In the case of the articles, the recognition memory 9,13 and recall 18,19,27,28 were impaired and this is observed in the literature, indicating a possible evolution to dementia in PD. 36 EFs are an umbrella concept that cover several aspects and, consequently, feature as the most evaluated functions and with the most discrepant results. Commonly, these functions appear to be impaired earlier in the disease and are directly associated with daily activities, impacting patient QoL. 38 Verbal fluency worsened in many studies. [9][10][11][12][14][15][16][22][23][24]27,28,31,35 In fact, worsening on category fluency tasks is the most frequent cognitive sequela reported after STN-DBS. This is in accordance with recent evidence suggesting that the STN is a potent regulator of basal ganglia and thalamocortical limbic and associative circuits. Frontal lobe-related cognitive changes after DBS may be determined by the modulation of these distinct neural networks. 39 Impairment of visuospatial skills, in which motor involvement is the main criteria, even in the early stages of the disease, is expected in PD -at odds with the fact that only 5 articles evaluated this function. 10,12,16,19,20 One of the inclusion criteria was surgery targeting the STN, and this was one of the limitations found in the studies. STNs are considered to produce more cognitive side effects in patients than when electrodes are implanted in the globus pallidus. 39 Patient age ranged from 51 to 67 years at the time of surgery and the literature indicates a higher risk of cognitive decline associated with older age. The medication or stimulation parameters in study participants were not controlled, and there may be an influence of reductions in postoperative medication or differences in DBS parameters. On top of this, there are differences regarding follow-ups, making it difficult to understand and establish "specific milestones", with which improvement or worsening of effects over months/years can be predicted. Thus, while certain articles reported follow-up effects for 36, 22,24 84 22,26 or up to 132 19 months, others had data for 12, 24,26 6 13,15,20,24 and up to 3 17,22,31 months. This discrepancy makes a fair comparison and reliable analysis of the data unfeasible. Using the same battery of tests at such widely varying time intervals may give the impression of an improvement simply by the learning effect of a short-term reassessment and a marked worsening as the disease progresses naturally over a long-term reassessment. 37 There was an absence of reports on the subjective impact of daily cognitive decline associated with motor symptoms 28 and of preoperative follow-up on cognitive function. 33 There were no other evaluations of impairment to activities of daily living associated with the disease, which interferes with the subjective perspec-tive of patient abilities. These aspects are directly influenced when motor improvement occurs. Thus, from the recovery of skills, new perspectives emerge, which can have a positive repercussion on non-motor symptoms, such as cognition. The angle of the surgical trajectory and proximity of the STN-DBS electrodes greatly influences the outcomes seen after surgery, where these aspects may be related to changes in the cognitive and emotional functioning of patients. 12,33 Thus, the results are expected to vary from one another -as has been seen. This disparity is mainly due to variations in the characteristics of patients selected for surgery across different centers (age, 21,26 preoperative state 10,24 and comorbidity with other conditions such as psychiatric disorders 11,12 ), making conclusions difficult to compare and analyze.
Thus, it was not possible to establish a neuropsychological profile of PD patients with STN-DBS. This is cause for concern since patients with MCI in PD are more likely to progress to dementia as the disease progresses, and it is necessary to understand which cognitive functions become impaired in this disease after DBS implantation to avoid miscalculating normal with worsening evolution. Much of this can also be attributed to the lack of a specific PD assessment protocol. 37 The results of this review highlight the need to establish a neuropsychological profile of PD patients to understand and investigate the effects of implantation of STN-DBS on cognitive non-motor symptoms. Future studies intend to develop a neuropsychological battery and evaluate patients with PD and STN-DBS to discriminate the aspects affected in these subjects and understand which factors contribute to outcomes.