DBS is an increasingly offered advanced treatment for Parkinson’s disease (PD). Neuropsychological assessment is considered to be an important part of the screening for selection of candidates for this treatment. However, no standardised screening procedure currently exists. In this study, we examined the use of our standardised neuropsychological assessment for the evaluation of surgical candidates and to identify risk factors for subsequent decline in cognition and mood. A total of 40 patients were assessed before and after DBS. Evaluation of mood and case notes review was also undertaken. Before DBS, patients with PD demonstrated frequent impairments in intellectual functioning, memory, attention, and executive function, as well as high rates of mood disorder. Post-DBS, there was a general decline in verbal fluency only, and in one patient, we documented an immediate and irreversible global cognitive decline, which was associated with older age and more encompassing cognitive deficits at baseline. Case note review revealed that a high proportion of patients developed mood disorder, which was associated with higher levels of depression at baseline and greater reduction in levodopa medication. We conclude that our neuropsychological assessment is suitable for the screening of candidates and can identify baseline risk factors, which requires careful consideration before and after surgery.
Drug therapies for advanced Parkinson’s disease (PD) can be unsatisfactory, with unwanted side effects, and/or insufficient control of disabling motor symptoms. Thus, there has been resurgence in interest in surgical treatments, with deep brain stimulation (DBS) now increasingly offered as an option. DBS is the chronic, high-frequency electrical stimulation of most usually the subthalamic nucleus (STN) or internal segment of the globus pallidus (GPi) [
DBS has been shown to be relatively safe, with few negative events occurring during or following surgery, when performed on appropriate candidates [
Decline in cognitive functioning following DBS has been found to be more common in patients who are older, especially above 70 years [
Despite the widespread agreement on the importance of appropriate screening and careful selection of surgery candidates, to the best of our knowledge no standardised neuropsychological assessment procedure currently exists. The Consensus on DBS for PD [
Moreover, there is scant official guidance. The British Psychological Society [
In lieu of such guidance, several studies have relied upon brief cognitive screens only, such as the Mini-Mental State Examination (MMSE) [
Moreover, such brief screening tools do not permit scrutiny of the wider cognitive profile, important for confirming diagnosis. Many cases of “failed” DBS have been later found to have atypical Parkinsonian syndromes, known not to benefit from DBS [
In addition to changes in cognition, there are also a few reports of dramatic deterioration in mood and greatly increased apathy following DBS [
Thus, the aims of this study were to evaluate the use of our standardised neuropsychological protocol in the evaluation of patients undergoing DBS for PD in order to identify any contraindication for surgery and to be sensitive to changes following DBS.
A total of 40 patients (29 male, 11 female) who underwent DBS took part in this study. All patients had had a diagnosis of idiopathic PD for at least five years (according to Queen Square Brain Bank criteria), were younger than 70 years, and suffered from disabling motor complications despite optimal treatment. Each patient underwent multidisciplinary evaluation to decide on suitability for DBS. Formal levodopa challenge confirmed dopaminergic drug responsiveness. A structural MRI was obtained to exclude surgical contraindications, such as advanced brain atrophy, white matter changes, or any other abnormality contraindicating surgery. Detailed neuropsychological and neuropsychiatric assessments excluded patients with significant cognitive impairment and/or psychiatric comorbidities. Contraindications for STN DBS included the presence of clinically relevant speech difficulties and cognitive impairment. The final decision regarding suitability for DBS and appropriate target for each patient was taken during a joint meeting of patient, immediate family, neurologist(s), and neurosurgeon(s).
Motor status was evaluated using part III of the Unified Parkinson’s Disease Rating Scale (UPDRS-III). Prior to surgery, patients were assessed in the practically defined “off state” after overnight withdrawal of anti-Parkinsonian drugs and the “on state,” following a levodopa challenge using a suprathreshold dose of oral levodopa. After DBS, motor assessments were sequentially performed under the following conditions, in open fashion: off medication/on stimulation (with stimulation switched on after 12 h medication withdrawal) and on medication/on stimulation (1 h after the administration of a routine dose of levodopa while stimulation was reintroduced). All medications before and after surgery were recorded, noting any dopamine agonist treatment, and levodopa-equivalent dosage was calculated (
All patients underwent assessment of neuropsychological and mood functioning before and after surgery, under optimal conditions. Thus, preoperatively, this was in the on medication and postoperatively on stimulation/on medication. The postoperative assessment was performed a mean of 19.60 months after surgery (range = 1–54; SD = 11.56). This broad range reflected the early recall of one patient following concern about cognition immediately following DBS, as well as later routine follow-up of cognitively intact patients after surgery.
The most appropriate DBS target was chosen on clinical grounds based on patient motor phenotype, imaging, and preoperative cognitive assessment. Twenty-eight patients underwent bilateral STN DBS and twelve bilateral GPi DBS.
The tests included general screening and IQ measures, as well as tests of specific cognitive functions. This was to enable both the quantification of any intellectual deficit and the elucidation of specific cognitive profiles. Thus, the measures included tests of general cognitive functioning, memory, language, visuoperceptual ability, attention, executive functions, and speed of processing. The tests chosen were considered to have acceptable test validity and reliability, as described below. The assessments took around two hours to complete and were as follows: The MMSE was used as a screening test of global cognitive functioning [ Vocabulary, similarities, arithmetic, and digit span subtest scores from the Wechsler Adult Intelligence Scale-Third Edition (WAIS-III) [ The National Adult Reading Test-Revised (NART-R) [ Memory was assessed using the following: The Warrington Words and Faces Recognition Memory Tests (RMTs) [ The People and Shapes subtests from the Doors and People Test were used to assess verbal and visual recall memory (D&P) [ The Graded Naming Test (GNT) [ The Silhouettes subtest from the Visual Object and Space Perception Battery (VOSP) [ Elevator Counting and Elevator Counting with Distraction subtests from the Test of Everyday Attention (TEA) [ Executive functioning was assessed using the following: FAS and Category subtests from the Delis–Kaplan Executive Function System (DKEFS) [ The Stroop [ The Hayling and Brixton tests [ The Symbol Search and Digit Symbol Coding subtests from the WAIS-III [
All patients were screened for mood disorder using the Hospital Anxiety and Depression Scale (HADS) [
The case notes were reviewed by one clinical neuropsychologist (JAF) to identify any change in cognition, mood, or behaviour since DBS, as highlighted by the surgery team, neurologists, or nursing staff. Any mention of decline in memory, attention, perception, language, reasoning, mood, anxiety, depression, or motivation was recorded, along with number of months elapsed since surgery. As discussed before, any mention of a de novo impulse control disorder was recorded.
Scores for each of the neuropsychological assessments were compared with published normative data. For each measure, patients were judged to be impaired if scores were ≤2 SD. When multiple measures were used, performance was classified as impaired when ≤2 SD on at least one of the measures used.
Normality of distribution was assessed using the Kolmogorov–Smirnov test and, if significant, by examining the
The research was done in accordance with the Helsinki Declaration and the Institute of Neurology Joint Research Ethics Committee UCLH, NHS Trust Research and Development Directorate.
As shown in Table
Patient demographic characteristics.
STN ( |
GPi ( |
|
|
---|---|---|---|
Gender (male) | 17 | 8 | 0.72 |
Age (at first assessment, years) | 57.50 ± 7.32 | 61.33 ± 6.30 | 0.12 |
NART Predicted Full-Scale IQ | 111.57 ± 11.08 | 103.42 ± 15.43 | 0.07 |
Age at PD diagnosis (years) | 45.55 ± 7.80 | 48.60 ± 6.35 | 0.29 |
PD disease duration (years) | 18.77 ± 6.12 | 19.00 ± 4.55 | 0.92 |
History of impulse control disorder ( |
9, 28.1% | 1, 3.1% | 0.08 |
At baseline, there were no significant differences between the STN and GPi DBS patient groups in UPDRS-III scores off or on medication, nor in baseline levodopa-equivalent dosage (as shown in Tables
Patient motor characteristics before and after DBS.
STN | GPi | |||||
---|---|---|---|---|---|---|
Before DBS ( |
After DBS ( |
|
Before DBS ( |
After DBS ( |
|
|
UPDRS-III off medication | 48.68 ± 14.10 | 28.67 ± 9.99 | 0.00 | 50.73 ± 11.09 | 35.20 ± 15.32 | — |
UPDRS-III on medication | 17.29 ± 7.967 | 15.83 ± 7.20 | 0.49 | 24.64 ± 10.97 | 20.75 ± 11.56 | — |
Patient medication characteristics before and after DBS.
STN ( |
GPi ( |
|||||
---|---|---|---|---|---|---|
Before DBS | After DBS |
|
Before DBS | After DBS |
|
|
Levodopa-equivalent dosage (mg/d) | 1321.82 ± 638.68 | 863.73 ± 583.92 | 0.00 | 1263.40 ± 971.08 | 1205.10 ± 626.68 | 0.81 |
Dopamine agonist treatment ( |
15, 46.9% | 6, 18.8% | 0.01 | 7, 21.9% | 5, 15.6% | 0.16 |
There were also no significant differences between the STN and GPi DBS patients groups in proportion of patients receiving dopamine agonist treatment before or after DBS.
When baseline neuropsychological assessment scores were compared with published normative data, impairment was documented on at least one domain of cognitive function in 85% of all patients (STN:
Cognitive performance before DBS: proportion impaired in each domain.
Cognitive domain | STN ( |
GPi ( |
|
---|---|---|---|
Screen |
|
|
|
IQ | 12, 42.9% | 8, 66.7% | 0.30 |
Memory |
|
|
|
Language | 1, 3.7% | 3, 27.3% | 0.07 |
Perception | 2, 7.4% | 1, 8.3% | 1.00 |
Attention | 5, 18.5% | 4, 33.3% | 0.42 |
Executive function |
|
|
|
Speed |
|
|
|
Results are given as number and percentage. Chi-squared significant group comparisons are indicated in bold.
When investigated further, we found that the GPi DBS patients obtained lower baseline scores on tests of general intellectual functioning (VIQ:
As shown in Table
Cognitive performance before and after DBS: mean scores on each test.
Assessment | STN ( |
GPi ( |
||||
---|---|---|---|---|---|---|
Before DBS | After DBS |
|
Before DBS | After DBS |
|
|
MMSE (30) | 28.64 ± 1.41 | 28.64 ± 1.68 | 1.00a | 26.75 ± 3.14 | 25.75 ± 2.87 | 0.31a |
WAIS-VIQ | 111.21 ± 12.40 | 107.54 ± 15.93 | 0.05a | 92.75 ± 13.10 | 92.67 ± 10.71 | 0.97a |
Vocabulary (66) | 51.57 ± 9.61 | 49.96 ± 10.00 | 0.13a | 40.64 ± 15.02 | 37.27 ± 14.14 | 0.16a |
Similarities (33) |
|
|
0.02a | 19.36 ± 5.43 | 18.36 ± 5.12 | 0.43a |
Arithmetic (22) |
|
|
0.01a | 10.36 ± 2.94 | 9.64 ± 3.04 | 0.90a |
Digit span (30) |
|
|
0.04a | 14.08 ± 3.09 | 14.00 ± 3.16 | 0.41a |
WAIS-PIQ | 106.37 ± 15.17 | 104.04 ± 19.57 | 0.50a |
|
|
|
Picture Completion (25) | 17.96 ± 4.25 | 17.19 ± 4.86 | 0.33a | 12.92 ± 3.66 | 11.33 ± 3.53 | 0.07a |
Matrix Reasoning (26) | 16.35 ± 5.61 | 15.31 ± 5.90 | 0.24a | 11.10 ± 4.33 | 9.20 ± 4.21 | 0.09a |
RMT-W (50) | 46.81 ± 3.50 | 45.12 ± 5.35 | 0.10b | 39.20 ± 9.45 | 40.30 ± 8.68 | 0.16a |
RMT-F (50) | 41.88 ± 4.13 | 41.08 ± 5.68 | 0.54a | 33.60 ± 6.85 | 34.00 ± 7.92 | 0.80a |
D&P People delayed (12) | 7.21 ± 3.68 | 7.50 ± 4.30 | 0.59b | 6.40 ± 4.65 | 5.60 ± 3.95 | 0.57a |
D&P Shapes delayed (12) | 10.50 ± 3.28 | 10.25 ± 2.82 | 0.22b | 8.86 ± 3.63 | 7.43 ± 3.78 | 0.30a |
GNT (30) | 23.69 ± 3.42 | 23.69 ± 3.28 | 1.00a | 17.91 ± 8.11 | 19.45 ± 6.65 | 0.34a |
VOSP Silhouettes (30) | 22.81 ± 3.25 | 21.92 ± 3.91 | 0.13a | 20.82 ± 3.52 | 19.00 ± 5.88 | 0.41a |
DKEFS FAS (SS) |
|
|
|
|
|
|
DKEFS Category (SS) |
|
|
|
|
|
|
Stroop (112) | 91.81 ± 21.36 | 83.77 ± 22.94 | 0.06a | 63.00 ± 20.44 | 58.50 ± 23.45 | 0.12a |
Hayling (SS) | 5.68 ± 1.07 | 5.32 ± 1.52 | 0.28b | 4.60 ± 1.84 | 4.70 ± 1.83 | 0.89a |
Brixton (SS) | 4.91 ± 1.53 | 5.00 ± 2.28 | 0.83a | 2.33 ± 1.66 | 2.56 ± 2.07 | 0.72a |
TEA EC (7) | 6.67 ± 0.96 | 6.75 ± 0.44 | 0.85b | 6.50 ± 1.41 | 5.88 ± 1.55 | 0.26b |
TEA EC-Distraction (SS) | 9.91 ± 2.66 | 8.96 ± 2.92 | 0.15a | 7.13 ± 3.14 | 5.75 ± 1.91 | 0.17a |
WAIS-SS (SS) |
|
|
|
7.89 ± 3.33 | 6.11 ± 2.42 | 0.86a |
WAIS-DSC (SS) | 8.20 ± 2.52 | 7.48 ± 2.65 | 0.22 | 4.89 ± 2.67 | 4.67 ± 1.87 | 0.86a |
Results are given as mean ± SD (apaired
Case note review revealed mention of decline in cognitive function in 15% (
Pearson correlational analysis revealed no significant baseline cognitive, mood, or motor correlates of decline in phonemic fluency after either STN or GPi DBS. Greater decline in category verbal fluency following STN DBS was associated with higher levels of apathy (
There were also no significant baseline correlates of decline in Symbol Search after STN DBS. However, greater decline in PIQ following GPi DBS was associated with slower baseline performance on the Digit Symbol Coding subtest. There were no other significant predictors of decline following DBS.
In order to identify baseline predictors of the subsequent global and irreversible cognitive decline following STN DBS noted in the one patient, Crawford and Howell [
When considering the remaining patients who demonstrated cognitive decline at least a year after surgery (as identified in the case note review), no significant difference in demographics or cognitive performance at baseline was identified.
Baseline mood assessment revealed high rates of anxiety disorder (
Mood scores before and after DBS.
Assessment | STN ( |
GPi ( |
||||
---|---|---|---|---|---|---|
Before DBS | After DBS |
|
Before DBS | After DBS |
|
|
HADS anxiety (21) | 7.50 ± 3.23 | 6.27 ± 4.85 | 0.19b | 8.55 ± 3.30 | 9.00 ± 4.12 | 0.69a |
HADS depression (21) | 6.15 ± 4.42 | 6.00 ± 4.04 | 0.86a | 7.00 ± 3.85 | 7.73 ± 4.63 | 0.67a |
Apathy (54) | 10.75 ± 6.02 | 13.96 ± 11.16 | 0.15b | 14.57 ± 6.71 | 20.86 ± 11.11 | 0.67a |
Results are given as mean ± SD (apaired
Incidence of case note indication of cognitive impairment or mood disorder, as a function of time, is depicted in Figure
Cumulative cases of cognitive impairment and mood disorder as a function of time following DBS.
One patient also developed de novo impulse control disorder, namely, hypersexuality, after GPi DBS.
Patients who had subsequent mood disorder were found to have significantly higher baseline levels of depression (
Mean levodopa-equivalent dosage of patients before and after DBS in patients split according to subsequent onset of mood disorder.
Logistic regression confirmed these as significant predictors of subsequent mood disorder (
The patient who developed impulse control disorder following GPi DBS did not experience a reduction in levodopa-equivalent dosage but rather an increase, with ongoing dopamine agonist treatment.
Neuropsychological assessment is considered to be an important part of the screening for selection of candidates for DBS. However, to the best of our knowledge, no standardised assessment procedure currently exists, with many studies relying upon brief screening tools only. Neuropsychological screening should comprise tests with sufficient reliability and validity, which are sensitive to cognitive impairment and dementia in PD, able to disambiguate between PD and other disorders, including atypical Parkinsonian syndromes, and be sensitive to the changes in cognitive and mood functioning associated with DBS.
In this study, we examined the use of our standardised neuropsychological assessment in a sample of patients undergoing DBS for PD. Our assessment tested a wide range of neuropsychological domains, including general intellectual functioning, verbal and visual recognition and recall memory, language, visuoperceptual functioning, attention, verbal fluency, executive functioning, and speed of processing. The tests were all standardised, with adequate psychometric properties, easy to administer, and suitable for routine clinical services.
Our neuropsychological assessment was sensitive to the cognitive impairment found in PD. At baseline, we documented frequent impairments in intellectual functioning, memory, attention, and executive function, with more frequent impairments, as expected, in the GPi group. Indeed, only six of all DBS patients (15%) did not demonstrate impairment in at least one cognitive domain. Despite this only one patient demonstrated immediate and irreversible cognitive decline following DBS. This highlights the limitation of using the presence of any baseline cognitive impairment as an exclusionary criterion for DBS. As low test scores may reflect a number of cognitive and noncognitive variables, such as high levels of fatigue, low scores on any one test should not be used to preclude surgery.
Our neuropsychological assessment was also sensitive to the cognitive impairments that warrant caution before proceeding with DBS. In the patient who demonstrated immediate and irreversible global cognitive decline, single-case statistics revealed that this patient was significantly older than the mean age of those who remained stable and had greater deficits in language and visuoperceptual processing at baseline. Of course, this is a single case, and therefore, these results may not be generalizable, but this finding supports earlier reports that decline in cognitive functioning following DBS is more common in patients who are older [
Although previous guidance on patient selection has tended to focus on memory impairment as a core contraindication for surgery [
Previous studies describing negative cognitive outcomes following DBS may have failed to identify such risk factors because of insufficient scrutiny of baseline cognitive performance. Previous reports of immediate and global decline following DBS have often stated that such deterioration has occurred despite satisfactory performance on neuropsychological testing at baseline [
In keeping with this, our patient who demonstrated immediate and permanent cognitive decline performed flawlessly on the MMSE and performed poorly on only two out of four tests of executive function but demonstrated unexpected impairments, most clearly in language and visual perception. This underlines the importance of a broad neuropsychological assessment, interrogating a wide range of cognitive domains, to reveal the full cognitive profile.
Our neuropsychological assessment was also sensitive to the changes in cognitive functioning associated with DBS. Pre- and after DBS assessments revealed that alongside improvements in the motor status and medication load are noted in the STN group at least, and our assessment detected significant declines in verbal fluency in both groups following DBS. This confirms the mild changes frequently noted in this cognitive function following DBS [
Although the exact cause of verbal fluency decline remains unclear, it has been linked with reductions in self-generation [
Irrespective of the underlying mechanism, deterioration in verbal fluency can deleteriously affect activities of daily living and quality of life [
In addition to this finding of reduced verbal fluency, our assessment detected declines in other aspects of cognitive functioning. Specifically, STN patients demonstrated significant slowing on the Symbol Search test and near-significant slowing on the Stroop and reduction in VIQ. GPi patients demonstrated significant reductions in PIQ. These findings confirm a slowing in the STN patients at least. In the absence of any other focal deficits, the heterogeneous reductions in performance on the WAIS (in both DBS groups) may also reflect the composite nature of this measure and the effortful, sustained, and speeded aspects of attentional functioning that it requires. Such reductions in speed of processing after DBS have rarely been discussed as most studies investigating cognitive changes have failed to include any measure of processing speed [
When considering the patients who went on to report cognitive decline at least a year after surgery (as identified by case note review), there were no significant predictors at baseline. This may suggest that the observed decline reflects the normal progression of the disease, rather than any preexisting vulnerability in the cognitive profile. It is important to recognise that the case note review was limited to qualitative and subjective comments only, precluding comment on the severity of any cognitive decline. However, the current findings do support previous studies which suggest that the risk of developing dementia following DBS is equivalent to that in medically treated patients [
Our study indicated no significant changes in mood or apathy, as measured by questionnaires, following DBS. However, case note review revealed a very high incidence of depression, anxiety, and/or apathy after surgery. These contrasting findings may be explained by the fact that assessment of mood relied upon self-reported symptoms of depression, anxiety, or depression, whereas case note review simply indicated clinicians’ observations. Discrepancy between self- and proxy-ratings of mood in Parkinson’s disease has been reported previously [
Mood disorder emerging after DBS has been largely attributed to reduction in dopaminergic medications [
Furthermore, the chance of developing mood disorder, as identified in the case notes, was even higher in those who endorsed clinically significant levels of depression at baseline. This may suggest that those who have a preexisting vulnerability in mood are at high risk of developing profound mood disorder following DBS. Of course, the high incidence of mood disorder as noted in the case notes may simply reflect clinicians’ recognition of (stable) low mood. However, its timing of onset and high incidence is consistent with several other studies [
High levels of postoperative apathy or mood disorder can negate any improvement in quality of life [
One of our patients developed de novo impulse control disorder following GPi DBS. The onset of hypersexuality occurred in the context of increased levodopa dosages following surgery, with ongoing use of dopamine agonists. Our findings were of course limited to clinician ratings only and may have missed other cases. Future research should further investigate the incidence of impulse control disorder following DBS by using a semistructured interview, such as the QUIP [
Following our findings, we propose an abbreviated version of our neuropsychological protocol, suitable for routine clinical use. We recommend that this protocol includes our measures of current and premorbid intellectual functioning (prorated version of the WAIS-III, NART-R) to gauge overall level of intellectual decline; memory recognition and recall (RMT Words and Faces and D&P People and Shapes) to ensure cognitive profile is not amnestic and thus atypical for PD; language and visuoperceptual function (GNT and VOSP Silhouettes) to detect the identified red flags for DBS; verbal fluency (DKEFS FAS and Category) and another measure of executive function (Stroop) to determine severity of executive dysfunction; speed of processing (Digit Symbol Coding and Symbol Search); and measures of mood and behavioural functioning, targeting depression, apathy (HADS and AES), and impulse control disorder (using a measure such as the QUIP). Of course, analysis of neuropsychological performance should consider any relevant cultural or linguistic factors, and it may be appropriate to replace some of the present neuropsychological assessments with suitable substitutions for specific populations.
This study has presented a standardised neuropsychological assessment procedure suitable for the selection of appropriate candidates with PD for DBS and identified clear baseline risk factors for subsequent decline in cognitive functioning and mood.
The authors declare that there are no conflicts of interest regarding the publication of this article.
This work was undertaken at UCLH/UCL, which received a proportion of funding from the Department of Health’s National Institute for Health Research Biomedical Research Centre’s funding scheme. The Unit of Functional Neurosurgery, UCL Institute of Neurology, is supported by the Sainsbury Monument Trust and Parkinson’s Appeal for Deep Brain Stimulation.