Abstract
INTRODUCTION
Although the definition of Parkinson’s disease (PD) is based on the motor triad of resting tremor, rigidityand bradykinesia, a variety of non-motor symptoms are associated, including disabling neuropsychiatric symptoms [1]. Neuropsychiatric symptoms encompass two opposite motivational expressions of a continuous behavioral spectrum with hypodopaminergic symptoms on one end and hyperdopaminergic symptoms on the other end of the spectrum [2–5]. Susceptibility for both hypo- and hyperdopaminergic symptoms may in part be governed by dopaminergic denervation [2, 6–8]. Apathy, depression and anxiety are hypodopaminergic symptoms observed in all stages of the disease, especially at its onset, with disease progression [9–12] and in the context of withdrawal of dopaminergic treatment [8, 13]. Prevalence of apathy is estimated between 17% and 70% in PD patients, whereas depression is present in 30–35% , depending on assessment tools, patient selection and antiparkinsonian treatment [11, 14–16]. Apathy is responsive to dopaminergic treatment in early-stage PD [17] and in the context of withdrawal of dopaminergic medications [18–20].
Hyperdopaminergic symptoms include impulse control disorders (ICDs), behavioral addictions, punding and addiction to dopaminergic drugs [2–5]. Non-physiological dopaminergic stimulation by antiparkinsonian drugs is known to induce hyperdopaminergic symptoms [2, 3]. In addition to hypo- and hyperdopaminergic symptoms patients may also experience non-motor fluctuations, including mood brightening, euphoria, impulsive pleasure seeking behavior and mania during ON periods as well as anxiety, apathy and depression during OFF periods [2, 21].
Although psychiatric symptoms have traditionally been considered as a contraindication for deep brain stimulation, this has become matter of debate. Early reports recommended exclusion of PD patients with behavioral problems from surgery [22], due to decompensation and new appearance of behavioral disturbances (e.g. mania, aggressiveness, euphoria) related to STN deep brain stimulation [23–26]. Although more recent prospective studies found clear beneficial effects of STN DBS on ICDs, behavioral addictions and dopamine addiction, resulting from reduction of dopaminergic medications after surgery [4, 27], the existing literature does not consider ICDs or dopamine addiction as a primary treatment indication for DBS surgery [28].
The prevalence of many neuropsychiatric symptoms such as punding and hyperemotionality is unknown, because of the lack of consistent diagnostic criteria and the use of different evaluation tools [29]. Furthermore, a rigorous evaluation of the effect of STN DBS on neuropsychiatric symptoms is limited, as baseline neuropsychiatric symptoms experience has never been compared between surgical candidates and PD patients in general. Such an analysis however, is highly relevant, as surgical candidates by definition are suffering from severe motor complications thus typically take high dose dopaminergic drugs. Therefore we describe in this paper the behavioral spectrum in a large prospective single center cohort of PD surgical candidates. We compare this to the behavioral evaluation performed in a cohort of PD patients in different stages of the disease, more representative of PD patients in general.
In analogy with the motor complications of dopaminergic treatment, i.e. motor fluctuations and dyskinesia, we assumed that non-motor complications of the treatment in the form of (1) psychic non-motor fluctuations and (2) hyperdopaminergic behavioral symptoms would be more frequent in surgical candidates compared to a more general population of PD patients at different stages of their disease. Following this logic, we hypothesized that (3) hypodopaminergic symptoms might be less frequent in PD surgical candidates. To test these hypotheses we compared the distribution of whole spectrum of neuropsychiatric symptoms among these populations based on evaluation with the Ardouin Scale of Behavior in PD (ASBPD) [3, 5].
PATIENTS AND METHODS
This study is an observational, multicenter, cross-sectional study comparing two cohorts. The evaluations of both cohorts were conducted prospectively using the same tool for behavioral evaluation (ASPBD).
Cohort of PD surgical candidates, assessments
234 PD STN-DBS surgical candidates recruited between January 2005 and January 2015 at the Movement Disorder Center in Grenoble (France) were included. All patients had routine evaluation performed at the Grenoble Institution. Inclusion criteria were those for surgery, i.e. clinical diagnosis of PD by a movement disorder neurologist according to the United Kingdom PD Brain Bank criteria, disabling levodopa-related motor complications despite optimal adjustment of anti-parkinsonian medication and age less than 70 years. Exclusion criteria were inability of the patient to participate in complete standardized behavioural evaluation based on the French version of the ASBPD, an atypical Parkinson’s syndrome, candidates for re-implantation, DBS target other than STN, ongoing severe depression at high risk of suicide or florid psychosis, a Hoehn & Yahr stage (HY) greater than IV in the ON state, and dementia [30]. Active ICD was not an exclusion criterion. Assessments took place in the month prior to surgery. Unified Parkinson’sDisease Rating Scale (UPDRS) parts I-V were performed, the motor evaluation took place in both the ON and OFF medication condition using supra-threshold doses of levodopa [31]. Neuropsychiatric symptoms were evaluated by a neuropsychologist familiar with PD and neuropsychiatric disorders in movement disorders, using the ASBPD (as described below). Beck depression inventory II (BDI-II) [32] was used to determine the severity of depressive symptoms and the Starkstein apathy scale (SAS) [33] to measure apathy.
Cohort of PD patients in general, assessments
The second cohort presented PD patients in general, with different disease stages. The study included 260 PD patients from 13 centers across four countries: France, Spain, the United Kingdom and the United States. This population has been described previously in the validation study of the ASBPD [3]. The primary outcome was behavioral evaluation as measured with the ASBPD, which was completed by a psychiatrist, a neuropsychologist or a clinical psychologist familiar with PD and neuropsychiatric disorders in movement disorders. UPDRS part I-V was assessed on chronic treatment condition. Montgomery and Asberg Depression Rating Scale (MADRS) [34] was used for the assessment of depression and the Lille Apathy Rating Scale (LARS) [35] for apathy. Patients with dementia according to the Diagnostic and Statistical Manual of Mental Disorders IV (DSM IV) were excluded.
Specific assessment of neuropsychiatric symptoms: The ASBPD
Changes in mood and behavior were evaluated using the recently validated Parkinson-specific ASBPD [3, 5]. This scale consists of 21 items, grouped into three sections: hypodopaminergic behavior (part I), psychic non-motor-fluctuations (part II) and hyperdopaminergic behavior (part III). Part I successively evaluates depressed mood, anxiety, irritability or aggressiveness, hyperemotionality and apathy. Part II evaluates the psychological state that is associated with motor symptoms in the ON and OFF states in fluctuating patients. Part III assesses the presence and severity of behavioral disorders induced by dopaminergic treatment. Hyperdopaminergic behaviors include: hypomanic mood, psychotic symptoms, nocturnal hyperactivity, diurnal somnolence, eating behavior, creativity, hobbyism, punding, risk-taking behavior, compulsive shopping, pathological gambling, hypersexuality, dopaminergic addiction and excess in motivation. The frequency and intensity of a symptom’s occurrence in the preceding month is rated on a five-point scale (absence of disorder 0; mild disorder 1; moderate disorder 2; marked disorder 3; severe disorder 4). A score of 2 indicates a modification in behavior that is significant enough to require therapeutic adjustment. A score >2 reflects clear-cut maladaptive pathological behavior, requiring therapeutic intervention.
Total levodopa equivalent daily dose (LEDD), levodopa-only LEDD and dopamine agonist-only LEDD were calculated according to Tomlinson et al. [36] in all patients, except for non-dopaminergic drugs (amantadine).
Ethics
Procedures were performed in compliance with existing laws and institutional guidelines. Patients participating in the validation study had given their written informed consent for anonymous data evaluation. Surgical candidates are part of a prospective single-center cohort follow-up of surgical patients from the Grenoble center and preoperative evaluation belongs to clinical routine performed in all patients.
Statistical analysis
Statistical analysis was performed using Stata 13 software (StataCorp LP, College Station, TX, US). The tests were two-sided, with a type I error set at
RESULTS
Demographic and clinical aspects of both cohorts are listed in Table 1. Both groups were similar in gender. Surgical candidates were significantly younger than in the general PD population but had longer disease duration. Between-group differences were seen for UPDRS III motor score in the ON condition, with a better motor score for surgical candidates. UPDRS III OFF score was 39.1 for surgical candidates. No significant difference was seen in mean Hoehn & Yahr stage. Mean antiparkinsonian treatment doses were significantly higher in PD surgical candidates, for total LEDD, levodopa-only LEDD and dopamine agonist-only LEDD.
Overall cognitive performance, measured with Mattis Dementia Rating Scale was preserved in surgical candidates. Surgical candidates reported less depressive symptoms as measured with BDI-II compared with the general PD population, assessed with MADRS. The prevalence of apathy was similar between the cohorts, although again different scales were used.
Descriptive statistics of neuropsychiatric symptoms as measured with the ASBPD are represented in Table 2 as scale percentages of prevalence of each symptom (score ≥2). Non-motor fluctuations were significantly more frequent in surgical candidates, in both the ON and OFF condition (Fig. 1). All the 14 hyperdopa-minergic neuropsychiatric symptoms were more frequently encountered in surgical candidates, except for diurnal somnolence and compulsive shopping (Fig. 2). After adjusted multivariate logistic regression analysis, the following neuropsychiatric symptoms remained significantly more frequent in surgical candidates: nocturnal hyperactivity, hobbyism, dopaminergic addiction and excess in motivation.
Among the hypodopaminergic symptoms only depressive symptoms were significantly less frequent in PD surgical candidates (Table 2). Irritability/aggressiveness, hyperemotionality and apathy were more frequent in PD surgical candidates; however, these results were not statistically significant.
DISCUSSION
This comparative study demonstrated that PD surgical candidates experience clinically relevant neuropsychiatric symptoms more frequently than PD patients in general. All the 3 of the initial hypotheses were confirmed. (1) Non-motor fluctuations and (2) hyperdopaminergic symptoms were more frequent in PD surgical candidates. Even after adjustment on age, disease duration, UPDRS III-score and treatments, there was a remarkable difference for dopaminergic addiction, with an odds ratio of 10.83. The odds ratio for excess in motivation in PD surgical candidates was 4.02. Surgical candidates also experienced more frequent nocturnal hyperactivity and excessive hobbyism. Among the hypodopaminergic symptoms (3) only depressive symptoms were less frequent in surgical candidates compared to controls.
PD surgical candidates were younger, had longer disease duration and had much higher antiparkinsonian treatment doses than PD patients in general, particularly dopamine agonists. PD surgical candidates also used antipsychotic and anxiolytic drugs morefrequently.
The confirmation of our hypotheses is compatible with a shared pathophysiological mechanism for motor and neuropsychiatric complications of dopaminergic treatment in PD. In conjunction with the denervation of midbrain dopamine neurons in the substantia nigra and ventral tegmental area, the capacity for dopamine storage in presynaptic intra-axonal vesicles is reduced. Both motor- and non-motor effects of dopamine thus become increasingly dependent on blood concentrations of dopaminergic medications. In addition to these pharmacokinetic differences, postsynaptic sensitization mechanisms in different cortico-basal ganglia thalamo-cortical loops will also add to motor and behavioral complications [37].
While under-stimulation of the dopamine receptors is responsible for the PD-specific motor symptoms, antiparkinsonian medication induces motor fluctuations and dyskinesia through non-physiologic, pulsatile stimulation of dopamine receptors. We postulate a common underlying mechanism for motor- and non-motor fluctuations as well as for dyskinesia and hyperdopaminergic behaviors. As motor complications of dopaminergic treatment are an indication for STN DBS, a higher prevalence of non-motor symptoms is to be expected in a PD population of surgical candidates.
Non-motor fluctuations and hyperdopaminergic neuropsychiatric symptoms
After adjustment on age, disease duration, UPDRS III-score and treatments, dopaminergic addiction (score ≥2) was almost 11 times more frequent in our cohort of PD surgical candidates. Eusebio et al. [27] reported 16.3% of compulsive dopamine users in their population scheduled for surgery, based on psychiatric evaluation using the following criteria: rapid increase in dopaminergic drug dosage, automedication, inability to reduce dopaminergic drugs, marked OFF dysphoria, and euphoria and hypomania in the ON state [38]. Lhomme et al. [4] found a comparable prevalence of 19.4% in a surgical population, using the ASBPD. The 11-fold increased risk for developing dopaminergic addiction could be explained by selection bias. Almost 45% of the surgical candidates experienced relevant non-motor OFF symptoms (score ≥2). Patients report these often as an extremely intense and bothersome dysphoria, which probably contributes to the patients’ decision making for surgery. The psychological suffering related to non-motor OFF symptoms lies at the heart of the dopamine dysregulation syndrome (DDS) [39]. OFF-period anxiety may also explain why surgical candidates use anxiolytic drugs more frequently.
We also observed more frequent “excessive motivation”, a behavioral item of the ASBPD that evaluates overall behavior on a hyperdopaminergic mode. Excessive motivation can be considered the opposite of apathy. Excessive hobbyism and nocturnal hyperactivity are often egosynthonic and are generally not a complaint of patients and their families. Given their frequency and potential devastating social impact/repercussion, systematic evaluation for these symptoms is useful in PD. The largest study on ICDs (hypersexuality, pathological gambling, pathological shopping and bulimia) showed that one or multiple ICD was identified in 13.6% of patients and in 17.1% of patients treated with a dopamine agonist [40]. The data of our cohort representing the general PD population approximate this, except for eating behavior, which was more frequently observed in our cohorts. Comparing our cohort of surgical candidates with our cohort of PD patients in general, we did not find any significant differences on ICDs, except for increased eating behavior in surgical candidates. After adjustment on age, disease duration, UPDRS III-score and treatments however, statistical significance was no longer observed. A previous study has found postoperative improvement in all ICDs except for increased eating behavior which obviously needs a more complex management than only decrease in medication [27, 41].
Hypodopaminergic neuropsychiatric symptoms
Among the hypodopaminergic symptoms only depressive symptoms were less frequent in surgical candidates at baseline evaluation prior to surgery. The hyperdopaminergic behaviors of surgical candidates suggest excessive dopaminergic stimulation. It is known that dopaminergic treatment may have a beneficial effect on apathy and depression [18, 42]. The fact that surgical candidates consider DBS as possible solution of their problems, may also contribute to lower depression rates at baseline evaluations before surgery.
Clinically relevant depression is present in 30–35% in the general PD population [16]. We observed lower prevalence rates in both populations using the ASBPD, which measures a relative
Baseline characteristics
Selection bias occurred in both cohorts as dementia was an exclusion criterion for both cohorts; age younger than 70 was an inclusion criterion for only the cohort of PD surgical candidates. Not surprisingly, surgical candidates were significantly younger than the cohort representing the general PD population. Surgical candidates also had longer disease duration. Even after correction for covariates (multivariate logistic regression adjusted on age, disease duration, score UPDRS III and treatments), remarkable differences between both cohorts regarding behavioral profiles persisted. Previous reports described that young-onset PD patients typically develop motor fluctuations and dyskinesia earlier in life [44]. Young-onset PD patients represent the prototype of patients selected for surgery, suffering from severe motor symptoms, often treated with a dopamine agonist in combination with levodopa, earlier age at onset and non-demented, thereby fulfilling the selection criteria of STN DBS.
PD surgical candidates in our series used antipsychotic treatments more often than PD patients in general. Antipsychotic drugs can be part of the management of hyperdopaminergic behaviors [20]. Furthermore, at the Grenoble institution, clozapine is part of the preventive peri-operative management of patients judged at risk for psychotic decompensation during the often stressful surgical intervention without general anesthesia.
Clinical implications
While psychiatric symptoms are historically considered as a contraindication for surgery [22], our data demonstrate that the prevalence of neuropsychiatric symptoms is higher in typical candidates for surgery, compared to the general population of PD patients. While pathological hyperdopaminergic behaviors and non-motor fluctuations are difficult to manage, they should not be considered a contraindication but rather part of an emerging larger spectrum of symptoms that need to be considered when discussing surgical indications. Prospective studies have shown that STN DBS can improve dopaminergic addiction, non-motor fluctuations and behavioral addictions, although such improvement is highly dependent on post-surgery management of dopaminergic treatment [4, 27]. While occurrence of ICDs have been reported after DBS when dopaminergic medication was not reduced [45], excess drug decrease can lead to a typically delayed hypodopaminergic state [4, 46]. The level of postoperative dopaminergic stimulation needs to be adapted to both motor and non-motor state, considering that DBS has no effect on dopaminergic denervation and defective dopamine storage and release. Continuous, non-pulsatile treatment remains necessary despite the possibility of decreasing unitary drug doses. Drug requirements may have to be adapted over time if desensitization of motor and psychotropic effects of dopaminergic drugs occurs [47].
It is known that psychiatric symptoms such as depression, anxiety, apathy, non-motor fluctuations and ICDs have a higher impact on quality of life than the motor symptoms [46, 48–50]. The management of surgical patients therefore should aim at improving neuropsychiatric symptoms as much as motor symptoms. To achieve this, including a routine systematic behavioral evaluation is necessary, which is not the case today. In contrast, a recent survey has illustrated a striking under-illustration of psychiatric assessment of PD patients in DBS centers [51].
Our study contributes to a better understanding of the outcomes of neuropsychiatric symptoms after subthalamic nucleus DBS. Based on our findings, we encourage a systematic evaluation of the spectrum of neuropsychiatric symptoms during PD treatment.
CONFLICT OF INTEREST
The study received no specific funding, apart from Valrie Lamberti, who was supported by the ParkinsonVereniging and got an Erasmus plus grant from the Radboud University of Nijmegen for a research semester as part of her medical training.
Valrie Fraix has received a travel grant from Abbvie and honoraria for lecturing from Medtronic.
Anna Castrioto has received travel expenses reimbursement to scientific meetings from ORKYN and
Abbvie and grants from the Edmond J. Safra Foundation and Medtronic.
Frank Durif received personal money from Allergan, Novartis, Orkyn, Lundbeck and Teva for board membership. The Clermont-Ferrand Institution received grants from the health ministry, France Parkinson foundation, INSERM (French National Institute of Health and Research in Medicine), Fondation de l’avenir and Fondation de France.
Paul Krack was granted funds from Medtronic, St Jude and Boston Scientific for research purposes in the field of deep brain stimulation. He received research support from France Parkinson foundation, INSERM, Edmond J. Safra Foundation, Grenoble University Hospital, Orkyn, Novartis, UCB, LVL, Boston Scientific. He received reimbursement of travel expenses to scientific meetings or honoraria for lecturing or consultation from Medtronic, Euthrapie Company, Novartis Pharma, UCB, St. Jude, Lundbeck, Boehringer Ingelheim, Orkyn, Abbott, Orion, TEVA and Boston Scientific.
Bruno Pereira, Eugnie Lhomme, Amlie Bichon, Emmanuelle Schmitt, Pierre Pelissier, Andrea Kistnerand Rianne Esselink have no potential interests to declare.
Footnotes
ACKNOWLEDGMENTS
The authors thank Laurens Lambert for criticalreading of the manuscript.
