Movement disorders are frequently a result of prescription drugs and of illicit drug use. In this chapter, we focus on prescribed drugs but briefly mention drugs of abuse. The main emphasis is on movement disorders due to dopamine-receptor–blocking agents (DBAs). However, movement disorders due to other drugs are also briefly reviewed.

MOVEMENT DISORDERS DUE TO DBA

The advent of “atypical antipsychotic” led to the hope that the incidence of drug-induced movement disorders (DIMDs) will decrease significantly. Unfortunately, the hope has not been realized, and the movement disorders due to DBAs continue to be a significant clinical problem. Chlorpromazine was first developed in France in 1952 and was introduced in the United States in 1954. It was heralded as a major advance in the treatment of psychosis (1). However, there were significant acute side effects, such as akathisia, drug-induced parkinsonism (DIP), and acute dystonia. A more disturbing persistent dyskinesia was first recognized in the late 1950s (2). In the past 40 years or so, significant clinical experience has accumulated with the use of DBA, and today, the movement disorders caused by these drugs can be classified (Table 33.1).

ACUTE DYSTONIA

Background and Phenomenology

Acute dystonia occurs shortly after the introduction of DBA and occasionally after a dose increase. This reaction is particularly common with injectable high-potency DBA. There is usually a delay between the administration of the drug and the appearance of dystonia. About 50% of patients experience the first signs of dystonia within 48 hours of drug intake, and 90% show signs within 5 days of drug treatment (3). Acute dystonia is more likely to occur with typical DBA; however, the newer atypical drugs, even clozapine, are also associated with a small risk of inducing acute dystonia in the order of 2% to 3% (4).

Dystonic reactions are variable in location and can be painful. The usual manifestations are orofacial dystonia, back arching, and neck extension. Laryngospasm may occur and may be life-threatening (5). Repeated acute dystonic reactions, even with a single dose of DBA, have been observed but are uncommon. Rarely, other drugs as diverse as SSRIs, opioids, methylphenidate, rivastigmine, albendazole, gabapentin, cetirizine, foscarnet, quinine, as well as general anesthetics can cause acute dystonic reactions (6). A form of acute dystonic reaction appearing 3 to 10 days after starting dopamine-blocking agents is called the “Pisa syndrome”. It is characterized by tonic lateroflexion of the trunk (7). However, Pisa syndrome also may be seen as a manifestation of tardive dystonia (8).

An oculogyric crisis (OGC) is characterized by tonic conjugate ocular deviation that may last minutes to hours, and OGCs can occur in both acute and tardive dystonia (9). Rarely, recurrent OGCs despite withdrawal of neuroleptic drugs have also been reported (10). This may relate to the long half-life of these drugs or their metabolites.

Frequency and Risk Factors

The frequency of acute drug-induced dystonic reactions varies widely from 2.3% to 94% (11). Risk factors for dystonia include male gender, young age (under 30), high potency and high dose of neuroleptics used, familial predisposition, the type of underlying psychiatric illness, mental retardation, and a history of electroconvulsive therapy (11). There is a 2:1 risk of drug-induced dystonia in men compared to women. The same ratio holds true for young adults and children. Cocaine abuse may predispose to acute dystonic reactions (12). AIDS also has been associated with increased risk (13).

Mechanism

Two somewhat opposing hypotheses have been proposed as a mechanism of acute dystonia. One hypothesis is that dopaminergic hypofunction results in a relative overactivity of cholinergic mechanisms (14). This hypothesis is supported by a consistent amelioration of acute dystonia by anticholinergic drugs. Also supporting this hypothesis is the fact that acute dystonia can be suppressed in primates by the preadministration of dopaminergic drugs such as levodopa or apomorphine. Further support for this hypothesis comes from a marmoset model. In this model, acute administration of haloperidol results in a syndrome of excitation with sustained retrocollis, climbing upside down, biting the perch, repetitive turnings, and frequent backward movements. The dystonic movements lasted approximately 6 hours and were reduced but not completely extinguished by biperiden (0.1 mg/kg) (15).

The other hypothesis proposes that there is paradoxical dopaminergic hyperfunction induced by DBA through blockade of presynaptic dopamine receptors. Moreover, as the level of the DBA decreases, postsynaptic receptors are exposed to the natural release of dopamine from presynaptic terminals (16).

The possible contribution of other neurotransmitter systems, such as γ-aminobutyric acid (GABA), is unknown. The role of sigma receptors has been explored. It has been reported that the unilateral microinjection of sigma ligands into the red nucleus induces torticollis in rats (17). In animal models, the anticholinergic drug biperiden dose dependently ameliorates dystonia induced by two sigma ligands.

Management

Some evidence suggests that acute dystonic reactions may be prevented by the use of anticholinergic drugs (3). It is recommended by some that patients at high risk for acute dystonia (young patients, cocaine abusers, and AIDS patients) requiring DBA receive prophylactic anticholinergics.

Acute dystonia responds well to injectable anticholinergic drugs (18) or diphenhydramine (19). The response to anticholinergics is so consistent that if a patient with suspected DBA-induced acute dystonia fails to respond, clinicians should suspect phencyclidine (PCP)-induced dystonia (20) or some other etiology. Occasionally, diazepam has been used with success. At times, acute dystonic reactions, such as laryngeal dystonia, are severe enough to warrant life-saving measures (tracheostomy). The patients should be observed for recurrence of acute dystonia even if they receive no further DBA.

ACUTE AKATHISIA

Background and Phenomenology

The term “akathisia” (from Greek, literally “not to sit”) was introduced by Haskovec (21) in 1901, long before the development of neuroleptics. The phenomenon of akathisia is somewhat paradoxical given that the same drugs that are supposed to calm patients actually result in restlessness.

There are two aspects of akathisia: (1) a subjective report of restlessness or inner tension, particularly referable to the legs, with a consequent inability to maintain a posture for several minutes, and (2) the objective manifestations of restlessness in the form of movements of the limbs, a tendency to shift body position in the chair while sitting or marching while standing.

The temporal association with drug administration is an important feature in the diagnosis of akathisia. The most recognized form of akathisia usually starts within hours or days after the initiation or increase in DBA dosage or change in the type of DBA, and even a single exposure to the drug is sufficient for the diagnosis (22); acute akathisia usually starts within the first 2 weeks (23,24) and almost always within the first 6 weeks (25). The term “pseudoakathisia” applies to patients with the objective features of akathisia without subjective complaints (26). As such, pseudoakathisia is a true akathisia but with limited manifestations. Psychotic patients may have trouble reporting subjective feelings, and it may be difficult to distinguish akathisia from dyskinetic movements. In addition to acute akathisia, tardive akathisia (TA) is also recognized as a delayed side effect of DBA (see Classic Tardive Dyskinesia).

The incidence of akathisia ranges from 21% to 31% in psychotic patients treated with DBA (23,24). With the advent of newer atypical neuroleptics, the incidence of akathisia has clearly decreased, although not completely disappeared. As an example, a study of quetiapine across a dose range 75 to 750 mg/day found a prevalence of 0% to 2% versus 8% and 15% for patients randomized to placebo and haloperidol respectively (27). It is important to distinguish akathisia from psychotic agitation, which is more generalized and is often chaotic, disorganized, and even frenzied. Another disorder that must be distinguished from akathisia is restless legs syndrome (RLS). RLS can be interpreted as a focal akathisia mostly affecting legs but with a clear circadian pattern, being worse in the evening. In addition, unlike RLS, periodic limb movements of sleep are not a characteristic finding in patients with akathisia.

Mechanism

DBA causes akathisia by blocking dopamine receptors, especially D₂ receptors. This is supported by the observations that high-potency D₂ antagonists are more likely to cause akathisia, and akathisia is related to drug dose and may occur after the administration of a single dose. Two studies using positron emission tomography (PET) (28,29) demonstrated an association between D₂ occupancy in the striatum and the development of akathisia, with the latter authors suggesting a threshold between 74% and 82% D₂ receptor occupancy for the production of extrapyramidal effects, including akathisia. However, the D₂ antagonism hypothesis does not explain as to why cholinergic and β-adrenergic antagonists are effective in some cases of akathisia.

An alternative hypothesis was proposed by Marsden and Jenner (16), who proposed that DBA antagonism of mesocortical and mesolimbic dopaminergic projections leads to akathisia. This is supported by the observation that lesions of mesocortical dopaminergic neurons lead to increased locomotor activity in rodents.

Management

Two classes of drugs most commonly used in the treatment of acute akathisia are anticholinergic and antiadrenergic. Some literature supports the efficacy of these drugs in a proportion of patients; however, a recent Cochrane Database Review concluded that there was insufficient evidence to support their use (30,31).

Anticholinergics employed have included benztropine (dosage range 0.5–8 mg/day), trihexyphenidyl (1–15 mg/day), procyclidine (7.5–20 mg/day), biperiden (2–8 mg/day), and orphenadrine (100–400 mg/day). The optimal dosage should be titrated, starting with a small initial dose. Peripheral anticholinergic side effects (constipation, dry mouth, etc.) and central anticholinergic side effects (confusion, memory disturbance) should be monitored, especially in the elderly.

Of the antiadrenergic drugs, propranolol—a lipophilic, nonspecific β-blocker—has been used most extensively. The suggestion from the literature is that lower doses of propranolol are sufficient, with most researchers using doses on the order of 60 mg/day and rarely above 120 g/day. Propranolol seems to be well tolerated in this population, and the possible side effects of hypotension, bradycardia, sedation, and depression are not usually reported by appropriately selected individuals. Clonidine, an α₂-adrenergic agonist that reduces central noradrenergic activity, has been beneficial, but side effects such as sedation limit its practical utility.

A newer treatment option with a class of drugs—5-HT_2A antagonists—is rapidly emerging. A large trial of low-dose mirtazapine compared to propranolol showed that it was just as effective in treating neuroleptic-induced acute akathisia with a more convenient dosing and less side effects (32). Trazodone (Trz) is another agent demonstrating prominent serotonergic antagonistic properties. A recent placebo-controlled, double-blind, crossover study with Trz (100 mg/day) showed statistically significant clinical improvement in the symptoms of acute akathisia (33). These studies suggest a potentially important role for serotonin 2A receptor antagonists in the treatment paradigm of neuroleptic-induced acute akathisia.

DRUG-INDUCED PARKINSONISM

DIP may result from a variety of prescribed medications. The most common offending drugs are DBAs. Nonneuroleptic DBAs, such as metoclopramide, may also cause DIP (34). Table 33.2 and Table 33.3 list drugs that are likely to occasionally cause or exacerbate parkinsonism.

Prevalence of DIP

DIP is a common complication of antipsychotic drug use, occurring in 15% to 60% of patients treated with DBA (25). In one study, 51% of 95 patients referred to a geriatric medicine service for evaluation had parkinsonism associated with prescribed drugs (35). Another study found that in a general neurology practice, 56.8% of the 306 cases of parkinsonism were either induced by or aggravated by drugs (36). Frequently, these patients are misdiagnosed with idiopathic Parkinson’s disease (PD) and treated with dopaminergic drugs without benefit. In a community study, 18% of all cases initially thought to be PD were subsequently diagnosed as DIP (37).

Individual susceptibility to DIP has been postulated based on case reports indicating a familial predisposition to DIP and a bias toward the female gender (38–40). Family history also may be relevant, as Gartmann and colleagues (38) reported six patients with family history of PD who developed DIP on neuroleptics, whereas others without family history of PD tolerated neuroleptics without side effects. The dose and the potency of DBA therapy are of obvious importance. Although typical neuroleptics have been associated with a high rate of DIP, the second-generation (atypical) neuroleptics (e.g., risperidone, olanzapine, aripiprazole, etc.) are also associated with DIP, but may be at a lower rate and severity (41). However, in PD patients even atypical agents with the exception of clozapine and perhaps low-dose quetiapine can worsen parkinsonism and should be avoided.

Clinical Features

The symptoms of DIP are frequently indistinguishable from PD. Traditionally, DIP was characterized as symmetrical; however, asymmetry of signs and symptoms may occur in 30% of cases (42,43). Subgroups exist within DIP; some patients primarily experience bradykinesia, whereas others have tremor. In some patients, the symptoms are mixed. Postural reflexes may be impaired in some patients, and many patients have an abnormal gait. Festination is uncommon, and sudden transient freezing, a symptom of PD, is rare (43,44). The coexistence of a hyperkinetic movement disorder, such as orobuccolingual dyskinesia in the absence of levodopa treatment, will support a diagnosis of DIP rather than PD. Evidence is emerging that gait dysfunction and nonmotor features such as constipation, sexual dysfunction, and hyposmia may be more suggestive of a diagnosis of PD as compared to DIP (45).

Time Course

In patients who develop DIP, the condition typically develops between 2 weeks and 1 month following introduction of a neuroleptic or an increase in dose. It also tends to coincide with clinical improvement of schizophrenia (46). In one series, 50% to 70% of cases appeared within 1 month and 90% within 3 months (47).

The term “rabbit syndrome” refers to a perioral tremor that may develop in some patients at any time during neuroleptic treatment (48,49). The tremor has the typical characteristics of parkinsonian tremor and tends to respond to antiparkinsonian agents. The term adds nothing to our understanding of DIP and should be avoided.

DIP may resolve despite continuing DBA therapy, suggesting tolerance, but prospective studies to address this issue are lacking. Most cases resolve after discontinuing DBA; however, in some patients, symptoms persist for months (50). In some cases, underlying PD may be unmasked by DBAs, and in such individuals, the symptoms persist even after the discontinuation of DBAs.

The Problem of Underrecognition

Early in the era of DBA therapy, it was thought that extrapyramidal side effects and antipsychotic efficacy were tightly linked, and DIP was ignored or thought to be essential for efficacy. In one series, residents recognized DIP in less than half the patients afflicted with it (51). Metoclopramide-induced DIP is also frequently unrecognized, as suggested by Miller (52).

Metoclopramide-induced parkinsonism is not uncommon. This nonneuroleptic DBA is frequently prescribed for the treatment of gastroesophageal reflux and gastroparesis. We have observed that metoclopramide is a frequent offender in elderly diabetic patients with renal insufficiency (34). Parkinsonism typically develops when these patients are taking a prescribed dose of 40 mg/day. Metoclopramide is cleared by the kidneys, and in renal failure, the dose should be reduced by 50%.

DIP DUE TO VESTIBULAR SEDATIVES

In countries other than the United States, a common cause of DIP is the use of the vestibular sedatives cinnarizine and its derivative, flunarizine (53,54). In one study of 172 cases of DIP over 15 years, 74 cases were due to cinnarizine (53). Cinnarizine-induced parkinsonism is more common in women, as is the case with neuroleptic-induced DIP. Complete recovery usually occurs after drug withdrawal, but in some cases, signs persist (55). Drug dose is important, and the risk appears to be low in those taking less than 150 mg of cinnarizine. The mechanism of DIP induced by these drugs is unknown and may involve presynaptic dopamine depletion, postsynaptic dopamine-receptor blockade, and effects on nondopaminergic neurons. These drugs also have marked antihistaminergic and calcium channel–blocking activity.

Pathophysiology

Superficially, DIP is the most easily understood side effect of DBAs and relates to their D₂ receptor–blocking activity (16). DIP appears to develop in almost everyone given high doses of high-potency DBA, thereby achieving concentrations that would block about 80% of central dopamine receptors (56). However, the reasons why some individuals develop parkinsonism in the usual therapeutic dose range are unclear. Another unexplained aspect is the delay between the pharmacologic blockade of the D₂ receptors and the onset of DIP. While DBAs block dopamine receptors within minutes to hours, DIP typically appears many days or weeks following drug exposure. This suggests that there is a complex relationship between exposure to neuroleptic drugs and the development of DIP.

Risk factors for DIP include the potency and dose of the neuroleptic and individual susceptibility. A high ratio between serotonin (5HT₂) and D₂ dopamine-receptor antagonism will produce less in the way of extrapyramidal effects, including DIP (57). Also, a rapid dissociation of the drug from the receptor may have a role to play in reducing the motor side effects (58).

Individual risk factors include older age, female sex, and the presence of cerebral atrophy. Genetic susceptibility to DIP has been postulated based on case reports indicating a familial predisposition to DIP. Genetic differences in drug metabolism are postulated to be important in some patients (59,60).

It has been suggested that DIP may merely be unmasking subclinical PD. DIP in treated patients is more prevalent than PD in the general population. However, the frequency of older patients with incidental Lewy bodies at autopsy is approximately 15 times the frequency of patients with clinically apparent PD (61). Therefore, DIP may represent an unmasking of subclinical PD. The evidence for DBA unmasking PD comes from both clinical (62) and pathologic observations (63). In some cases of DIP, the condition may persist after discontinuing DBA, and some patients may go on to develop PD. A retrospective study done at the Mayo Clinic in Rochester, Minnesota, showed that IPD emerged over time in 2 out of 24 patients (8%) followed for DIP. A comparison with the expected number of PD patients in the general population yielded a relative risk of 24.3 (64). Rajput et al. (63) reported pathologic evidence of nigral Lewy body disease in two patients who developed DIP while on neuroleptic treatment but had a complete recovery on withdrawal of neuroleptics. After death from unrelated causes, autopsy revealed slight-to-moderate loss of nigral dopaminergic cells and Lewy bodies.

More recently, functional neuroimaging has been utilized to explore this problem. Dopamine transporter (DAT) imaging is now commercially available to study presynaptic nigrostriatal deficit which is a hallmark of PD. A DAT study can be helpful in differentiating true nigrostriatal dysfunction seen in PD from DIP. A normal DAT study is supportive of DIP and may also be suggestive of recovery upon stopping the offending agent (65). For a detailed discussion on DAT imaging in movement disorders, please see Reference 65.

Management

Prevention of DIP should be the goal. DBA should be administered only when absolutely necessary, and atypical antipsychotics should be used when possible because of their comparatively lower propensity to cause extrapyramidal syndrome (EPS), including DIP. The use of anticholinergic drugs in prophylaxis is debatable. Anticholinergics may worsen psychiatric problems and cause confusion and memory difficulties. A reasonable approach may be to treat prophylactically patients who are at a higher risk (e.g., AIDS patients) with anticholinergics. This would be particularly useful if a high-dose, high-potency DBA is used. However, there is little prospective data to support this approach.

The treatment of clinically manifest DIP is difficult. An obvious choice is to reevaluate the need for DBA and to withdraw the drug whenever possible. This is done with a risk of recurrence of psychosis. In most cases, the condition is reversible once the offending agent is withdrawn. In cases where DBA cannot be withdrawn, substitution with an atypical agent should be attempted. Clozapine, an atypical neuroleptic, has a low acute extrapyramidal symptom profile, but it may cause agranulocytosis in approximately 1% of patients. The risk of DIP is also low with quetiapine, but higher doses of olanzapine, risperidone, and aripiprazole may cause DIP.

Mild DIP may be left untreated. Where necessary, the symptoms can be managed with anticholinergic agents, antihistaminic agents, or amantadine. Amantadine is a useful drug and may be superior to anticholinergics, as shown in two studies (66,67).

There are limited data on the use of levodopa or dopamine agonists in the treatment of DIP. In one study, dopaminergic drugs seemed to worsen psychosis (68), but this approach has not been systematically studied.

TARDIVE MOVEMENT DISORDERS DUE TO DBAS

DIP may occur in a delayed fashion, but the problems more frequently encountered in a tardive fashion include classic tardive dyskinesia (TD) and its variants and TA.

Classic Tardive Dyskinesia

TD has been defined by the American Psychiatric Association Task Force as an abnormal involuntary movement following a minimum of 3 months of neuroleptic treatment in a patient with no other identifiable etiology for movement disorders (69). DSM-IV criteria, however, specify that the duration of exposure to neuroleptics may be only 1 month in individuals aged 60 and older (70).

Many of the writers in the preneuroleptic era described spontaneous abnormal movements in patients with schizophrenia. However, many motor phenomena seen in patients with schizophrenia are highly idiosyncratic and tend not to appear qualitatively different from normal movements in the way that dyskinesia, dystonia, and tremor do. Most of the reports from the preneuroleptic era suggest the presence of stereotypies and posturing (71). Some patients had orofacial dyskinesia, but it is unclear if they had other organic brain diseases, such as neurosyphilis. Patients with schizophrenia treated with antipsychotic agents exhibit relatively delineated, recognized syndromes of abnormal involuntary movement, which are abnormal in their nature and appear unrelated to the normal range of expressive gesture. Although motor problems such as tardive orobuccolingual dyskinesia are generally exacerbated by increased arousal and anxiety, they seem otherwise not to be directly driven by aberrant thought content, affect, or perception.

Epidemiology and risk factors. In a review of 56 studies that spanned from 1959 to 1979, Kane and Smith (72) reported point prevalence of TD ranging from 0.5% to 65%, with an average point prevalence of 20%. In a more recent review from 2008, an analysis of 12 studies since 2004 (n = 28,051, followed for 463,925 person-years) revealed that the annualized TD incidence was 3.9% for second-generation antipsychotics and 5.5% for first-generation antipsychotics (73). The clinical significance of both above figures is limited as they were derived from studies that differed in assessment criteria, methodology, and population characteristics. However, overall, the incidence of TD seems to have diminished with the advent of the modern antipsychotic drug era as compared to the older neuroleptics. But TD has not disappeared with the use of newer, more expensive antipsychotics. In fact, at higher doses, some of the newer atypical antipsychotics carry a substantially high risk similar to the older neuroleptics (74). In addition, the atypical antipsychotics are also associated with weight gain, diabetes, and increased cardiovascular risk. So, the treatment has to be individualized and clinical awareness and continued vigilance are necessary.

Risk factors for development of TD most consistently defined by various epidemiologic studies include affective disorder, old age, female gender, total cumulative drug exposure, diabetes, alcohol and cocaine abuse, persistence of neuroleptic drug use after the development of TD, and a history of electroconvulsive treatment (ECT). Advancing age is the most consistently established risk factor for TD, and there appears to be a linear correlation between age and both the prevalence and severity of TD. A number of studies have indicated a higher risk for women to develop TD (75). There is some evidence that diabetes increases the risk of TD. Woerner et al. (76) reported a risk ratio of 2.3 for diabetics exposed to neuroleptics compared to similarly treated nondiabetics, with the risk greater in aged diabetics.

Clinical features. Classic TD manifests as repetitive, coordinated, seemingly purposeful movements affecting mainly the orofacial area. Some prefer to use the term “tardive stereotypy” to describe these movements. True chorea may occur but usually in the setting of withdrawal dyskinesia (see below).

Many patients with TD exhibit a combination of movement disorders. In general, the presence of multiple movement disorders in a patient should alert the physician to the possibility of a DIMD. Most frequently, stereotypy of classic TD is combined with choreic movements of the hands, fingers, arms, and feet or with dystonia. The diaphragm and chest muscles are frequently involved in respiratory dyskinesias, resulting in noisy and irregular breathing, leading to pulmonary investigations. The abdominal and pelvic muscles may also be involved, producing truncal or pelvic movements known as “copulatory dyskinesia”.

Orobuccolingual dyskinesia may occur in other clinical situations (Table 33.4) that should be considered in the differential diagnosis before blaming the DBA.

Withdrawal and Covert Dyskinesias

Dyskinesias may first appear when the DBA is discontinued or the dosage is reduced. This form of dyskinesia has been called “withdrawal or covert dyskinesia” (77). Withdrawal dyskinesia disappears within 3 months of drug withdrawal, whereas covert dyskinesia becomes apparent upon reduction of neuroleptic therapy and persists for longer periods. This distinction is arbitrary, because some of the covert dyskinesias will disappear after prolonged follow-up.

Tardive Tourettism

This is a rare tardive syndrome with a review of the literature revealing only 41 published cases until 2011(78). Tardive tourettism has been recognized as a complication not only associated with neuroleptics but with other drugs such as anticonvulsants, antidepressants, and stimulants (78). Patients may develop abnormal movements and vocalizations following chronic neuroleptic treatment. Further, the symptoms exhibited by the patients are indistinguishable from those of classic Tourette’s syndrome. It is a requirement for the diagnosis of tardive tourettism that these patients had no tics as children. The neuropharmacology underlying tardive tourettism likely parallels that of TD and idiopathic Tourette’s syndrome (79).