Psychosis

Psychosis, as described in the previous chapter, is a departure from reality, is an unfortunate yet all too common side effect of PD management. Psychosis can be intrinsic to PD or drug induced, and therefore is discussed twice in this book. The presence of psychosis in those receiving PD treatment ranges from 5 to 30% [6]. Psychosis in PD may manifest as a range of perceptual disturbances and abnormal thought content. Hallucinations in PD most frequently present as visual hallucinations but may also present as auditory, olfactory, tactile or gustatory disturbances. Patients describe the presence of very vivid animals or people, and often retain insight into the artificial nature of these hallucinations, especially early in the disease process. Patients are typically not threatened by the hallucinations, in contrast to schizophrenia. In fact, unless educated proactively, patients may not volunteer this information out of embarrassment. They may also worry that divulging this information may lead to treatment being withheld or discontinued. Delusions, defined as fixed false beliefs, may surface and range from bizarre to subtle. False accusations by patients of spousal infidelity are frequently seen in PD and create significant distress for spouses and families. Delusions may also be persecutory or somatic in nature. Illusions, the incorrect interpretation of an actual sensory stimulus, may also be reported in PD. Illusions in PD have been classified as “minor hallucinations” [7]. Other minor hallucinations may include reports of brief hallucinations in the peripheral vision or a feeling of someone or something nearby (without persecutory ideation). Because of the difficulties in capturing minor hallucinations that are mild and brief, the true prevalence of psychosis in PD has been difficult to quantify. A recent community-based study approximated that a quarter of nondemented patients with PD likely experience psychosis, with half of these experiences manifesting as minor hallucinations [8]. To improve detection and reporting accuracy, the National Institute of Neurological Disease and Stroke (NINDS)/ National Institute of Mental Health (NIMH) workgroup recently proposed a criterion set for psychosis in PD that encompasses any of the above manifestations of psychosis occurring for greater than 1 month [7]. Reported predictors of psychosis in PD include longer disease duration, visual dysfunction, higher use of DRT (particularly dopamine agonists), presence of rapid eye movement (REM) sleep behaviors, and cognitive dysfunction [6].

Disturbance in mesocorticolimbic dopaminergic neurotransmission has historically been thought to be largely responsible for psychosis in PD. This association in PD was generated primarily from the observation that dopamine replacement therapy has been associated with the development of psychosis, while dopamine antagonists have demonstrated efficacy in reducing psychotic phenomena in PD. The precise mechanism, however, has been poorly understood, as not all patients with PD develop psychosis in response to DRT. Recent evidence has suggested the contribution of other neurotransmitters in the generation of psychosis in PD. Psychosis in PD is often reported as a predictor of dementia in PD. As a result, acetylcholine deficiency has been implicated based on overlapping clinical features with Alzheimer’s disease and evidence of degeneration of the nucleus basalis of Meynert in PD [9]. Moreover, cholinesterase inhibitors, whose mechanism is aimed at limiting the breakdown of acetylcholine, have shown some promise in the treatment of mild psychosis in PD [10], while anticholinergic agents are notorious for triggering hallucinations and altered mental states. Serotonin imbalance is another consideration in the pathophysiology of psychosis. Subtypes of serotonin receptors, particularly 5-HT2A, have been implicated in the generation of psychosis based on: (i) modulation of dopamine neurons in the ventral tegmental area by hyperstimulation of 5-HT2A [9]; (ii) awareness of agents that are known to act via 5-HT2A and produce psychosis, such as LSD (lysergic acid diethylamide); and (iii) evidence of atypical antipsychotics (which antagonize 5-HT2A) to successfully treat psychosis [11]. Particularly striking and supportive of 5-HT2A’s prominent involvement in PD psychosis is the relatively low doses of atypical antipsychotics that can effectively treat psychotic phenomena in PD. These doses are often 10-fold lower compared with primary psychotic disorders [5], potentially suggesting a predominant serotonergic mechanism in PD. Studies examining visual hallucinations in PD have found clustering of 5-HT2A alterations in the temporal cortex and ventral visual pathway of patients with PD experiencing visual hallucinations, further supporting the contribution of serotonin to psychosis in PD [12].

The balancing act of providing optimal motor control without triggering psychosis is a common, yet sometimes challenging, dilemma in PD management. On the one hand, patients may require and/or request DRT to manage deteriorating motor symptoms while on the other hand experiencing vivid hallucinations that may predispose them to fear, confusion and/or falls. Prior to discontinuation of DRT, the first step in evaluating a patient with psychosis is to conduct a thorough evaluation to rule out any acute medical disturbances, such as infection, cerebrovascular event, respiratory insufficiency and/or metabolic derangements, which may predispose to confusion or delirium [13]. Delirium may differ from treatment-related psychosis in PD due to the presence of altered sensorium characterized by waxing and waning leves of consciousness, greater disorientation and no apparent association with DRT timing or dosing. Even in the setting of clear DRT-related psychosis, these should be investigated, since the underlying disturbance may have increased susceptibility to psychosis. Resolution of a systemic disturbance in some cases may then allow the patient to better tolerate DRT. Because of the frequent reports of hallucinations during the evening and upon awakening during the night (when visual acuity is lessened), the differential diagnosis should include consideration for Charles Bonnet syndrome, REM sleep behavior disorder, hypnogogic and hypnopompic hallucinations, and dementia [5]. Finally, evaluation of environmental, or extrinsic, factors should not be underestimated. In the setting of intrinsic vulnerabilities, patients may be prone to experience DRT-related psychosis in environments with ambiguous noise and poor lighting, particular if the setting is new or unfamiliar. Psychological support may confer some augmenting benefit for those with mild psychosis where insight is retained, but this has not been well studied.

Once the evaluation for metabolic abnormalities, coexisting medical and psychiatric conditions and environmental factors has not yielded any obvious contributing factors, the patient’s PD medication regimen should be closely reviewed to eliminate any nonessential agents. Patients may, on occasion, still be taking medications that derive minimal motor benefit but contribute significantly to the emergence of psychosis. Dose reduction and discontinuation of these higher-risk agents (in the order anticholinergic agents, amantadine dopamine agonists and then COMT inhibitors, high risk to low risk) in a stepwise fashion [13] may be prudent as side effects multiply with increasing numbers of drugs [2]. Patients may ultimately be best managed with levodopa therapy alone.

Patients with significant or functionally disabling psychosis may require the addition of a psychotropic agent in order to treat DRT-related psychosis. Clozapine has demonstrated the best efficacy in PD psychosis [14] but remains grossly underutilized due to the burden of use (e.g. patient registry, coordination with pharmacy, frequent laboratory visits, side effects, warnings). Some patients may also express reservation due to knowledge of its reserved use in chronic or “severe” schizophrenia. Quetiapine is prescribed more frequently for PD psychosis due to ease of use, augmenting benefits (i.e. nighttime sedation) and favorable extrapyramidal risk profile. One retrospective study in a Veterans Affairs population found that quetiapine was prescribed in 66% of those with PD psychosis [15]. Unfortunately, there has been insufficient evidence in randomized controlled trials to recommend quetiapine in the treatment of psychosis in PD [14]. All other atypical antipsychotics have not demonstrated a clear benefit in controlled trials and are more likely to worsen motor symptoms in PD. Typical or conventional dopamine antagonists (such as haloperidol) should be avoided altogether due to the strong propensity to worsen parkinsonism. Pimavanserin, a selective 5-HT2A inverse agonist with minimal to no dopaminergic effects, is gaining attention. Its promise has been further strengthened by a recent successful randomized, placebo-controlled phase 3 trial in PD psychosis [16].

Impulse control disorders

According to the Diagnostic and Statistical Manual of Mental Disorders, 4th Edn, Text Revision (DSM-IV-TR), the category of impulse control disorders (ICDs) was characterized by a failure to resist an impulse, drive or temptation to perform an act that is harmful to the person or to others, with increased tension or arousal preceding the act, and pleasure or relief upon completing the act [17]. Pleasure-seeking impulse control behaviors seen in PD include pathological gambling, compulsive sexual behavior, compulsive buying, and compulsive or binge eating [18]. Impulse dyscontrol in PD commonly arises from DRT and may range from these pleasure-seeking behaviors to less purposeful behaviors, such as “punding” (compulsive fascination with and performance of repetitive, mechanical tasks, such as assembling and disassembling objects, collecting or sorting) or “walkabout” (unnecessary and aimless wandering). Patients rarely endorse intrusive egodystonic thoughts as a reason for their behavior, as typically seen in obsessive-compulsive disorder, and may not always have a conscious explanation for their actions. Impulse control disorders may occasionally present with elevated or expansive mood. Presence of clear hypomania and mania should prompt suspicion for dopamine dysregulation syndrome, or in some cases may also exist as its own distinct DRT-related entity [19]. The inconsistent presence of pleasurable (vs nonpleasurable) urges or drive may suggest the potential for differing mechanisms among the ICDs [20]. In the most recent edition of DSM-V [21], the simple category of ICDs has now been eliminated. Instead the prior ICDs have been reclassified into several different categories, comprising: (i) disruptive, impulse control and conduct disorders; (ii) obsessive-compulsive and related disorders (i.e. hoarding disorder); and (iii) substance-related and addictive disorders (i.e. gambling disorder). This presumably better reflects the heterogeneity in impulsive–compulsive conditions.

Similar to non-PD patients, pathological gambling may present in a variety of forms, including irresistible urges to purchase scratch-off tickets, participate in online gambling games or play slot machines at the casino. Gambling on items with delayed gratification (i.e. sports betting) is relatively uncommon. Hypersexuality presents as a maladaptive preoccupation with sex, ranging from compulsive masturbation and viewing of pornography to excessive requests for sex from a partner, extramarital affairs and habitual promiscuity. Compulsive shopping manifests as an intense preoccupation with buying or shopping above what can be afforded and/or is necessary. Compulsive or binge eating commonly includes cravings for sweets. Punding may represent its own distinct entity based on the purposeless nature. In some cases, punding may be referred to as “hobbyism” based on the intensification of an individual’s already existing personal, professional and recreational interests [22]. Impulse control disorders and punding may also arise in the setting of excessive and/or compulsive use of DRT, referred to as dopamine dysregulation syndrome. Diagnostic criteria, including duration and functional impairment, have been proposed for each of these ICDs to improve detection and reporting [20], although it is possible to overlook subsyndromal cases that may still cause distress to patients and caregivers [23].

Observational studies have indicated that ICDs may occur at a higher frequency in PD than in the general population. An early retrospective database review identified a link between pathological gambling and dopamine agonists (DAs) in PD [24]. The largest epidemiological study (over 3000 patients) demonstrated that an ICD was identified in 13.6% of PD patients, and this was subdivided into gambling (5%), sexual behavior (3.5%), compulsive buying (5.7%) and binge-eating disorder (4.3%), with 4.9% of patients having two or more ICDs [25]. Dopamine agonists were strongly linked to an increased risk of ICDs in this cross-sectional study, and ICDs also appeared to be more frequent in men and associated with a greater mean daily DA dosage. Other independent variables associated are levodopa use, living in the USA, younger age, being unmarried, current cigarette smoking and a family history of gambling problems [25]. The degree of risk of ICDs among different DAs is believed to be similar, although recent evidence suggests a potential for lower risk with transdermal DA preparations [26].

Impulse control disorders involve alterations in neurotransmission and neural circuitry, particularly in reward pathways within the mesocorticolimbic pathways [18]. Dopamine agonists have a higher ratio of D3:D2 and D3:D1 striatal activation. D3 receptors are concentrated in limbic structures, and stimulation contributes to psychiatric and behavioral effects, compared with D1 and D2 receptors within the dorsal striatum [27]. The regions that appear most implicated include the ventromedial and orbitofrontal regions of the prefrontal cortex, and structures of the limbic system including the ventral striatum and amygdala [28]. These structures, particularly the nucleus accumbens within the ventral striatum, are associated with modulating motivated behaviors underlying engagement in risky behaviors, while the dorsal striatum is involved in motor habits [28]. Altered ventral striatum activation and dopamine transporter density, particularly in the right striatum, have been demonstrated in patients with ICD in PD [29, 30]. Moreover, patients receiving DRT, particularly DAs, are more sensitive to positive reinforcement, leading to impaired ability to avoid negative outcomes [18].

Practitioners should routinely obtain a careful history of preexisting symptoms or conditions (such as psychiatric or substance abuse) prior to starting DRT, most notably DA therapy. In addition, clinicians should educate patients on the risks prior to treatment, while monitoring for ICD behaviors during follow-up visits. Patients may withhold reporting these behaviors due to uncertainty about their relationship to DRT or embarrassment. There are no widely accepted or routinely utilized ICD screening instruments in PD. Although there are several general instruments available to screen for impulsivity (i.e. Minnesota Impulsive Disorders Interview), these instruments do not comprehensively screen for the spectrum of ICDs in PD. The Questionnaire for Impulsive–Compulsive Disorders in Parkinson’s Disease – Rating Scale (QUIP-RS) may therefore be considered based on its intent for this purpose. The QUIP-RS is a rating scale designed to assist in the diagnosis and severity measurement of ICDs and related disorders in PD [31]. Routine use of this instrument may potentially serve a dual role: (i) to inform patients of the potential behaviors that may arise during the course of treatment (better vigilance); and (ii) to normalize the behaviors as a product of treatment when they do occur (better reporting).

Upon the discovery of an emerging or active ICD, patients should first be evaluated for any immediate safety concerns. Spouses and families may need to be alerted to damaging or serious consequences of the ICD behaviors, such as financial or sexual indiscretion that can quickly escalate. Management may also involve referral for inpatient psychiatric management if the behaviors cannot be de-escalated, or an acute threat to self or others is present. Ultimately, the management of an ICD in PD will entail an alteration in DRT. Studies have demonstrated recovery or improvement with discontinuing or significantly reducing DA treatment [32]. Discontinuation of DAs should be done gradually to minimize the risk of withdrawal symptoms. It may be necessary to implement additional specific strategies, such as preventing access to bank accounts or credit cards, while DRT is being lowered. Upon discontinuation of DAs, additional supplementation of DRT in the form of levodopa may be necessary to maintain motoric stability. Caution should be maintained for triggering compulsive behavior, particularly if higher doses of levodopa are needed. Intestinal levodopa infusion may potentially be a safer consideration [33]. The use of DBS, particularly subthalamic nucleus (STN)-DBS, may be a potential intervention due to the ability to reduce DRT postoperatively. These options are considered in the following three sections. Lastly, the utility of certain psychotropic agents, such as atypical antipsychotics and mood stabilizers, in alleviating impulsivity has been reported but not yet proven for use in treating ICDs [14].

Dopamine dysregulation syndrome

Dopamine dysregulation syndrome (DDS) refers to a severe behavioral condition that arises as a result of compulsive use of DRT in PD, commonly manifesting as hoarding, deceit, impulsivity, impatience, manipulation, aggression and poor insight [34]. As PD progresses, there is a need to escalate doses of DRT to gain better control of motor symptoms. A subset of predisposed patients may increase doses of DRT in excess of what is necessary for symptomatic control. Patients will often engage in escalation of DRT despite obvious disabling psychiatric and/or motoric consequences (i.e. dyskinesias). The compulsive aspects of DDS share some overlap with ICDs but appear to be a separate entity, driven by negative reinforcement, habit formation and/or incentive sensitization [34]. As a result, patients with DDS are motivated by a pathological “wanting” for DRT, rather than a “liking,” similar to proposed addiction models [35]. Surreptitious use of DRT may therefore be undertaken to avoid the negative “off” effects (dysphoria) more so than to achieve euphoria or pleasure from “on” states. The potential for different subtypes (pleasure seeking versus dysphoria avoiding) has been suggested, but the exact pathophysiology still remains unclear [23]. Dramatic affective cycling and thought disturbances (such as hypomania, mania and psychosis) may be part of the clinical presentation. Dopamine dysregulation syndrome has synonymously been referred to as “hedonistic homeostatic dysregulation” in order to highlight the desperate, pathological, compulsive and behavioral disturbances of this condition [36].

Proposed diagnostic criteria for DDS include: (i) clinical diagnosis of levodopa-responsive PD; (ii) a need for increasing DRT in excess of what is necessary; (iii) pathological use despite psychiatric and motoric consequences; (iv) functional impairment; and (v) observed dopaminergic withdrawal state with dose reduction [36]. The exact prevalence of DDS is not well defined due to its overlap with ICD, limited reporting and lack of prospective studies but appears to be in the range of 4% [36]. There appears to be a higher association with levodopa and other short-acting DAs, such as apomorphine, suggesting that repeated pulsatile stimulation of shorter-acting oral agents may contribute to the emergence of DDS. Associated factors that have been reported include previous drug abuse, higher-sensation-seeking personalities, depression and creative professional backgrounds [34]. A recent retrospective longitudinal population-based cohort study identified 35 cases of DDS over a 6-year period of which 91% were associated with compulsive levodopa use and 9% due to DA addiction [37]. Patients with DDS had younger PD onset and more severe motor fluctuations, and DDS was also more likely to be associated with a history of depression, a family history of PD, and personal or family history of drug use. Similar to ICD in PD, there are alterations in dopamine neurotransmission within limbic circuits during DDS [38], potentially representing a behavioral counterpart to levodopa-induced dyskinesias [39].

There is currently no evidence-based treatment for DDS, but the approach must be aimed at reducing the dopaminergic burden. All rescue or taken-as-needed DRT should be discontinued. This, however, can be difficult due to resistance from the patient with clear “dependency” for the DRT regimen. Similar to ICD, management of medication adjustment must be done with caution, as there may be increased anxiety or depression as doses are adjusted or decreased. Insight may be poor and assistance from caregivers to provide supervision is essential, as lack of monitoring is an associated trigger for relapse [37]. Caregivers should be alerted that aggression may surface when doses are not provided early or in excess as demanded by the patient craving DRT. Improvement may be possible with patients who successfully reduce doses of DRT; however, long-term outcome data is not available at this time. Some psychotropic agents investigated include treatment with antidepressants, mood stabilizers and atypical antipsychotics, which have shown a mixed response with no generalizable recommendations [34]. Cognitive-behavioral strategies aimed at disengaging negative somatic interpretations from the reactive need for additional DRT dosing has been reported as useful in supporting “off”-period distress in DDS [23]. Conversion to intestinal levodopa infusion or DBS implantation has been reported as a potentially helpful consideration in resolving DDS [40], while continuous apomorphine infusions appear much less favorable [37], although these recommendations need further investigation.

Dopamine agonist withdrawal syndrome

It is commonly recognized that substances with the propensity to trigger acute pleasurable states and reward-driven behavior (i.e. cocaine) have the potential for negative discontinuation effects. This may occur with abrupt or, in some cases, gradual cessation. Recently, a distinctive and stereotyped constellation of symptoms has been recognized with DA discontinuation, referred to as dopamine agonist withdrawal syndrome (DAWS) [4]. The syndrome is characterized by several key features, including: (i) the presence of psychological and physical disturbances with DA dose reduction or discontinuation; (ii) correlation of worsening symptoms with the extent of DA dose reduction; (iii) improvement of distress with resumption of the DA; (iv) lack of improvement with other dopaminergic agents (i.e. levodopa); and (v) significant psychosocial and functional impairment [4]. Similar to other disturbances discussed in this chapter, the reason for its occurrence in some patients and not others remains elusive. The symptoms of DAWS range from psychiatric and autonomic to generalized. Symptoms include anxiety (panic, restlessness and phobia), mood changes (irritability, dysphoria, depression and hopelessness), autonomic effects (diaphoresis, flushing and orthostasis) and generalized disturbance (agitation, pain, sleep difficulties and fatigue). The frequency of DAWS in PD ranges from 7.8 to 19% [41]. The syndrome may be difficult to identify in PD as many of the symptoms of DAWS may be confused with the sequelae of PD progression or the expected “wearing off” phenomena. Moreover, unlike other drug-withdrawal syndromes, there are no clear objective findings (e.g. vital sign changes, pupil reactivity, piloerection).

Several studies have noted a strong association (up to 100% of subjects) of DAWS occurring in those with a prior history of ICD [41]. As noted earlier, DAs have relative selectivity for D3 dopamine receptors, which are found disproportionately in limbic pathways and therefore frequently responsible for ICDs. Furthermore, drug cravings are often seen in both presentations, similar to other addictions. This raises the question of whether DAWS may be a manifestation along a continuum of behavioral disturbances in PD arising from the same etiological dysfunction as ICD and DDS [42]. In the initial published report of DAWS, the authors hypothesized the presence of a “mesocorticolimbic” variant of PD that is characterized by disproportionate dopaminergic dysfunction in the mesocorticolimbic connections, as opposed to nigrostriatal [4]. This hypothesis is supported by: (i) DAWS occurring in patients with less severe motor symptoms (Unified Parkinson’s Disease Rating Scale [UPDRS] scores) than those who did not experience DAWS with DA taper [4]; and (ii) functional imaging evidence of mesolimbic dopaminergic denervation [43].

The discussion about management of DAWS must begin with strategies to avoid the occurrence of DAWS altogether. Dopamine agonist withdrawal syndrome has been associated with higher DA doses and longer DA exposure [44]. Subsequently, practitioners should be perpetually cognizant of this in the setting of longer-term management of PD. Providers should also be attentive of the risk in the context of postoperative management of DBS surgery (see below) when there may be pressure to reduce the dopaminergic burden. At present, there is no evidence of a difference in risk of DAWS between one DA over another or in the rate of DA taper [44], although conventional wisdom would likely favor a more gradual reduction. Once the symptoms of DAWS appear, the management of DAWS should first start with an investigation for other contributing factors, such as parallel systemic aberrations (e.g. infection, primary psychiatric disorder, other drug-withdrawal effects). If there is no clear contributing reason, and a recent reduction in DA dosage was conducted or found, there should be a high suspicion for DAWS.

There are no definitive evidence-based recommendations available for the treatment of DAWS. Frequent communication (via office visits and telephone communication with patients or caregivers) is essential for those at high risk for or actively experiencing DAWS. A milder, self-limiting form of DAWS may occur and resolve with supportive management (or without intervention) in a subset of patients [45]. However, for some, a more severe and debilitating form of DAWS will present. As noted above in the defining features of DAWS, the supplementation of other dopaminergic agents is not effective. Therefore, reintroduction of DAs may be the only consideration for palliation. The full reinstatement of DA dosage should be done with extreme caution, as reports have noted the recurrence of ICD in a chronic and debilitating fashion in these situations [4].

The lowest effective dosage is recommended when resuming or increasing DA dosage to alleviate DAWS. One report observed a necessary dose titration to a DA–levodopa-equivalent daily dose of 160mg to treat DAWS in three patients [45]. It is also important to mention that, although other dopaminergic agents may be ineffective in managing DAWS, patients may inappropriately and unknowingly attempt to self-medicate their distress with compulsive or higher-than-prescribed doses of other DRT (i.e. levodopa), potentially exposing a vulnerability to DDS [4]. Therefore, patients should be informed in advance of this inappropriate consideration and approach in management. For those who ultimately are unable to cease DA use (due to triggering DAWS), close monitoring should be established and education of long-term risks must be discussed.