Efforts of subthalamic nucleus deep brain stimulation on cognitive spectrum: From explicit to implicit changes in the patients with Parkinson's disease for 1 year

Abstract Objective To evaluate the cognitive function of Chinese patients with Parkinson's disease PD postsubthalamic nucleus deep brain stimulation (STN‐DBS). Methods Cognitive function was assessed by neuropsychological methods in PD patients. Twenty matched healthy persons served as normal controls. t test, analysis of variance, and chi‐square analysis were used to compare the difference among the groups. Reliable change index was utilized to analyze the changes in cognition from the individual level. Results (a) Improvement in motor function was significantly better after STN‐DBS (P < .01). (b) Notably, the increase error rates of implicit SRTT (serial reaction time task) was significantly higher after STN‐DBS as compared with the conservative therapy group (P = .03). (c) The decline of verbal fluency (explicit) was also significantly higher after STN‐DBS than that in the medication therapy group (P = .03). (d) In the explicit clock‐drawing test, scores had significantly improved after STN‐DBS (P = .04). Conclusions STN‐DBS as a neuromodulatory tool in the Chinese PD population not only improves motor symptoms but also cognitive function to a certain extent, such as the decline of executive function and verbal fluency. The explicit cognitive decline was significantly quicker than that in patients on medication therapy. The improvement of visiospatial function was also noted. Implicit memory impairment during the 1‐year follow‐up period was not observed.


| INTRODUC TI ON
Parkinson's disease (PD) refers to the disruption of the motor loop between the basal ganglia and motor cortex induced by the loss of dopaminergic neurons in the substantia nigra, resulting in resting tremor, muscle tension increase, slow motion, and postural disturbance. 1 PD has become a common neurodegenerative disease in middle-aged and elderly people. Besides motor symptoms, PD patients also show nonmotor symptoms, such as cognitive impairment, autonomic dysregulation, and sleep disorders, which seriously affect the quality of their life. 2 Currently, deep brain stimulation (DBS) is gradually being used on a large scale, owing to the significant success it has shown in improving motor symptoms, quality of life, and reducing daily drug dependency. 3,4 Effects of STN-DBS on PD patient's cognitive function are still not understood. It is believed that STN-DBS significantly alleviates motor symptoms through cortex-basal ganglia-thalamus-cortex loop modulation. 5 Memory is one of the key components of cognition. Previous studies have shown that STN-DBS could interfere working memory. 11,12 It has been reported, instances of some patients that could not recognize facial stimuli 13 or discern vocal emotion. 14 We supposed STN-DBS might impact patients' implicit memory and cause some slight cognitive change. This study attempts to explore the effects of STN-DBS on cognitive function and its potential mechanism by means of a specific neuropsychological tool.

| SUBJEC TS AND ME THODS
The ethics committee of Shanghai Jiao Tong University School of Medicine approved the study (for project number 81071065#). The authors received written informed consent from every participant.

| Patients and groups
Forty-three PD patients were selected from the Shanghai JiaoTong University School of Medicine in China from August 2013 to August 2015. All subjects were in accordance with the diagnostic criteria for idiopathic Parkinson's disease of United Kingdom Parkinson's Disease Society Brain Bank. 15 None of the patients met the following conditions: (a) invalid dopaminergic therapy; (b) severe psychiatric symptoms, such as schizophrenia symptoms, apathy, suicidal tendencies, and severe depression; (c) history of stroke, brain injury, and intracranial infection; (d) severe heart, liver, lung, kidney, and blood system, endocrine system diseases; (e) history of drug abuse. Three cases were dropped out due to good motor symptoms control and patients delayed follow-up time point more than 3 months. The dropout rate was less than 10%. Twenty PD patients received bilateral STN-DBS in the Department of Neurosurgery as the STN-DBS group (DBS group). And twenty PD patients without surgery were matched with STN-DBS group, received the optimal medication therapy as the medication therapy group (MED group).Twenty ( All the patients were evaluated for motor symptoms and cognition impairment at the beginning of the study and after 1-year follow-up.
As a control, 20 healthy candidates were also evaluated for cognition impairment. This control group was then compared with DBS and MED group at baseline. The declination of cognition between DBS group and MED group was compared, read the flowcharts in Figure 1.

| Assessment of cognition
Tests were administered by the neuropsychologist (Dr Wang Xiaoping), included Mini Mental State Examination (MMSE), 18 Wechsler Memory Scale, 19 Verbal fluency, Clock-drawing, Degraded picture naming, and serial reaction time task(SRTT). 20 The assessment of cognition was launched at "On time" on the day after motor symptoms assessment. The SRTT software version can count response time (T avg ), error rates, and minimum reaction time (T min ).
Attention is quantified by minimum reaction time (T min ).

| Leads localization
The Medtronic S7 Neuro Navigation System (Medtronic Navigation) was used to merge preoperative 3.0T-MRI images with postoperative CT images. The CT scan images were calibrated at 1 mm thickness and without pneumocranium. The Lead-DBS software 21 was used to show the electrodes and active contacts the same as our previous work done. 22

| Statistical analysis
Comparison of the overall level-Data were analyzed with SPSS 19.0 software. Measurement data were expressed as mean ± SD or minmax. A value of α = .05 was considered statistically significant difference. Self-control group was evaluated with paired t tests. Group comparison evaluated with student t test. Data that not obey normal distribution were analyzed with Mann-Whitney U rank sum test. Count data were compared by using chi-square test.
Comparison of the individual level-Some individual cognitive differences may be masked after averaging. Several reliable change indices (RCIs) exist to evaluate statistically significant individual change with repeated neuropsychological assessment. We present reliable change indices (RCIs) to compare statistically reliable cognitive changes between the groups for each neuropsychological measure over time. 23 According to the reliable change indices (RCIs), is the patient's baseline score, X2 is the patient's 1-year score, and S diff is the standard error of the difference between the test scores. RCI > 1.96 represents remarkable changes before and after test.

| Two patients' groups
STN-DBS group and MED group were both including 20 patients (10 female and 10 male). Two groups had no significant difference at gender, age, disease duration, H -Y grade, UPDRS sore, and LED at baseline (see Table 1).

| Assessment of motor symptoms
No significant difference in UPDRS scores was detected between the DBS group and MED group at baseline. As displayed in Table 1, motor function were significantly improved 49.7% in the DBS group postoperation (MED off IPG on) compared with that preoperation F I G U R E 1 Flowcharts of the study designed (MED off) (P < .001), postoperative required drug dose significantly decreased 48.2% (P < .001). Two patients totally stop taking medication at the optimal DBS stimulation.

| Memory
Memory quotient measured by Wechsler Memory Scale reflects explicit memory. Student-Newman-Keuls pairwise comparison test results showed that memory quotient was significantly lower in the DBS and MED groups than in the control group (F = 8.62, P < .01; Figure 3A).
There was no significant difference in long-term memory and instantaneous memory obtained by the superposition of the scale of the same type of subtest among DBS, MED and control groups (F = 2.13, P > .05；F = 1.54; P > .05). Short-term memory was significantly decreased in the DBS and MED groups compared with the control group (F = 4.56, P < .05; Figure 3B).There was no significant difference in memory quotient and each subtest before and after surgery in the DBS group (P > .05). There was no significant difference in memory quotient and each subtest before and after 1-year follow-up in the MED group (P > .05). And there was no significant change in the range of memory quotient of the two groups. Changes in memory quotient were considered as an observed value in the DBS and MED groups before and after 1-year follow-up. No significant difference in the change was detected between the two groups (F = 0.07, P = .800; Table 2). The SRTT was a task of procedural learning. Analysis of variance showed significant differences in T avg among DBS group, MED group and control group (F = 4.90, P < .05). Student-Newman-Keuls pairwise comparison test results showed that T avg was significantly higher in the DBS and MED groups than the control group. No   Table 2).

MED group (1-y follow-up)
Memory quotient (min-max) Note: Change = postoperation-preoperation or after 1-y follow-up-before 1-y follow-up. In the clock-drawing test, Mann-Whitney U rank sum test was used to compare the difference among DBS, MED, and control groups. Paired rank sum test was utilized to compare the difference before and after surgery or 1-y follow-up.
*P < .01, vs control group; # P < .05, vs preoperation; trend of error rate was significantly higher in the DBS group after surgery compared with the MED group (F = 6.16, P = .03; Table 2).

| Speech function
There was a significant difference in verbal fluency among the control group, DBS group (preoperative) and MED group (before 1-year follow-up) (P < .01). This could have been as a declined language function in PD patients ( Figure 3F). Verbal fluency decreased after DBS (t = 2.48, P < .05). No significant difference in verbal fluency was determined between 1-year follow-up and baseline data in the MED group (t=−1.06, P = .302). The decreased trend was significantly higher in the DBS group than in the MED group (F = 6.27, P = .03; Table 2).

| Visuospatial ability
Clock-drawing test showed a significantly lower test score in the DBS and MED groups at baseline compared to the control group (DBS group: z = −2.76, P < .01; MED group: z = −2.55, P < .01; Figure 3G).
The score was significantly increased after surgery in the DBS group (z = −2.07, P < .05). Moreover, no significant difference was detected between DBS group postoperatively and control group, suggesting that postoperative clock-drawing test results were close to normal levels. Clock-drawing test score was decreased after 1-year follow-up in the MED group, but the difference was not statistically significant (z = −1.31, P = . 19). Variations were compared between the two groups, multivariate analysis of variance of levodopa equivalent dose variation demonstrated significant difference (F = 5.71, P = .04; Table 2).

| Attention
T min was significantly increased in the DBS and MED groups at baseline compared with the control group (t = 2.41, P < .05; Figure 3H).
T min had a significant decrease after BDS (t = 2.83, P < .05). T min also exhibited a decreased trend in the MED group after 1-year follow-up, but no significant difference (t = 1.18, P = . 26). No significant difference in variations was found between the two groups (F = 2.83, P = .12) ( Table 2).

| Comparison of reliable change index from the individual level
Each test values were calculated in the DBS and MED groups based on reliable change index. The declined, remained stable, and improved percentages of each test are listed in Table 3. Significant differences on verbal fluency and error rate were seen between the DBS and MED groups (x 2 = 7.03, P = .03; x 2 = 6.72, P = .04).

| D ISCUSS I ON
Several and nonexclusive mechanisms have been proposed to underlie the beneficial effects of DBS, such as electrical and neurochemical effects of stimulation, modulation of oscillatory activity, synaptic plasticity, and, potentially, neuroprotection and neurogenesis. 24 It is generally believed that cognitive function is closely associated with the prefrontal lobe-basal ganglia-thalamic circuit. 25  Our results demonstrated that explicit memory was really impaired in PD patients, mainly in short-term memory impairment. Abbreviations: DBS, deep brain stimulation; MED, medication therapy.
With the in-depth study of memory, a theory of implicit memory is presented which describes it as a form of general plasticity within processing networks that adaptively improve function via experience. 27 The line graph of normal controls in serial reaction time task test showed the priming effect of implicit memory on later learning.
Compared with the health, implicit memory was impaired in PD patients, but STN-DBS did not evidently deteriorate implicit memory.
Different types of memory have different neural structures and loop. The medial temporal lobe and the hippocampus are involved in explicit memory, while basal ganglia is involved in implicit memory. 28 Explicit and implicit memory systems act the memory function through different cognitive mechanisms, but they influence each other. 29 The basal ganglia are likely to be the linked between the two memory systems. 30 However, the effects of basal ganglia on the memory system and its relationship with explicit and implicit memories deserve further investigations. It is emphasized that the healthy control group also was scored by the implicit and explicit cognition tests, so to observe and adjust the age-dependent tendency of natural cognitive outcomes for a year. In conclusion, STN-DBS obviously improved motor symptoms, and simultaneously altered cognitive function, such as the decreased executive function and verbal fluency. The decline is more obvious than the natural cognitive decline in patients with PD. Visuospatial function was improved, and attention was possibly improved.
Explicit memory and overall cognitive function did not noticeably alter. Implicit memory was retained, and STN-DBS did not affect implicit memory in a short period. Long-term effects need to be made clear by longer follow-up and more in-depth study.

ACK N OWLED G M ENTS
The authors express many thanks to Drs Lassonde M, Hu XW, and Wu B. WXP is funded by National Natural Science Foundation of China (81071065# and 81671103#), and Key special topics of National Ministry of Science and Technology (2016YFC0105900).

CO N FLI C T O F I NTE R E S T
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

E TH I C A L A PPROVA L
This study was performed in accordance with the guidelines of the Declaration of Helsinki (1964) and was approved by the medical ethics committee JiaoTong University School of Medicine. All subjects provided written informed consent prior to their inclusion in the study.