Correlates of Deep Brain Stimulation Consensus Conference Decision to Treat Primary Dystonia

Abstract

Background:

Deep brain stimulation (DBS) is an effective treatment for motor disturbance in people with primary dystonia (PWD). Numerous factors are considered by an interdisciplinary consensus conference before deciding candidacy for DBS surgery (e.g., demographic, medical, cognitive, and behavioral factors). However, little is known about which of these factors are associated with PWD DBS surgery consensus conference decisions.

Objective:

Our goal was to examine whether pre-operative demographic, medical, and cognitive/behavioral variables are associated DBS consensus conference decisions in patients with dystonia.

Methods:

Thirty-two PWD completed comprehensive presurgery workup included neurological and neuropsychological exams, and neuroimaging in consideration for DBS surgery. An interdisciplinary conference committee either recommended or did not recommend DBS surgery based upon these data. Demographic and medical data (e.g., dystonia disease characteristics, medical comorbidities, medications) were also collected. We also examined impact from cardiovascular disease factors, using a Revised Cardiac Risk Index. PWD were grouped based on DBS conference decision (eligible: n=21, ineligible: n=11) and compared across demographic, medical, and cognitive/behavioral variables.

Results:

Across clinical variables, PWD who were deemed ineligible for DBS surgery had a higher Revised Cardiac Risk Index. PWD who were classified as ineligible displayed lower global cognitive functioning, working memory, phonemic fluency, memory retrieval, and cognitive flexibility.

Conclusions:

Consensus decision making regarding DBS surgery eligibility involves a multifactorial process. We found that deficits in executive functioning were associated with the DBS consensus committee decision. We also observed elevated cardiac risk among these individuals, likely reflecting the relation between vascular health and cognition. Implications, and clinical and scientific applications of these findings are discussed.

1. Introduction

Dystonia is a movement disorder involving sustained muscle contractions frequently causing twisting, repetitive movements, and/or abnormal postures [1]. Although prevalent in only 16 per 100,000 [2], dystonia is the third most common movement disorder and can be either primary or secondary [3]. Primary dystonia refers to dystonia in the absence of another neurological disorder or other neurological symptoms, with the exception of tremor and myoclonus, and often has a genetic basis. In contrast, secondary dystonia is caused by another disease state, brain injury, or exposure to certain medications or environmental toxins [3, 4]. Primary dystonia patients demonstrate difficulties with executive functioning, particularly cognitive flexibility [5-7].

Dystonia may result from aberrant inhibition or plasticity of motor programs in the basal ganglia, as well as the cerebellum [8]. Deep brain stimulation (DBS) of the globus pallidus internus (GPi) is an effective treatment with minimal adverse cognitive and psychiatric side-effects in PWD [9-12]. Whether DBS surgery is appropriate for a particular patient is complex and depends upon numerous variables including demographics (e.g., age), clinical characteristics (e.g., disease duration), cognitive and psychiatric factors [13], medical comorbidities (e.g., heart disease), and dystonia phenomenology [14]. In practice, decisions about which patients undergo DBS surgery are made at multidisciplinary consensus conference meetings (including neurology, neuropsychology, nursing, and neurosurgery) [14], and factors that could negatively affect surgical outcomes and the recovery process are discussed before a “yes” or “no” consensus is reached for each patient.

While there is a paucity of research investigating the factors associated with DBS consensus committee decisions in PWD, several papers have put forth consensus-based guidelines for determining surgical candidacy in persons with Parkinson’s disease (PD). Common reasons for exclusion based on Guideline statements include sociodemographic (older age [>70], lack of caregiver/family support, nursing home residence), medical/neurologic (dysarthria, repeated falls, medical comorbidity/surgical risk factors), cognitive (mild, moderate, or severe dementia), neuropsychiatric (impulse control disorder, uncontrolled depression, troublesome hallucinations, psychosis), and patient attitude (fears of side-effects, unrealistic goals/expectations) factors are all considered contraindications [15, 16]. Consensus conference data from DBS clinics treating Parkinson’s disease (PD) indicate that the leading reason for ineligibility include significant cognitive decline and cognitive disorders, followed by psychiatric factors and behavioral dysfunction, early disease duration or insufficient motor disability severity, significant medical comorbidities, and abnormal brain imaging [13, 17, 18].

While the presence of dementia/major neurocognitive disorder is widely regarded as a definite contraindication for DBS surgical candidacy, there is no clear consensus among experts on how more mild forms of cognitive impairment (e.g., mild neurocognitive disorders, ‘moderate’ graded cognitive impairment) are considered in the decision-making process [13, 19]. In a different vein, age as a contraindication has been debated in the DBS PD literature. As older age is associated with more medical comorbidities/complexity, higher rates of cognitive decline, and greater disease severity/occurrence of levodopa-resistant symptoms in PD, there is general consensus of higher risk of surgical complications/adverse outcomes in these patients [20].

The literature exploring DBS decision-making issues and establishing consensus regarding candidacy and contraindications is sparse in primary dystonia. Some studies have investigated which factors are associated with positive outcomes following DBS surgery in PWD [12, 21-23]. These studies indicate that positive DYT1 gene status, larger right pallidal volume, shorter disease duration (<15 years), younger age at surgery (<21 years), lower disease severity, and absence of skeletal deformities are all associated with better motor outcomes following DBS surgery in primary dystonia patients. Additionally, primary dystonia appears to be more responsive to DBS intervention targeting the GPi than secondary dystonia [24]. Based on available literature (despite lack of clear consensus guideline-based set of criteria), good candidates for DBS surgery among PWD have clear diagnostic history of isolated dystonia with predominate symptoms in limbs, neck, and trunk [25]. This includes persons with idiopathic segmental and generalized, inherited (DYT1), and acquired (tardive dystonia) forms of dystonia [25]. In contrast with the PD DBS literature, a recent review indicated that there is no established consensus or set of criteria for excluding PWD based on cognitive impairment or mental health problems [19]. Rather, it is recommended that these risk factors are considered on a case-by-case basis [19]. In addition, , a recent review suggested lower threshold for considering PWD for DBS surgery, especially if younger in age [26], and accompanied by functional and/or social impairment [25].

Research investigating which factors influence DBS consensus conference decisions have the potential to impact clinical practice. To our knowledge, no research has examined which factors are associated with a DBS consensus conference decision regarding PWD. To address this knowledge gap, we examined whether demographic, medical, and cognitive/behavioral factors and clinical characteristics are associated with consensus conference decision on DBS eligibility in PWD. In line with the rich body of literature examining cognition in DBS for PD, we broadly hypothesized that PWD designated eligible for surgery would show higher cognitive functioning compared to those deemed ineligible.

2. Materials & Methods

2.1. Participants

Participants were 32 PWD who were candidates for DBS surgery at the University of Alabama at Birmingham (UAB). Evaluations occurred from 2004-2016, and the study consisted of a retrospective chart review. The UAB Institutional Review Board approved all study procedures. All participants provided informed consent for their data to be used in future research at the time they were initially evaluated.

We only evaluated patients with primary dystonia and excluded secondary dystonia (e.g., dystonia due to head injury, medication use, or other neurological condition) in this study. All participants were aged 19 and older. Of the 32 PWD evaluated for DBS surgery, 21 (65.6%) were deemed appropriate for surgery by consensus conference and 11 were considered by consensus conference to be poor surgical candidates. We obtained comprehensive neuropsychological evaluations for all 32 PWD who are being evaluated for potential DBS surgery. In addition, all PWD received comprehensive neurological evaluations and confirmed primary dystonia diagnosis prior to the neuropsychological evaluation and consensus conference.

2.2. Consensus Conference

The monthly DBS consensus conference at UAB is composed of neurologists, neuropsychologists, nurse practitioners, and neurosurgeons. At consensus conference meetings, each potential DBS surgical candidate is discussed and the following three main components are reviewed: (1) neurologic work up, including patient information and disease semiology, (2) neuroimaging data, and (3) cognitive and psychiatric findings.

2.3. Demographic and Clinical Characteristic Factors

We examined the following demographic factors in relation to DBS consensus conference decision: age, years of education, gender, race, relationship status, and handedness. In addition, the following clinical characteristics were the subject of DBS consensus conference decision analysis: age of primary dystonia diagnosis, primary dystonia disease duration, and primary dystonia subtype. We also considered the Reading subtest of the Wide Range Achievement Test-3 (WRAT-3) as an ancillary measure of baseline intellectual functioning; standard scores were used with M=100 and SD=15 in the normative sample [27].

2.4. Medical and Pharmacologic Risk Factors

We examined current and past medical therapies, including neurological medications as well as psychotropic and cardiac medications. In addition, to further characterize our examination of medical factors, we computed the perioperative cardiac risk factors derived from the Revised Cardiac Risk Index (RCRI) criteria [28]. The RCRI was not used in the clinical decision-making process, but rather as a tool to further facilitate our understanding of the patient’s medical status for purposes of this analysis. This included a history of ischemic heart disease, congestive heart failure, cerebrovascular disease, and diabetes mellitus managed by insulin. Preoperative serum creatinine was not available and was therefore not included in risk factor calculations. The resulting index provided a relative risk approximation in terms of the likelihood of perioperative cardiac complications. By virtue of its computational elements, the RCRI also broadly represents an index of vascular risk factors [28].

2.5. Cognitive Measures

All PWD completed a neuropsychological test battery comprised of the following measures (corrected for demographic factors according to the respective test normative samples).

2.5.1. Global Cognitive Functioning

The Dementia Rating Scale-2 (DRS-2) was used as a measure of global cognitive functioning [29]. The DRS-2 is a 144-point measure that includes five domain subscales: Attention, Conceptualization, Memory, Construction, and Initiation/Perseveration. The DRS-2 Total Score is the sum of the five domain subscales.

2.5.2. Attention

The Digit Span subtest from the Wechsler Adult Intelligence Scale-Third Edition [30] measures auditory attention and working memory. Examinees are asked to repeat strings of digits both forward and backward. The Attention subscale of the DRS-2 measures working memory and attention abilities while completing verbal and visual tasks of varied complexity [29].

2.5.3. Verbal Fluency

We evaluated phonemic fluency with two tasks. In one, participants were asked to say as many words as possible in one minute apiece for the letters “C,” “F,” and “L,” or alternatively the letters “F,” “A,” “S” [31]. To evaluate semantic verbal fluency, participants were asked to generate as many animals, beginning with any letter, as possible in one minute [31].

2.5.4. Visuospatial

The DRS-2 Construction subscale [29] consists of copying simple designs and the examinee signing his or her name.

2.5.5. Memory

The California Verbal Learning Test-Second Edition (CVLT-II) [32] is a measure of verbal learning, memory, and recognition. The present study used the following CVLT-II components: Trials 1-5 (i.e., total learning across 5 presentation trials), Trial B (i.e., interference task of immediate recall for a distractor word list), Short-Delay Free Recall (SDFR; spontaneous recall of the target word list after the interference task), Long-Delay Free Recall (LDFR; spontaneous recall of the target word list after 20-minute delay), and Total Discriminability (i.e., overall ability to accurately recognize target words without making false positive errors). Of note, Trial B and SDFR examine proactive and retroactive interference effects, respectively.

The 10/36 Spatial Recall Test is a measure of visuospatial memory [33, 34]. The test materials consist of a 6 x 6 check board and 10 checkers. This task produces two scores: Immediate Recall (IR) and Delayed Recall (DR). The former represents total spatial recall across three trials, where the examinee is shown the board for 10 seconds on each trial. The latter score represents spatial recall after a 15-minute delay.

The DRS-2 Memory subscale measures orientation and knowledge of basic information. It also measures recall of verbal material after a brief delay and recognition of verbal and visual information [29].

2.5.6. Conceptualization

The DRS-2 Conceptualization subscale evaluates abstract reasoning with recognizing similarities and differences between different objects when presented verbally or visually [29].

2.5.7. Executive Functioning

The WCST is a measure of executive functioning (i.e., concept formation, novel problem-solving, cognitive set-shifting ability) [35]. Examinees face an ambiguous situation where they are given vague directions to match each card in a deck to a key card according to a category that conspicuously switches after a standardized number of successful sorts. The present study used the 64-card version. Performance was examined based on Categories (i.e., total number of categories correct), Total Errors, and Perseverative Responses [35].

The DRS-2 Initiation/Perseveration subscale is comprised of a semantic verbal fluency task, articulation of vowel and consonant patterns, double alternating motor movements, and graphomotor abilities [29].

2.5.8. Overall Test Battery Mean

As an additional index of broad cognitive ability, we calculated an overall test battery mean (OTBM) for each PWD. The OTBM represents the arithmetic mean performance across standardized cognitive scores. The OTBM was used for research purposes only and was not part of the DBS consensus decision. For OTBM computation, all standardized data were transformed to z-scores using the equation: $\frac{(X_{is}-M_s)}{S{D}_s}$ where X_is=individual standardized test score and M_s and SD_s=mean and standard deviation for the standardized metric, respectively (e.g., T score M=50, SD=10). Of note, our OTBM was based on all variables displayed on Table 3 with exception of DRS-2 Total Score and Wisconsin Card Sorting Test (WCST) Categories.

Table 3.

Neurocognitive Performance of Grouped by DBS Surgery Decision

Measures	DBS Surgery Conference Decision				Test Statistics
	Eligible (n=21)		Not Eligible (n =11)
	M (SD)	Mz	M (SD)	Mz	t (d)	p a
Global Cognition
OTBM b	−0.4 (0.6)	--	−0.8 (0.5)	--	2.16 (0.82)	.020
DRS-2 Total Score	129.3 (12.0)	−0.8	127.7 (13.8)	−1.0	0.31 (0.12)	.378
Attention
DRS-2 Attention	35.1 (0.9)	0.0	33.7 (3.2)	−0.1	1.63 (0.58)	.058
Digit Span Forward	9.2 (2.4)	−0.3	8.5 (1.0)	−0.6	1.04 (0.43)	.153
Digit Span Backward	5.9 (1.9)	−1.0	4.0 (1.8)	−1.6	2.65 (1.00)	.007
Expressive Language
Phonemic Fluency	30.9 (12.2)	−0.9	20.4 (11.2)	−1.7	2.37 (0.90)	.013
Semantic Fluency	17.9 (7.4)	−0.4	13.8 (5.5)	−1.0	1.61 (0.63)	.059
Visuospatial
DRS-2 Construction	5.6 (1.3)	0.0	4.9 (1.9)	−0.4	1.19 (0.46)	.123
Verbal Reasoning
DRS-2 Conceptualization	34.1 (3.1)	−0.6	33.0 (4.3)	−0.4	0.77 (0.30)	.225
Memory
DRS-2 Memory	21.6 (3.0)	−0.8	21.8 (1.9)	−0.8	−0.17 (0.07)	.435
CVLT-II Trials 1-5	44.0 (11.8)	−0.1	37.9 (14.9)	−0.8	1.24 (0.45)	.112
CVLT-II Trial B	4.4 (2.3)	−0.7	3.0 (2.1)	−1.4	1.65 (0.63)	.055
CVLT-II SDFR	9.4 (3.8)	0.1	6.2 (4.3)	−1.2	2.15 (0.79)	.020
CVLT-II LDFR	9.6 (4.0)	−0.3	6.7 (4.9)	−1.1	1.78 (0.64)	.043
CVLT-II Total Discriminability	2.8 (0.8)	−0.1	2.7 (0.8)	−0.3	0.19 (0.07)	.427
10/36 Spatial IR	18.0 (5.1)	−0.2	18.0 (4.3)	0.0	0.00 (0.00)	.500
10/36 Spatial DR	6.1 (2.4)	0.0	6.4 (1.9)	0.3	−0.34 (0.14)	.368
Executive Functioning
DRS-2 Initiation/Perseveration	32.9 (6.4)	−0.3	33.7 (5.9)	−0.3	0.33 (0.13)	.372
WCST 64 Categories c	2.7 (1.9)	79%	1.8 (1.2)	82%	1.39 (0.55)	.088
WCST 64 Total Errors	20.8 (10.7)	−0.4	31.6 (8.5)	−1.4	−2.85 (1.11)	.004
WCST 64 Perseverative Resp.	15.0 (13.8)	−0.3	28.5 (16.2)	−1.5	−2.41 (0.89)	.012

Note. DBS=Deep brain stimulation; M_z is the mean of norm-calibrated z-scores. t=Independent t-test test statistic with corresponding significance (p) and effect size (d). DRS-2=Dementia Rating Scale, 2nd ed; CVLT-II=California Verbal Learning Test, 2nd ed.; SDFR=short delay free recall; LDFR=long delay free recall; 10/36 Spatial IR and DR=10/36 Spatial Recall Test, Immediate Recall and Delayed Recall, respectively; WCST=Wisconsin Card Sorting Test, 64 card version.

^ap values are 1-tailed to reflect directional hypotheses for cognitive variable analyses.

^bOTBM scores are based on the mean of norm calibrated z-scores across all cognitive data, barring WCST 64 Categories.

^cIn place of norm-calibrated scores, % within normal limits (based on normative data) shown for WCST 64 Categories.

2.6. Data and Statistical Analyses

To evaluate demographic differences between PWD deemed eligible and ineligible for DBS surgery, chi-square (χ²) tests of independence and independent samples t-tests were used for categorical and continuous variables, respectively. To examine the strength of association of the χ² analyses, Cramér’s V(φ_c) was used (interpretation: weak=.10, moderate=.30, strong=.50) [36]. Results of t-tests were converted to Cohen’s d as an index of magnitude of group differences (interpretation: small=0.20, moderate=0.50, large=0.80) [36]. In line with our non-directional hypotheses for demographic differences between eligibility groups, level of statistical significance (α) was set to .05, partitioned across two tails.

To compare cognition between groups, we first examined the normative ranges each group fell in across areas of cognition based on average standardized (z) scores (see Table 2 for z-scores). We then conducted a series of independent samples t-tests across cognitive tests and subtest raw scores. Again, Cohen’s d was used as an index of effect size and according to our a-priori predictions (i.e., PWD eligible for DBS surgery cognitive functioning > than those deemed ineligible), α was set to .05, one-tailed. SPSS version 25 was used for all statistical analyses. Participants with missing data on any variables were excluded from respective statistical analyses.

Table 2.

Pharmacologic and Medical Characteristics by DBS Surgery Consensus Conference Decision

	DBS Surgery Decision n(%)
Medication Type and # Rx	Eligible (n=21)	Not Eligible (n=11)	P(φc)
Cardiac			.599 (.18)
0	8 (38.1)	5 (45.5)
1	5 (23.8)	1 (9.1)
>1	8 (38.1)	5 (45.5)
Anti-PD			.977 (.04)
0	10 (47.6)	5 (45.5)
1	6 (28.6)	3 (27.3)
>1	5 (23.8)	3 (27.3)
Psychotropic			.092 (.39)
0	4 (19.0)	4 (36.4)
1	10 (47.6)	1 (9.1)
>1	7 (33.3)	6 (54.5)
Adjusted RCRI			.016 (.43)
1 (Lower Risk)	18 (85.7)	5 (45.5)
2 (Higher Risk)	3 (14.3)	6 (54.5)

Note. # Rx=Number of medications prescribed are displayed within each medication type. PD=Parkinson's disease; RCRI=Revised Cardiovascular Risk Index; RCRI was modified for the present study to reflect a history of ischemic heart disease, congestive heart failure, cerebrovascular disease, and diabetes mellitus managed by insulin; preoperative serum creatinine data was not available for analysis.

3. Results

3.1. Sample Characteristics

Table 1 displays sample demographic and clinical characteristics stratified by DBS consensus conference decision (eligible vs. ineligible). Of the total sample of PWD (n=32), missing data was noted on the DRS-2 (n=6; 19%) and 10/36 Spatial Recall Test (n=7; 22%). The majority of participants missing data on these measures were in the eligible DBS surgery group. The study sample mostly consisted of middle aged (M=58.4, SD=15.9), high school to college educated (M=13.4, SD=2.9), white (88%), female (69%), right-handed (91%) PWD. There was considerable variability in duration of primary dystonia, but on average, PWD had the disease for a little over 9 years (M=9.3, SD=8.4). The most common primary dystonia subtype was cervical (59%) followed by generalized (28%). In addition, the target for almost all PWD deemed eligible for DBS surgery was GPi (n=19; 8 left, 1 right, 7 bilateral, and 3 staged bilateral), though two had the ventral intermediate nucleus of the thalamus as a target (1 left and 1 bilateral). From available psychological data (most participants had either Beck Depression Inventory-II or Geriatric Depression Scale), there was no difference in depressive symptoms between groups, both measuring in the mildly depressed range. There was a considerable amount of participants missing WRAT-3 Reading data (n=8, 25%); the PWD with valid data had a mean standard score of 91.38 (SD=11.70), indicating broadly average estimated baseline intellectual functioning.

Table 1.

Demographic and Clinical Characteristics by DBS Surgery Consensus Conference Decision

	DBS Surgery Decision
Variable M (SD)	Eligible (n=21)	Ineligible (n=11)	p (effect size)
Age	56.1 (16.9)	62.9 (13.5)	.256 (0.45)
Age of Dx	48.3 (16.2)	50.7 (17.3)	.701 (0.14)
Disease Duration	7.8 (7.2)	12.2 (10.2)	.162 (0.50)
Years of Education	13.9 (2.8)	12.4 (2.9)	.152 (0.54)
WRAT-3 Reading	92.40 (12.92)	89.31 (9.19)	.553 (0.28)
Gender n (%)			.725 (.06)
Male	7 (33.3)	3 (27.3)
Female	14 (66.7)	8 (72.7)
Race n (%)			.482 (.12)
White	19 (90.5)	9 (81.8)
Black	2 (9.5)	2 (18.2)
Relationship Status n (%)a			.864 (.03)
Married	14 (66.7)	7 (63.6)
Divorced	3 (14.3)	2 (18.2)
Single	2 (9.5)	--
Separated	1 (4.8)	1 (9.1)
Widowed	1 (4.8)	1 (9.1)
Handedness n (%)			.968 (.01)
Right	19 (90.5)	10 (90.9)
Left	2 (9.5)	1 (9.1)
Primary Dystonia Subtype n (%)			.357 (.42)
Generalized	7 (33.3)	2 (18.2)
Cervical	12 (57.1)	7 (63.6)
Segmental	1 (4.8)	--
Hemidystonia	1 (4.8)	--
Axial	--	1 (9.1)
Multifocal	--	1 (9.1)

DBS=deep brain stimulation; Age of Dx=age of primary dystonia diagnosis; Disease duration=duration of primary dystonia, in years; WRAT-3=Wide Range Achievement Test; p values for Age, Age of Dx, Disease Duration, Years of Education, and WRAT-3 Reading reflect independent t-test analyses with effect size d; p values for Gender, Race, Relationship Status, Handedness, and Primary Dystonia Subtype reflect χ² analyses with effect size (φ_c.

^aRelationship Status categories truncated for χ² analysis to ‘Married’ and ‘Not Married’.

Approximately 31% of the total sample of PWD had one or more cardiovascular or cerebrovascular comorbidities (e.g., myocardial infarction, coronary artery disease, atrial fibrillation, angioplasty, mitral valve prolapse, transient ischemic accident, cardiomegaly). Polypharmacy was common among participants, with approximately 69% taking five or more pharmacologic agents. The majority of the study sample was prescribed one or more cardiovascular (75%), anti-PD (53%), and psychotropic (75%) medications. Most common cardiac medications included beta-blockers, calcium channel blockers, and anti-thrombotic pharmacologic agents. Anti-PD medications commonly included dopamine agonists/precursors (e.g., carbidopa/levodopa), anti-GABAergics (e.g., baclofen), and anticholinergics (e.g., trihexyphenidyl). Psychotropic medications mostly consisted of benzodiazepines (e.g., clonazepam) followed by selective serotonin re-uptake inhibitors (e.g., citalopram). Across the total sample, PWD prescribed psychotropic medications showed greater auditory-verbal encoding on CVLT-II Trials 1-5 [t(29)=−2.44, p=.021]. While we did not explicitly exclude PWD with a history of dementia, there was no diagnostic incidence of major neurocognitive disorders/dementia in our PWD sample.

Since we observed a difference in cardiac risk factors, we then explored its association with cognition. PWD with higher perioperative cardiovascular risk (RCRI) demonstrated worse broad attention (DRS-2 Attention; [t(24)=2.25, p=.034]). Furthermore, those with high RCRI risk scores showed a pattern of worse letter-guided fluency (M_z higher-risk=−1.43 vs. M_z lower-risk=−1.06), auditory-verbal retrieval (SDFR; M_z higher-risk=−0.93 vs. M_z lower-risk=−0.14), and perseverative responses (WCST; M_z higher-risk=−1.12 vs. M_z lower-risk=−0.59).

3.2. Sample Characteristics by DBS Consensus Conference Decision

Tables 1 and 2 show demographic, clinical, and pharmacologic/medical information. There were no statistically significant demographic differences between DBS surgery eligibility groups. Notwithstanding, the PWD deemed eligible for DBS surgery approached significance for younger mean age (56.1 vs. 62.9, d=0.45), shorter mean disease duration (7.8 vs. 12.2 years, d=0.50), and higher mean years of education (13.9 vs. 12.4 years, d=0.54) compared to the ineligible group (these are broadly moderate effect sizes). There were also no appreciable differences in symptom lateralization between PWD deemed eligible (5 right predominant, 5 left predominant, 11 indeterminate [i.e., no clear lateralization or bilateral]) and ineligible (4 right predominant, 2 left predominant, 5 indeterminate) for DBS surgery. There were no differences between groups across medication categories. Cardiovascular risk status, as assessed by our adjusted RCRI, was significantly associated with DBS consensus conference decision. A greater proportion of PWD in the DBS-eligible group had lower cardiovascular risk (85.7% vs. 45.5%; p=.016, φ_c=.43). Consistently, the odds of being deemed ineligible for DBS surgery were seven times higher in PWD with higher cardiovascular risk (OR = 7.20, 95% CI = 1.31-39.56, p =.016). A greater proportion of PWD who were deemed ineligible for DBS (72.7%) had any history of vascular comorbidity (e.g., myocardial infarction, mitral valve prolapse, atrial fibrillation, coronary artery disease) compared to those in the eligible group (9.5%, p<.001). Lastly, groups did not significantly differ on baseline intellectual functioning as estimated by the WRAT-3 reading test (t=−0.60, p=.55, d=0.28).

3.3. Cognitive Performance by DBS Consensus Conference Decision

Group mean standardized performances across cognitive variables are presented on Table 2. Of note, both study groups exhibited normatively intact mean performance on DRS-2 Attention, DRS-2 Construction, DRS-2 Conceptualization, DRS-2 Initiation/Perseveration, and 10/36 Spatial IR and DR. PWD in the eligible group generally measured in the low average to average normative range across cognitive measures. Conversely, PWD in the ineligible group showed lower mean normative performance across several tasks, especially Digit Span Backward (M_z=−1.6), Phonemic Fluency (M_z=−1.7), CVLT-II Trial B (M_z=−1.4), WCST 64 Total Errors (M_z=−1.4), and WCST 64 Perseverative Responses (M_z=−1.5), which were all impaired. The ineligible group also showed borderline impaired mean performance on CVLT-II SDFR (M_z=−1.2).

Table 3 shows the cognitive performance of PWD deemed eligible or ineligible for DBS surgery. There was a significant group difference in OTBM (p=.020, d=0.82) such that PWD deemed eligible for DBS surgery (M=−0.41, SD=0.64) had higher overall cognitive performance compared to those in the ineligible group (M=−0.85, SD=0.52). This finding generally held when correcting OTBM for PWD with any missing data (p=.058, d=0.76). In addition, relative to PWD deemed ineligible for DBS surgery, those in the eligible group had significantly higher performance on digit span backward (p=.007, d=1.00), Phonemic Fluency (p=.013, d=0.90), CVLT-II SDFR (p=.020, d=0.79), CVLT-II LDFR (p=.043, d=0.64), and WCST 64, exemplified by lower Total Errors (p=.004, d=1.11) and Perseverative Responses (p=.012, d=0.89). The eligible group also approached significance for higher performance on DRS-2 Attention (p=.058, d=0.58), Semantic Fluency (p=.059, d=0.63), CVLT-II Trial B (p=.055, d=0.63), and WCST 64 Categories (p=.088, d=0.55), all with moderate effect sizes. See Table 2 for other non-significant findings.

4. Discussion

The current study represents the first empirical investigation into which variables influence DBS consensus conference decision in a sample of PWD. We examined various demographic, clinical, medical, and cognitive variables in PWD who were evaluated for DBS candidacy by consensus conference at a large academic medical center. Results showed that demographic, clinical, and medication characteristics were not significantly associated with DBS consensus conference decision. Incidentally, relative perioperative cardiovascular risk (and more broadly, vascular disease burden) based on our adjusted RCRI (one or two vascular risk factors) was significantly associated with DBS consensus conference decision. Higher perioperative cardiovascular risk was accompanied with an approximate seven-fold increase of not being a surgical candidate. There is limited research on the relation between cardiovascular risk factors, such as cerebral vascular disease, hypertension, diabetes, atrial fibrillation, and past myocardial infarction, and DBS outcomes in movement disorders. It is unclear whether these risk factors increase the perioperative or postoperative risk, since most studies deem these patients ineligible to participate in their research protocols [14]. Thirty days post-operative DBS-related complications are rare with mortality and morbidity rates of only 0.4% and 1%, respectively, with complications more common in older adults as well as people with PD, rather than PWD [37]. Therefore, it is unlikely that the PWD who were determined to be ineligible for DBS were rejected based solely on their vascular risk factors. Rather, it is more likely that the observed executive dysfunction associated with primary dystonia is exacerbated by their underlying vascular disease, and they are denied by the committee due to their pronounced executive functioning deficits.

Interestingly, there were no diagnostic instances of psychiatric illness or major neurocognitive disorder in our sample of PWD. Thus, it was likely that the 11 PWD in our sample were considered ineligible for DBS surgery by a consensus conference in the absence of frank dementia and/or mental health factors. Given our findings, these patients were more likely to be excluded due to executive dysfunction, which does not meet the level of impairment required for a dementia syndrome diagnosis. This is consistent with the PD DBS literature, where patients with cognitive dysfunction in the absence of dementia (or those with considerable medical comorbidities) may still be deemed poor candidates [13, 15-19, 38]. Though there were no group difference on the WRAT-3 Reading, we attempted to compare these scores to the DRS-2 Total Score to identify cases of detectable cognitive decline (≥ +1 SD difference in standardized score between the WRAT-3 Reading minus DRS-2 Total Score). Unfortunately, we could only analyze 59% of the sample (due to concurrently missing WRAT-3 and DRS-2 data), which resulted in four cases (3 eligible and 1 ineligible) with possible relative cognitive decline from their presumed baseline. While we were unable to appreciate any relation between cognitive decline and DBS eligibility among our sample of PWD, it is possible this relation may be seen in a larger sample. It should be noted that of these four cases where there was a relative cognitive decline, none of these PWD were diagnosed with a dementia syndrome during their presurgical neuropsychological assessment.

Six separate cognitive variables were significantly associated with DBS consensus conference decision in our sample. Across cognitive analyses, the largest group difference effect size was seen on a test of executive functioning, followed by a test of letter-guided verbal fluency and free delayed recall, both of which have a frontally mediated retrieval component [39, 40]. This aligns with DBS literature in PD, in which those with already low verbal fluency ability are at risk for further post-surgical decline, and subsequent decrement to quality of life [13, 41, 42]. Following DBS surgery, further verbal fluency decline is common, irrespective of target [41, 43, 44], possibly raising concern of the risks to quality of life outweighing the benefits of the procedure in PWD.

Deficits in executive function and attention/working memory have been found in individuals with primary dystonia [5-7, 45-49]. Functional neuroimaging studies have demonstrated evidence of altered fronto-striatal activity in PWD, consistent with our observed pattern of executive dysfunction. Specifically, resting-state functional MRI studies in patients with cervical dystonia have demonstrated reduced functional connectivity between the right mid-dorsal putamen and right external globus pallidus basal ganglia regions and the fronto-parietal network [50], diffuse alterations in functional connectivity among the bilateral postcentral gyrus, dorsal cingulate, basal ganglia, and thalamus [51], reduced global brain functional connectivity in the right M1-supplementary area and precentral gyrus [52]. Similarly, research using magnetoencephalography in patients with cervical dystonia has shown evidence of broadly reduced intrahemispheric and interhemispheric network connectivity across fronto-frontal, fronto-parietal, fronto-striatal, fronto-occipital, parieto-parietal and temporo-parietal brain regions [53]. Deficits in executive function and attention/working memory have also been observed in people with significant cardiovascular risk factors such as a myocardial infarction, atrial fibrillation, large vessel and microvascular disease, diabetes, hypertension, and congestive heart failure in need of a transplant or placement of a left ventricular assist device [54-62]. It is likely that the presence of mild attention and executive functioning difficulties in PWD are exacerbated in those who also have high RCRI scores, thus making them more likely to be deemed ineligible for DBS surgery.

Cognition, particularly executive functioning, predicting DBS consensus conference decision has important clinical implications. Attention and executive dysfunction may be associated with DBS postsurgical outcome in movement disorders, and executive functioning has been shown to be an important predictor of postsurgical outcome in other patient groups as well. In a study of non-demented older adults, poor executive function predicted postoperative delirium following major, elective noncardiac surgeries [63]. Similarly, executive dysfunction has been shown to predict postoperative delirium following coronary artery bypass graft surgery [64]. Impaired attention may also predict DBS surgical outcome. In a PD sample, impaired attention predicted longer hospitalizations following DBS surgery [38]. These studies illustrate that specific cognitive abilities (i.e., executive functions and attention) likely affect postsurgical outcomes.

The study has several limitations. Since this was a retrospective study, we were unable to obtain granular information regarding the specific reasons for DBS ineligibility. To this end, there are many reasons candidates may not be deemed appropriate candidates for DBS surgery which were not addressed in the current study. For example, some potential DBS recipients may lack family or caregiver support or have other contraindicating medical conditions. In addition, as indicated in the PD DBS decision-making literature, cognitive decline in the context of dementia is a strong contraindication for surgical candidacy [15]. However, our ability to thoroughly examine the statistical relationship between cognitive decline and candidacy in PWD was limited by the absence of PWD and a co-occurring dementia diagnosis in our sample. Moreover, the base rate of possible significant cognitive decline in our sample was very low (n=4). In a similar vein, our relatively small sample size was insufficient to capture adequate instances of psychometrically defined cognitive decline to formally investigate this relation. Finally, only 59% of our sample had both valid WRAT-3 Reading and DRS-2 data (further compounding our sample size issue). Clinically though, none of the PWD were diagnosed with a dementia syndrome by their neurologist or neuropsychologist.

As we were unable to consider several factors in this study, future studies may investigate type of gene mutation (e.g., DYT1), pallidal volume, white matter abnormalities on neuroimaging, and cognitive decline in relation to DBS decision-making in primary dystonia. Moreover, further research examining cognition and vascular risk factors in PWD (among other movement disorders) would provide additional support for our findings. It should be noted that this study included a small, but well-characterized and studied sample from a single consensus conference team at one movement disorder center. Along these lines, our small sample limited our ability to parcel out the unique effects in our cognitive and vascular analyses. Other DBS consensus conference teams at different centers may weigh certain variables differently to determine whether DBS surgery is performed. A future multi-site study could examine how multiple DBS consensus conference teams reach their clinical decisions. Finally, attention/working memory deficits are a significant predictor of medical decision-making capacity in cognitively impaired populations such as mild cognitive impairment, Alzheimer’s disease, and traumatic brain injury [65, 66], as well as executive dysfunction in PD [67]. Therefore, more research is needed to determine if the constellation of executive and attention deficits found in primary dystonia has a functional impact (e.g., medical decision-making capacity) and the extent to which it influences DBS consensus conference decisions.

5. Conclusions

In summary, cognitive dysfunction (i.e., attention, working memory, executive functioning) was associated with DBS consensus conference decisions not to proceed with device implantation. While PWD with higher vascular risk factor scores were seven times more likely to deemed not be eligible for DBS surgery, this is likely due to vascular risk factors exacerbating the PWD-related executive dysfunction, rather than a concern about peri- and post-operative risk given the low mortality and morbidity rates associated with DBS surgery.

Highlights.

• Deep brain stimulation (DBS) is an effective treatment for motor disturbance in patients with primary dystonia (PWD).

• Numerous factors are considered by an interdisciplinary consensus conference before deciding whether a candidate for DBS surgery (e.g., demographic, medical, cognitive, and behavioral factors). Little is known about which of these factors are associated with consideration in PWD DBS surgery consensus conference decisions.

• Thirty-two dystonia patients completed comprehensive neurological and neuropsychological exams in consideration for potential DBS surgery.

• An interdisciplinary conference committee either recommended or did not recommend DBS therapy based upon these data.

• Across clinical variables, PWD who were deemed ineligible for DBS surgery had a higher cardiac risk factors.

• PWD who were classified as ineligible displayed lower global cognitive functioning, working memory, phonemic fluency, memory retrieval, and cognitive flexibility.

• Incidentally, we observed elevated cardiac risk among these individuals, likely reflecting the relationship between vascular health and cognition.