The Psychiatric and Neuropsychiatric Symptoms After Subthalamic Stimulation for Parkinson’s Disease

Consider the function of the basal ganglia in the selection and facilitation of movement, cognition and emotion. The STN has a regulatory role, providing an inhibitory NO-GO signal in the cortico-subcortical networks of the basal ganglia.⁹ A gradient of representations exists within the STN, with motor activity located in the dorsolateral aspect of the nucleus, cognitive-associative circuits in the intermediate zone, and limbic circuits in the ventromedial region.¹⁰ Functional inhibition of the STN using high-frequency stimulation facilitates movement but may release cognitive and affective disinhibition in certain individuals. The small size of the STN means that current diffusion from accurately placed electrodes may still inadvertently modulate networks implicated in mood, thinking, and reward, including the anterior cingulate cortex, the orbitofrontal cortex, and the ventral tegmental area. Electrodes situated in the ventromedial STN¹¹ or the substantia nigra¹² seem more likely to produce these nonmotor side effects.

In PD, variable patterns of cell loss within the basal ganglia also modify the outcome of DBS. Mesencephalic dopaminergic projections within motor and nonmotor networks degenerate at different rates,¹³ such that motor and nonmotor responses may vary for a given level of stimulation. Following DBS, STN stimulation sufficient to restore a degenerated hypodopaminergic motor circuit may cause a functional imbalance in an adjacent nondegenerated nonmotor circuit.

The postoperative presentation resembles mania or hypomania, with a variable incidence of between 1% and 15%,^14,15 although stimulation-induced reversible depressive symptoms have also been described following inadvertent electrode placement in the substantia nigra.¹⁶ Stimulation effects can be distinguished from intraoperative microlesioning by their persistence and temporal association with DBS adjustment. Clinically, the most common features are an elevated or irritable mood, racing thoughts, impulsivity, and an increase in goal-directed activity. Insomnia and agitation may also be present. Some cases manifest attenuated symptoms, making the diagnosis more difficult to capture and potentially accounting for the discrepancy in reported incidence. Close family members such as spouses are nonetheless likely to detect these changes and commonly complain that the patient is no longer himself. Impulsivity is frequently noted and may be a common endophenotype in this cohort, defined informally as the tendency to act prematurely, without foresight. Some neuropsychological correlates of impulsivity are known to change after STN DBS, such as impaired response inhibition¹⁷ and increased speed of decision making in situations of conflict.¹⁸

Treatment with DRT is also associated with neuropsychiatric complications that may persist or worsen after DBS. In these patients, exogenous dopamine disrupts the function of the basal ganglia in a manner that predisposes to impulse control disorders (ICDs) such as pathological gambling, hypersexuality, compulsive shopping, binge eating, and punding.¹⁹ These disorders appear unified by the concept of compulsivity, namely a maladaptive perseveration of behavior that is inappropriate to the situation and results in undesirable consequences. There is an overlap with impulsivity with the two differentiated by their relevance to different aspects of response control: impulsivity to response initiation and compulsivity to response termination. Correspondingly, patients with ICDs also score highly on neuropsychological measures of impulsivity,²⁰ are more likely to have a novelty-seeking temperament, and are more likely to have experienced medication-induced hypomanic symptoms.²¹

Preclinical evidence implicates the nucleus accumbens in the pathophysiology of these behaviors.²² A region of the ventral striatum that is vulnerable to dopaminergic destabilization, the nucleus accumbens is associated with hedonic tone and reinforcement learning. Chronic dopaminergic stimulation provided by DAs attenuates the regulatory drive of the prefrontal cortex to this region, with a predominance of appetitive limbic inputs and a loss of behavioral flexibility. This process is augmented by sensitization of the ventral striatal neurocircuitry²³ and the affinity of DAs for D₃ receptors that predominate in this region.²⁴ Functional imaging indicates that patients with ICDs have a specific vulnerability to deactivation of frontal regulatory circuits,²⁵ and recent work identified a structural vulnerability to impulsivity in gray matter atrophy of fronto-striatal regions.²⁶

The relationship between STN DBS and ICDs is controversial. By permitting substantial postoperative reduction in DRT, STN DBS has been considered a potential treatment for ICDs. However, this is not consistently supported in the literature, with 70% of patients with ICDs remaining unchanged or worsening after DBS in one series.²⁷ New-onset ICDs have also been reported after DBS.²⁸

If postoperative mania is caused by disinhibition of limbic and associative networks, then underactivation of these same nonmotor circuits also produces characteristic neuropsychiatric sequelae. Substantial reduction of DRT following STN DBS is common practice as electrical stimulation in motor regions of the STN predominates. It may, however, unmask symptoms such as apathy, depression, and anxiety in up to 50% of patients.²⁹ For these individuals, DRT has a psychotropic effect and withdrawal symptoms arise on discontinuation,³⁰ particularly in those treated with dopamine agonists.³¹ The symptoms of this dopamine agonist withdrawal syndrome include anxiety, agitation, depression, irritability, insomnia, and suicidal ideation, and respond to DA repletion. Dopamine agonist withdrawal syndrome, therefore, may be a driver of postoperative neuropsychiatric complications after STN DBS.³² As DRT withdrawal is not a problem for all patients, the PD-related denervation pattern may explain this hypo-dopaminergic vulnerability, with afflicted individuals having greater mesocorticolimbic dopaminergic denervation than those without symptoms.²⁹

We contend that many post-DBS neuropsychiatric disturbances have a biological basis in neurostimulation, dopaminergic tone, and neurodegeneration. STN DBS and the attendant alterations in DRT represent an acute insult to a compensated but brittle homeostatic system regulating movement, mood and cognition—already anatomically and physiologically disrupted prior to surgery. There may arise a postoperative dissociation between motor and nonmotor outcomes as stimulation parameters and DRT are optimized to restore degenerated motor circuits, with resultant imbalance in topographically distinct functional networks.

Despite the psychiatric sequelae of these complex disruptions to brain neurocircuitry, the natural history of post-DBS neuropsychiatric symptoms appears to trend to an attenuation of symptoms (including suicide risk) in the months and years that follow DBS. It is unclear whether this represents the progression of neurodegeneration or a slow recovery of stable homeostasis via preserved mechanisms of neuroplasticity. This process plays out in an individual with a unique temperament, specific constellation of psychiatric morbidity, and a support network with a variable capacity to respond to disruption. It is particularly in the early postoperative course that the psychiatrist’s expertise is relevant, when nonmotor circuits seem at particular risk of dysregulation and the patient at risk of an adverse outcome.

Suitability for STN DBS can be considered from a psychiatric and surgical perspective. This provides both clinical and ethical challenges for the psychiatrist. Neuropsychiatric symptoms may be present only at subclinical levels prior to DBS, and it is difficult to identify those patients who may have postoperative deterioration. Alternatively, more florid behavioral problems may be concealed by patients and their families, fearful that psychiatric symptoms will affect surgical eligibility. It is difficult to conceive of denying surgery to patients identified as possessing risk factors, particularly if they have exhausted other treatment options and are capable of informed consent. The psychiatrist may be asked to stratify patients as high or low risk for poor nonmotor outcome, but this is a complex matter shaped not only by the patient’s psychiatric history but also by factors such as rapport with the psychiatrist and willingness to engage in close postoperative follow-up. Hopefully this issue can be addressed in subsequent work.

Should all patients undergo a psychiatric assessment prior to STN-DBS? Neuropsychiatric symptoms are prevalent in the population presenting for surgery, even after screening by a neurologist.³³ The advantages of preoperative engagement with the patient are that it allows assessment of the psychiatric baseline and treatment of existing psychiatric problems. For example, having a history of a depressive disorder may be less important than being in remission at the time of surgery, such that the psychiatrist may delay DBS to allow the initiation of antidepressant therapy.

The authority of the psychiatrist in the multidisciplinary team is not yet established. At what stage should the psychiatrist intervene to veto or delay DBS? Should the psychiatrist be able to suggest DBS of other brain nuclei that may be less likely to lead to neuropsychiatric complications? Clearly, navigating these issues demands a collegiate working relationship with the neurologist and neurosurgeon at the DBS center.

The issue of suicide following STN DBS for PD was highlighted by a retrospective case series that found it to be a predominant cause of mortality in the first postoperative year,³⁴ although the quality of these data are diminished by its retrospective nature. The only randomized controlled trial to examine this issue did not find an increased rate of suicide in the DBS group at 6 months postoperatively,³⁵ although this study was arguably underpowered to detect a significant difference in such a rare event. Although DBS itself does not cause suicide, a constellation of downstream phenomena may make suicide more likely in vulnerable individuals. This is a nuanced concept that requires careful discussion with patients and their families.

In the aforementioned retrospective series, postoperative depression was the factor most strongly associated with completed suicide and suicide attempts after DBS.³⁴ Identifying those patients likely to suffer postoperative depression is more difficult. Depression is one of the most common nonmotor symptoms of PD, and a preoperative history of depression is thus relatively nonspecific. In the absence of clear evidence, the psychiatrist might consider the duration, severity, and treatment refractoriness of previous depressive episodes, as well as whether these appeared as a feature of PD or as part of a premorbid history of major depressive disorder. Suicidal ideation is also relatively common in PD,³⁶ but a history of suicide attempts is of more concern. In one sample, severe depression at baseline did appear to predict severe depression 3 years after DBS, emphasizing the importance of treating active depressive symptoms prior to surgery.⁷

A history of the dopamine dysregulation syndrome or an ICD was also associated with a risk of attempted suicide after DBS.³⁴ This may be a marker of both impulsivity and a liability to develop a dopamine agonist withdrawal syndrome–induced depressive state, as comorbidity of dopamine agonist withdrawal syndrome with ICDs and dopamine dysregulation syndrome is common. Prior to surgery, nonmotor fluctuations in mood and anxiety, responsive to a levodopa (LD) challenge, also predict a postoperative vulnerability to dopamine agonist withdrawal syndrome and a complicated post-DBS outcome.²⁹ In clinical practice, however, identifying this phenomenon is hindered by the frequent disavowal of the nonmotor effects of DRT by patients themselves, with OFF period dysphoria often expressed in motor terms alone. It is conceivable that stimulation and dopamine agonist withdrawal syndrome could interact within the same individual to effect a dangerous combination of depressed mood and impulsivity. This may be related to the observation that in this cohort, suicide attempts appeared more likely to result in death than other populations.³⁴

The presence of a psychiatrist in the multidisciplinary team is evidently prudent, but may not mitigate the likelihood of suicide in all cases. Certain individuals may be reluctant to engage with the psychiatrist, for reasons of additional cost, stigma, or alexithymia. This makes it substantially more challenging to accurately and effectively manage this group. As well as failing to report relevant features of the psychiatric history, such patients can have a great aversion to accepting psychiatric care, despite the presence of suicidal urges. Even if the psychiatrist successfully identifies these cases, direct intervention may still be resisted. Here the psychiatrist faces a dilemma: can the patient be managed indirectly through other clinicians at the DBS center or must the psychiatrist proceed with involuntary treatment under the provisions of the relevant local legislation? The latter course of action may be necessary in cases where the patient lacks the capacity to make treatment decisions, but is likely to be a profoundly distressing experience that may compromise ongoing rapport and worsen engagement in the longer term.

At present there is little evidence to guide identification of those patients at risk of postoperative impulsivity or manic symptoms. Major determinants of these phenomena are postoperative, such as inadvertent stimulation of the cognitive and limbic STN. Case reports⁸ suggest that a personal or family history of bipolar disorder and a history of medication induced mania may be risk factors. The significance of premorbid, subsyndromal bipolar traits is not established, but the concept of DBS unmasking a hitherto latent mood disorder bears consideration. The question of whether some patients have existing structural or functional vulnerabilities due to neurodegeneration or dopaminergic stimulation was raised in the preceding section, but it is not common practice to screen for these using neuroanatomical or neuropsychological methods and their connection to postoperative problems has not been established.

The use of STN DBS to treat preoperative ICDs is debatable. One group has demonstrated that abrupt cessation of DAs and marked reduction in DRT after surgery led to remission of baseline ICDs in their cohort.³⁷ However, switching rapidly from a hyperdopaminergic to a hypodopaminergic state brings its own complications, with emergent apathy and depression that may themselves require “rescue” with DAs. A higher rate of suicide attempts in this sample also supports the view that these patients are more psychiatrically unstable candidates. If ICDs have been detected preoperatively, a gradual taper of DA can begin prior to surgery, potentially smoothing the postoperative course. Whether severe preoperative ICDs should be a contraindication to STN DBS is another question: there is not clear longitudinal data to guide clinicians. Patients and their families should be warned that the ICD may persist or even worsen after surgery. In those patients with compulsive use of DRT, DBS is likely to treat motor symptoms but will not reduce the nonmotor requirements for dopamine. Again, exposing patients to DRT withdrawal prior to surgery allows them to develop a tolerance to these nonmotor OFF states. The psychiatrist may advocate for globus pallidus DBS in severe cases, based on the presumption that this target is associated with fewer neuropsychiatric complications. However, the outcome did not differ in one small retrospective series.³⁸

Aside from effects on neuropsychological constructs of impulsivity, the most consistent effect of STN DBS has been to impair semantic and phonemic verbal fluency (reviewed in Voon et al.⁸). For most patients, DBS has relatively benign cognitive sequelae, but many clinical studies exclude patients older than 70 years of age and those with significant preexisting cognitive impairment, variously defined. There are cases of significant cognitive deterioration that occur soon after DBS and seem temporally linked to the procedure rather than to progression of PD. Risk factors for this complication may include older age, total presurgical DRT requirement, and axial symptoms, potentially reflecting advanced disease.³⁹ Intraoperative variables are also relevant, such as damage to cortical tissue during the passage of electrodes.

The presurgical psychiatric assessment should routinely include an assessment of cognition. In PD, this should include measures of visuospatial and executive function. Both the Montreal Cognitive Assessment⁴⁰ and the Addenbrooke’s Cognitive Examination⁴¹ examine a distributed set of cognitive domains, including verbal fluency, yet can be administered rapidly. Patients older than 70 years of age should be counseled about the potential, yet poorly quantified, risk of accelerated cognitive deterioration after the procedure. Those with preoperative cognitive impairment should be warned that they may not benefit as greatly from DBS on quality-of-life outcomes.⁴² The finding of preexisting cognitive impairment should prompt an assessment of the patient’s capacity to consent to DBS. Although DBS is generally contraindicated in cases of dementia, the psychiatrist should work with the multidisciplinary team on a case by case basis when this is advocated, contributing to the process of substitute decision making.

Patients with preoperative hallucinations and other psychotic symptoms linked to excessive dopaminergic stimulation may improve following STN DBS with postoperative reduction of DRT. However, the presence of psychotic symptoms signals advanced disease and should prompt a careful screen for significant cognitive impairment that might exclude surgery. In cognitively intact patients with psychotic symptoms, however, the outcome of DBS appears favorable.⁴³

PD is a disease that classically disrupts the developmental transition from working life to early retirement. Long-held plans and goals are modified or given up. DBS offers a chance for the patient to recover the life that should have been, was it not for PD. Typically, the patient with PD exists in a family system, most often with a spouse or partner, in which roles are established to balance the restrictions of disability with the demands of everyday life. As part of the reversal of disability, roles in the marital and family system must now be renegotiated to take account of postoperative changes. The patient and their spouse may have divergent opinions regarding the ideal outcome from surgery and how much the existing system should change. This process of adaptation can be disrupted by emergent neuropsychiatric symptoms, which alienate the sufferer from his support network at a crucial point in the postoperative journey. A loyal partner is discarded in a euphoric behavioral activation as the patient seeks to indulge his newfound freedom from motor disability. Other spouses struggle to accept that DBS is not a cure for PD and expect the depressed and apathetic sufferer to resume an entirely normal role within the family.

The psychiatrist generates a unique understanding of each case, including the internalization of PD as part of the patient’s self-concept, associated negative cognitive distortions, coping styles, and expectations of postoperative life. Negotiating a mutuality of expectations between the patient, family, and surgical team is an important step in mitigating the fantasy of cure or the unrealistic belief that DBS will resolve existing psychosocial difficulties. Families should also consider their capacity for increased forbearance in the immediate postoperative period during which neuropsychiatric symptoms are more common.

Although the psychiatrist can neither predict nor prevent all neuropsychiatric complications, screening for risk factors alerts the clinician to a potential worsening of neuropsychiatric symptoms. Certainly, the outcome after DBS seems worse if preoperative ICDs go undetected.²⁷ If known, risks can be addressed by ensuring that neuropsychiatric issues are treated at the time of surgery, and the patient and family are educated about behaviors they may encounter following DBS. Unknown risks result from a preoperative denial of neuropsychiatric symptoms and the multiplicity of biological, psychological, and social determinants that may contribute to a poor nonmotor outcome following DBS. A key skill of the DBS psychiatrist is communicating these unknowns to the surgical candidate in a manner that increases his capacity to make an informed decision about DBS.

The psychiatrist’s role in the postoperative phase is to detect and manage neuropsychiatric symptoms before they substantially impair functioning and quality of life. It is neither necessary nor feasible to longitudinally follow all patients, so the psychiatrist relies on the patient and his support network for the prompt detection of behavioral disturbances. This highlights the importance of preoperative education to raise awareness of these issues. The psychiatrist is skilled in nonmotor assessment and may be able to elicit nonmotor symptoms that are overlooked by or concealed from other clinicians at the DBS center.

With regard to treatment, the psychiatrist should also be familiar with adjustment of stimulation settings and comfortable with dosing DRT. These appear to be key interventions for post-DBS neuropsychiatric problems. Postoperatively, clinicians in the DBS team must balance titration of stimulation and withdrawal of medication, a process likened to walking a tightrope between the hyper-and hypoactivated clusters of nonmotor symptoms. Tuning of the device and adjustment of medication to manage neuropsychiatric problems may necessitate the sacrifice of some motor response and thus requires collaboration with the treating neurologist and neurosurgeon. In refractory cases, the psychiatrist may also provide expertise in additional pharmacological and psychotherapeutic strategies, although the evidence base for these in the DBS population is limited.

These symptoms can be devastating, especially for families who were not forewarned about a potentially rocky postoperative course. The link between ICDs and caregiver burden in PD is recognized.⁴⁴ Even less florid disinhibition can be burdensome to the patient’s spouse, who is often ill prepared to tolerate this personality change. Mood changes commonly appear to be substantially stimulation dependent: switching to a dorsolateral stimulation contact, using bipolar rather than monopolar stimulation, or reducing voltage can be helpful. When in bipolar mode, reversing the polarity of the active electrodes changes the shape of the stimulation field and also reduces diffusion to unwanted targets. However, affecting these changes may meet with resistance. Mood elevation is pleasant, and a return to euthymia is often accompanied by substantial deterioration in motor performance. Managing this situation is challenging and demands a working alliance with the patient.

Pharmacological treatment with an atypical antipsychotic or mood stabilizer may permit preservation of motor benefits⁴⁵ and is an option for refractory cases. Due to their rapid dissociation from the D₂ receptor, quetiapine and clozapine are the antipsychotics of choice in PD, being least likely to induce extrapyramidal symptoms. However, data for the efficacy of quetiapine for psychotic symptoms in PD are inconclusive,⁴⁶ and clozapine is associated with considerable monitoring requirements to reduce the risk of myocarditis, cardiomyopathy, and agranulocytosis. Efficacy of either drug for the treatment of post-DBS neuropsychiatric symptoms has not been examined.

The psychiatrist can be called on to provide serial assessments of mental state during a phase of DBS reprogramming by the movement disorder specialist. This may also involve an opinion on capacity, safety, and the need for hospitalization to reduce the risk of misadventure or damage to reputation. Sensible behavioral management in this period includes placing temporary limits on financial transactions and restricting access to motor vehicles. Once these issues have abated, the psychiatrist can anticipate working with the family to process some of the grief and anger consequent to these behaviors, allowing the patient to reestablish his role in the family system. In the most severe cases, there may also be considerable legal, social, and financial matters to address, such as criminal convictions, divorce, disputes over testamentary capacity, and outstanding debts.

In chronological order of appearance, depression can be ascribed to stimulation effects, levodopa withdrawal, psychological adjustment, and nonmotor progression of PD.

Direct stimulation-induced depressive symptoms are rare. When optimally placed, the immediate effects of stimulation seem to have a modest mood-elevating effect.⁴⁷ Depression and anxiety arising as a consequence of DRT withdrawal are most likely to emerge in the initial postoperative months. In such patients, increasing LD or reintroducing a DA reverses these mood changes. The role of pramipexole in the treatment of depression in PD is established.⁴⁶ In the longer term, it is not clear whether reductions in DRT should remain modest or whether neuropsychiatric symptoms are simply part of an unpleasant withdrawal syndrome that will abate in time. Antidepressant medication may address these symptoms, although again the evidence base is inconclusive. In depressed patients with PD, there is more evidence for the efficacy of tricyclic antidepressants over selective serotonin reuptake inhibitors,⁴⁶ potentially reflecting the importance of noradrenergic reuptake inhibition, although the former are likely to be less well tolerated due to anticholinergic side effects. During this period, verbalized suicidality should be taken seriously in an impulsive population where attempts carry high lethality. The psychiatrist may again be required to closely monitor the patient’s mental state and consider inpatient treatment to mitigate the risk of suicide. Where depression arises in the setting of a complicated adjustment to postsurgical life, the psychiatrist may take more of a psychotherapeutic role, revisiting the expectations of the patient and his family and investigating social determinants of wellbeing.

Characterizing the more distal effects of STN DBS on mood is challenging. Variability in outcome is likely to be governed by such factors as personality, family support, and adjustment to illness, with some evidence to suggest that patients who are poorly adapted at baseline continue to suffer psychosocial impairment.⁴⁸ Post-DBS behavioral disorders may isolate the PD sufferer. Nonmotor disease progression is also relevant, given that the prevalence of depression in pharmacologically managed PD patients is substantial.

Apathy includes both loss of motivation and loss of emotionality. This is manifest in diminished goal-directed activity and reduced spontaneous or evoked emotional display.⁴⁹ It overlaps with but is distinct from depression and dementia. Following DBS, apathy is relatively common in the setting of DRT withdrawal and may be part of the spectrum that includes anxiety, depression, and suicidal ideation. It responds to reintroduction of DAs,⁵⁰ suggesting that apathy is a feature of dopamine agonist withdrawal syndrome in some patients. Apathy emerging later in the postoperative course is more likely to represent nonmotor disease progression, and addition of DAs in these patients might lead to confusion rather than improvement. Behavioral measures such as prompting by caregivers or structured environmental stimulation are more suitable.

Biological aspects of post-DBS neuropsychiatric symptoms inform the psychiatrist’s management. Balancing stimulation with dopaminergic medication parallels the balancing of motor and nonmotor outcomes. The psychiatrist will gather information from other clinicians in the DBS center and the patient’s family to get a comprehensive picture of postoperative behavioral change. Working with families to contain neuropsychiatric disturbance is key, particularly when caregiver burden is onerous, and relationship rupture seems imminent. Alliance with patients may be complicated by denial and preference for motor performance.