Clinical manifestations

Presentation and course

Tardive dyskinesia can manifest as early as three months after continuous treatment with a dopamine-blocking medication (or one month in elderly patients); however, in the majority of cases it occurs one to two years after the beginning of treatment (04; 152). Reduction of the dose or sudden discontinuation of the offending drug can also cause tardive syndrome within four weeks for oral or eight weeks for depot medication. Tardive dyskinesia has insidious onset, with the symptoms usually evolving over a few weeks.

Tardive syndrome is a group of phenomenologically different involuntary movements manifesting as orobuccolingual stereotypy, dystonia, akathisia, tics, tremor, myoclonus, chorea, or a combination of a few types of abnormal movements (43; 127). A stereotypy is an involuntary, patterned, repetitive, continuous, coordinated, purposeless, or ritualistic movement.

Classic tardive dyskinesia. Classic tardive dyskinesia manifests as orobuccolingual stereotypy involving the tongue, lips, and jaw and producing repetitive licking, lip smacking, and chewing movements (62; 152). There may be torsional movements of the tongue inside of the mouth, quick tongue protrusions (flycatcher tongue), or pushing of the tongue against the inner cheek creating a bulge (bonbon sign). Facial movements are typically limited to the lower face, less frequently involving the forehead.

In contrast to idiopathic craniocervical dystonia (which consists of blepharospasm and cervical dystonia), lip pursing, lip smacking, and facial grimacing are more likely to be found associated with tardive dyskinesia (143). Other characteristically (though less commonly) affected areas include the respiratory muscles, causing erratic breathing, stridor, and respiratory noises such as humming or moaning (162).

Besides orobuccolingual stereotypy of classic tardive dyskinesia, some patients have limb and trunk stereotypy manifesting as repetitive flexion and extension of the fingers and toes, tapping of the foot while sitting, flexion and extension of the thighs when supine, rocking, swaying trunk movements, and thrusting pelvic movements (38; 135). The repetitive, patterned, seemingly purposeful nature of the movements in tardive dyskinesia often resembling ritualistic gestures of mannerism has led to their designation as "stereotypies" (140). This can be contrasted to the choreiform movements seen in Huntington disease wherein the movements are random and unpredictable (137).

Involuntary movements in tardive dyskinesia are often suppressible by voluntary actions involving the affected part. Many patients are unaware of their tardive dyskinesia (24), but others with severe orolingual dyskinesias experience difficulty with eating and talking and have dental damage, along with embarrassment. Generalized dyskinesias may lead to respiratory difficulties, dysphagia, and gait and balance impairment leading to falls. Changes in gait can be described as “dance-like” (repetitive short steps on toes followed by long stride) or “duck-like” (wide-based gait with short stride length and mild steppage features), broad-based, spastic, or generally unsteady (82).

Tardive dystonia typically manifests as retrocollis with truncal backward arching or scoliosis, internal rotation and adduction of the arms, extension of the elbows, and flexion of the wrists (14; 152). A tardive Pisa syndrome, a lateral flexion of the trunk, has been reported in a young patient treated with aripiprazole (30). Focal tardive dystonia can affect facial muscles, tongue, jaw, or cervical muscles.

Tardive akathisia refers to an inner sense of restlessness and inability to sit still (92). Patients may have associated repetitive movements and vocalizations resembling tardive stereotypy such as body rocking, crossing and uncrossing of the legs, rubbing the scalp, marching in place, and moaning. However, tardive akathisia has a sensory component of internal restlessness whereas tardive stereotypy does not. In addition to a generalized sense of restlessness, there may be focal symptoms described as a burning sensation, most commonly in the oral and genital areas (40). Indeed, sensory aspects of tardive dyskinesia, which include tardive akathisia, are increasingly recognized in many movement disorders (120).

Other tardive syndromes include tardive tics, tardive myoclonus, tardive chorea, and tardive tremor phenomenologically resembling hyperkinetic movement disorders not caused by the dopamine-blocking agents (143; 148; 41; 152).

Withdrawal emergent syndrome is a rare condition described in children following sudden discontinuation a dopamine-blocking agent. It manifests as generalized chorea, usually sparing the face, days or weeks after withdrawal of the offending drug and self-limiting within a few weeks (104).

Prognosis and complications

In the majority of patients, tardive dyskinesia persists for years or decades even following discontinuation of the offending medication. In many cases it can be irreversible. Rate of remission after eliminating a dopamine-blocking medication (or in rare cases even while continuing treatment) varies widely in the literature, from 2.5% per year for patients still on a suspected offending agent up to 33% two years after stopping a drug (70; 146). Likelihood of remission increases if dopamine receptor antagonists are withheld earlier in the course of tardive dyskinesia, whereas a longer exposure to the drugs is associated with worse prognosis. The majority of remissions occur within one year of drug discontinuation, but persistence of movements for up to 5 years may still be followed by complete spontaneous resolution (77). Once established, tardive dyskinesia does not tend to continue to progress in severity; however, the course can be waxing and waning.

Clinical vignette

A 79-year-old woman with chronic low back pain complained to her physician of insomnia. She was treated with haloperidol, amitriptyline, and zolpidem for about five years. In the past two months, she began having involuntary movements of her mouth, which progressed from the onset until her speech became dysarthric. Examination at that time showed “constant mouth movements,” and the haloperidol was discontinued. She returned to the original physician three months later, and haloperidol was restarted for unclear reasons. She soon developed “shaking” of her head and feet and dysphagia. The haloperidol was stopped again after one month. Treatment with anticholinergics, botulinum toxin injections, and diazepam yielded little or no benefit.

On examination 15 months after onset of the involuntary movements, there were continuous orofacial dyskinesias. Alternating protrusion and retraction of the tongue at rest were diminished by active tongue protrusion. Chewing movements of the jaws were abolished by voluntary mouth opening. There was intermittent gentle flexion and extension of the neck, and mild rocking of the trunk when standing. There were dyskinesias of the feet (greater on the right) and slight dyskinesias of the left upper limb as well.

Biological basis

Etiology and pathogenesis

Tardive dyskinesia occurs as a result of chronic exposure to dopamine receptor-blocking agents, drugs used mainly to treat psychiatric or gastrointestinal disorders (152). Due to their high D2 receptor occupancy, the first generation, or typical, neuroleptics carry greater risk than second- or third-generation neuroleptics, which are often referred to as atypical (31). The latter medications have low D2 receptor occupancy, and hierarchical risk of causing tardive dyskinesia from least to greatest is as follows. Clozapine has a lower risk than quetiapine, which has a lower risk than olanzapine and ziprasidone; risperidone carries a lower risk than the last two at low doses, but with increased risk at higher doses (147). Despite a low prevalence of extrapyramidal symptoms, newer atypical neuroleptics, such as aripiprazole, have been reported to cause tardive dyskinesia (98; 123; 132). Essentially all atypical neuroleptics have been documented to cause tardive dyskinesia (75).

Although historically neuroleptics were to blame for most cases of tardive dyskinesia, metoclopramide is a common reported cause (119). Since approval by the Food and Drug Administration in 1979, metoclopramide has become one of the most widely used drugs for gastrointestinal motility disorders. One of the major advantages of metoclopramide is that it acts centrally as an antiemetic, antagonizing dopamine receptors in the chemoreceptor trigger zone, and peripherally as a prokinetic drug. Several studies have indicated that metoclopramide usage often continues despite the development of tardive dyskinesia (44). In a study, the average duration of exposure to metoclopramide before onset of hyperkinesia was 12 months and therapy was continued for an average of 6 months after symptom onset (107). Metoclopramide and other antiemetics with dopamine receptor blocking properties such as prochlorperazine are also used to treat migraine, and carry the risk of causing tardive dyskinesia (158).

Rare cases of tardive dyskinesia-like syndrome have been reported as a result of long-term treatment with antidepressants, including selective serotonin and norepinephrine reuptake inhibitors and tricyclics, or lithium (48; 22; 98; 02). Calcium channel blockers (flunarizine, cinnarizine) are also associated with tardive dyskinesia in some countries (32). The topic of tardive dyskinesia caused by non-dopamine blocking drugs remains open for debate. Some clinicians and scientists believe that non-dopamine blocking drugs are unlikely causative agents in tardive dyskinesia but they rather provide a “priming” effect that can unmask or exacerbate dyskinesia in patients who have been exposed to dopamine-blocking drugs, even in the remote past (29).

The pathophysiology of tardive dyskinesia remains poorly understood. A number of lines of evidence point to the involvement of dopaminergic systems in the pathogenesis of tardive dyskinesia (152; 12). Experimental animals chronically exposed to antipsychotics demonstrate behavioral supersensitivity to dopamine agonists and increased striatal dopamine binding (78). In a rat model of haloperidol-induced oral dyskinesias, the severity of dyskinesia was correlated with degenerative changes in substantia nigra cells (11). The long-term use of antipsychotics has been associated with a 3-fold reduction in density of striatal dopamine terminals in patients with tardive dyskinesia, but not in individuals who had no tardive dyskinesia as measured by imaging of dopamine transporters (131). This increased loss of striatal terminals was attributed to accumulation of antipsychotics in the neuromelanin of the substantia nigra. PET studies in humans implicate D2 receptor upregulation in the generation of tardive dyskinesia (136).

The weight of evidence favors the notion that chronic blockade of D2 and possibly D3 dopamine receptors leads to receptor upregulation and increased receptor sensitivity. Primate studies indicated that upregulated striatal D3 rather than D2 receptors correlate with the development of tardive dyskinesia (95; 56). Yet it is not clear how this leads to the development of a chronic hyperkinetic movement disorder not remitting even after discontinuation of a causative medication, or why only a minority of people with similar drug exposure develop the syndrome. One widely studied mechanism invoked to explain these gaps in our knowledge is the ubiquitous “oxidative stress” hypothesis (81). Evidence of a role for oxidative stress in the pathogenesis of tardive dyskinesia comes in several forms. Tardive dyskinesia patients have been shown to have elevated plasma activity of manganese superoxide dismutase compared to schizophrenics without tardive dyskinesia and normal controls, and the level of activity is correlated with the clinical severity of dyskinesia (168). The same investigators have found that a polymorphism of the superoxide dismutase gene interacts with a polymorphism of the dopamine D3 receptor gene to produce an increased susceptibility to development of tardive dyskinesia (169). A study demonstrated lower levels of manganese superoxide dismutase in schizophrenic patients with tardive dyskinesia compared to those without (160). A rodent study found that haloperidol led to an increase in vacuous chewing movements and an increase in S100B (a potential biomarker of structural brain damage) in the prefrontal cortex, striatum, substantia nigra, and globus pallidum (08). Treatment with antioxidants gingko biloba leaf extract (EGb761) or vitamin E was associated with a decrease in abnormal movements and a decrease in S100B expression. However, a study of the effect of another antioxidant lipoic acid on haloperidol-induced vacuous chewing movements in rats reported only nonsignificant decreases of movement on lower doses of lipoic acid, but significant worsening of dyskinesia on higher doses of lipoic acid (91).

Alternative theories regarding the pathophysiology of tardive dyskinesia exist (21; 12). One hypothesis relates tardive dyskinesia to a GABAergic hypofunction and degeneration of GABAergic interneurons regulating balance between direct and indirect basal ganglia pathways (49). Opioid pathways and excessive production of neurotensin may be involved (99; 100). A preliminary report on proton magnetic resonance spectroscopy in patients with and without tardive dyskinesia suggests that the latter is associated with accentuation of metabolic abnormalities, reflected by increased choline/creatinine ratios, in the basal ganglia found in schizophrenia (10). Several animal and human studies provide evidence that chronic use of antipsychotics may result in structural changes in brain with neuronal loss and gliosis in basal ganglia (76; 58). Structural changes in cerebral cortex with decreased gray matter volume in cuneus and lingual gyri on MRI were reported in schizophrenic subjects with tardive dyskinesia versus schizophrenics without tardive dyskinesia or healthy controls (163). Degree of gray matter volume loss in those two areas correlated with severity of tardive dyskinesia on AIMS (Abnormal Involuntary Movement Scale). Some studies have shown that serum levels of brain-derived neurotrophic factor (BDNF), which protects against neuronal damage in the nigrostriatal dopaminergic system, are lower in patients with tardive dyskinesia than those without (144; 167). Immune dysregulation is a proposed mechanism underlying tardive dyskinesia. Alterations of the levels of IL-2, IL-6, and IL-8 were associated with schizophrenia patients with tardive dyskinesia compared to those without and normal controls (07). The role that these changes play in the development of tardive dyskinesia is unclear. A maladaptive NMDA-medicated synaptic plasticity hypothesis suggests that dopamine-receptor hypersensitivity and neurodegenerative changes due to oxidative stress caused by chronic exposure to dopamine-blocking medications can alter synaptic plasticity and lead to a formation of miscoded motor programs and abnormal hyperkinetic movements (149).

The majority of patients on similar drug regimens do not develop tardive dyskinesia, and those affected vary considerably in the severity of the movements. Unaffected siblings of schizophrenic patients with tardive dyskinesia have a higher prevalence of spontaneous dyskinesias than patients without siblings with tardive dyskinesia (101). Thus, individual susceptibility plays a large role, and this may have a genetic basis (81). Gene candidates include those coding for dopamine D2 and D3 receptors, serotonin 5HT2A receptors, catechol-O-methyltransferase (COMT) enzyme, manganese superoxide dismutase, and CYP2D6. Findings from genetic studies have reported on certain polymorphisms such as CYP2D6*10, DRD2 Taq1A, DRD3 Ser9Gly, HTR2A T102C, and MnSOD Ala9Val, although these results remain controversial (84). Furthermore, genome-wide association studies have identified genes associated with tardive dyskinesia such as rs7669317 on 4q24, GLI2 gene, GABA pathway genes, and HSPG2 gene (84). Several studies have implicated polymorphisms in exon 1 of the dopamine D3 receptor gene and the 5-HT2A receptor gene (142; 87). One report supports the role of COMT and dopamine D3 receptor genes in a population of schizophrenics in North India (139). However, a review of 11 case-control studies did not find significant association between COMT Val158Met gene polymorphism and risk of tardive dyskinesia (94). In Caucasians, but not African Americans, the APOE2 allele may increase the risk of developing tardive dyskinesia (51). In a study of 491 Caucasian patients with schizophrenia, a significant association was found between the functional mutation N251S-polymorphism of the PIP5K2A gene (rs10828317) and tardive dyskinesia (37). A mutation in this gene that codes for phosphatidylinositol-4-phosphate-5-kinase type IIa suggests a role of this kinase in tardive dyskinesia and neurotoxicity. In a genome-wide association study of 505 patients with paranoid schizophrenia, 95 patients had tardive dyskinesia, and 503 healthy controls revealed that paranoid schizophrenia and orofacial dyskinesia were associated with the genomic loci on chromosomes 3p22.2, 8q21.13, and 13q14.2 (89). A review of 43 animal studies of tardive dyskinesia pathogenesis identified eight genes that also reached statistical significance in at least one clinical study (150). Gene candidates include variants of SLC18A2 gene encoding for vesicular monoamine transporter 2 protein (VMAT2); dopamine receptor genes DRD1, rs4532, and DRD3 rs6280 polymorphisms; serotonin receptor gene HTR2A gene variants; NMDA glutamate receptor genes GRIN2A and GRIN2B variants; oxidative stress-related gene SOD2 rs4880 variant; and CYP2D6 variants leading to decreased metabolizing capacity of the enzyme). Pharmacogenomic studies have produced conflicting results regarding the association of genetic polymorphisms with tardive dyskinesia, and they deserve further study (23; 93).

Epidemiology

Reported incidence and prevalence rates of tardive dyskinesia vary widely, depending on the population, study design, and diagnostic criteria used. A review of 56 prevalence studies found that an average of 20% of 34,555 neuroleptic-treated patients demonstrated involuntary movements compatible with tardive dyskinesia (69). The cumulative incidence increases for the first five years of treatment; the prevalence subsequently tends to remain stable as remission rates equalize with incidence rates (47). A prospective study of schizophrenics suffering their first episode of psychosis and treated with haloperidol found an incidence of tardive dyskinesia of 12.3% in the first 12 months (116). Meta-analysis of 57 randomized controlled studies including 32 first-generation antipsychotics arm and 86 second-generation antipsychotics arm reported annualized incidence of tardive dyskinesia of 6.5% in the former arm versus 2.6% in the latter arm (19). A systematic review of studies from 1957 to 2015 estimated that the incidence of probable tardive dyskinesia was 23% for patients treated with first-generation typical neuroleptics versus 7% for patients using second-generation neuroleptics at one year (112). The prevalence of tardive dyskinesia in patients treated with metoclopramide ranges from 1% to 10% in different studies (125).

A retrospective study of 123 patients treated with antipsychotics for more than six months in an outpatient psychiatric clinic (96.8% had schizophrenia, 66.7% were women, mean age 45.6 years) reported 35 patients with at least one episode of a tardive syndrome (85). The prevalence of the various phenomenologies of tardive syndromes included 21.1% tardive dyskinesia, 12.5% tardive dystonia, 2.4% tardive tremor, and 2.4% tardive akathisia. The term “tardive dyskinesia” in this study was presumably used to describe the classic orofacial stereotypy.

Prevalence rates for tardive dyskinesia increase with age due to a higher incidence and a lower remission rate in older patients. Older patients may have difficulty metabolizing the medication and may be more susceptible to side effects (97). Older women appear to be the most vulnerable group (110). Younger men have a greater risk of developing tardive dystonia. The prevalence in children and adolescents is distinctly lower, but tardive dyskinesia and withdrawal emergence syndrome are increasingly recognized in children and even infants (27; 103; 104). In a 1-year, multicenter observational study in a pediatric population of 265 subjects exposed to antipsychotics, 5.8% developed tardive dyskinesia at follow-up (45). In this study, younger age, longer exposure to antipsychotics, and a history of psychotic symptoms were associated with an increased risk of tardive dyskinesia. Other proposed risk factors in adults include a greater total drug exposure, preexisting drug-induced parkinsonism, treatment for an affective rather than a psychotic disorder, alcoholism, previous brain damage, and smoking (65; 25). A study of 77,022 adult inpatient admissions for mood disorders and schizophrenia with secondary diagnosis of tardive dyskinesia and age-matched controls without tardive dyskinesia found that most tardive dyskinesia patients were older (50-64 years; 40%), with nearly equal proportions of men and women (121). African Americans had 2-fold higher odds of tardive dyskinesia. Higher likelihood for cardiometabolic comorbidities—obesity (OR 1.61, 95% CI 1.481-1.756), hypertension (OR 1.78, 95% CI 1.635-1.930), and diabetes (OR 1.54, 95% CI 1.414-1.680)—compared to controls were reported.

Prevention

The only absolute method of prevention is to avoid using causative drugs. Therefore, physicians must consider alternative medications before instituting treatment, particularly in patients with nonpsychotic disorders. In patients for whom dopamine receptor antagonists are the best therapy available, a major goal should be to minimize the dose and duration of exposure to the drug as possible. Physicians should also reevaluate the need for ongoing treatment at every opportunity. Another approach is to use an atypical antipsychotic such as quetiapine or clozapine. The American Psychiatric Association Practice Guideline recommends assessment of involuntary movements with a structured scale such as the Abnormal Involuntary Movement Scale (AIMS) at baseline and every 6 months for high risk and every year for low risk patients with schizophrenia on neuroleptic medication (06). Prophylactic administration of vitamin E has been advocated but without strong evidence supporting its use (46). Small, limited trials report vitamin E may protect against worsening tardive dyskinesias without evidence of improvement in symptoms (138).

Differential diagnosis

Confusing conditions

The differential diagnosis of the orolingual movements of tardive dyskinesia includes spontaneous buccal and lingual dyskinesias of the elderly (in the appropriate age group) and edentulous dyskinesias (in the setting of compromised dentition).

Generalized movements similar to those of tardive dyskinesia may be seen in Huntington disease and other genetic chorea, Sydenham chorea, cerebral infarcts, hyperthyroidism, systemic lupus erythematosus, antiphospholipid antibody syndrome, and other conditions causing chorea and athetosis. In general, chorea exhibited in these conditions is random in nature, whereas movements in tardive dyskinesia are repetitive and stereotypic. It is also important to remember that patients with schizophrenia may show stereotypic movements without ever being exposed to antidopaminergic drugs (157).

Associated or underlying disorders

Patients with tardive dyskinesia might present with other drug-induced movement disorders such as drug-induced parkinsonism or tremor. Unlike tardive dyskinesia, drug-induced parkinsonism and tremor usually improve or resolve within a few weeks or months after discontinuation of a causative drug or even after reduction of the medication dose.

Diagnostic workup

Tardive dyskinesia is a purely clinical diagnosis, and no laboratory tests support it. Diagnostic evaluations are usually performed only to rule out other possible causes of abnormal movements. These might include brain imaging and appropriate laboratory studies. The most important step in the diagnosis of tardive dyskinesia is a careful medication history. Patients are often unaware of the types of medications they have received in the past. A diligent search includes solicitation of information from family members, prior treating physicians, and medical and pharmacy records.

Management

The development of tardive dyskinesia should prompt a reevaluation of the necessity for the causative agent. This is particularly true when a dopamine receptor-blocking drug is being used for a nonpsychotic indication. An attempt should be made to reduce the dose of the offending drug if it cannot be withdrawn entirely. Many patients will not improve despite discontinuation of their medication, whereas others may have spontaneous resolution of their movements while continuing to take the responsible drug at the same dose. However, it is widely believed that a patient's best chance for a remission lies with cessation of exposure to the offending agent. Before embarking on this course, it is important with psychiatric patients to weigh the potential benefit of improving tardive dyskinesia, often unnoticed by patients, against the risk of relapse of psychotic symptoms (102). For patients with mild tardive dyskinesia who are not distressed by the symptoms, slow tapering off an offending drug followed by watchful observation may be a reasonable management option. Those who cannot discontinue taking antipsychotic medication can be candidates for therapy with clozapine or quetiapine, atypical dopamine receptor antagonists that rarely cause tardive dyskinesia. A study of 35 patients with schizophrenia or bipolar disorder who switched to clozapine from another antipsychotic drug after developing tardive dyskinesia reported remission of symptoms in 65.7% of patients, especially those of younger age and with milder tardive dyskinesia (83). Meta-analysis of 48 studies (including 13 clinical trials and 35 case reports) reported significant improvement of tardive dyskinesia symptoms in 86.7% of clozapine-treated patients with schizophrenia spectrum disorders (average dose of clozapine = 355 mg/day) and 93% of patients with other psychiatric disorders (average dose of clozapine = 152.5 mg/day) (159). Although potentially life-threatening, agranulocytosis occurs in about 0.8% to 3% of patients treated with clozapine; therefore, close monitoring of white blood cell counts is required (80; 03).

A wide variety of pharmacologic agents has been reported to lessen the involuntary movements, but may produce only modest benefit (145). Increasing the dose of a dopamine receptor antagonist can acutely improve tardive dyskinesia, but this maneuver may contribute to a more firmly entrenched movement disorder over the long term. Therefore, it is useful only if immediate control of the severe dyskinesia is imperative. An evidence-based guideline for the treatment of tardive syndromes was published by the American Academy of Neurology in 2013 (15). Based on the available evidence, clonazepam and ginkgo biloba were found to probably improve tardive syndromes (level B), and amantadine and tetrabenazine might also be effective (level C).

In small scale studies, tetrabenazine, a vesicular monoamine transporter 2 (VMAT2) inhibitor (which in 2008 was FDA approved for treating Huntington disease in the United States), has been shown to be an effective medication for the treatment of tardive dyskinesia (63; 115; 72; 35; 64; 66; 61). However, dopamine depleters may cause drug-induced parkinsonism, akathisia, depression, and excessive daytime drowsiness. There has only been one published report of tardive dyskinesia caused by tetrabenazine (90). The author describes oral-buccal-lingual movements and chorea of the hands and feet in a 55-year-old man treated with tetrabenazine 12.5 mg twice daily for cervical dystonia. The movements resolved after discontinuation of the medication. Although tetrabenazine has been implicated in this man’s movement disorder, it should be noted that he was treated for his underlying depression with paroxetine for 10 years, and later with desvenlafaxine until the onset of the dyskinesia. It is, therefore, possible that the selective serotonin reuptake inhibitors have caused or contributed to the tardive dyskinesia (86). Selective VMAT2 inhibitors (deutetrabenazine and valbenazine) were approved by the FDA in 2017 for the treatment of tardive dyskinesia, and they exhibited at least the same effectiveness as tetrabenazine, but with a better side effect profile (61; 130). Bhidayasiri and colleagues consider deutetrabenazine and valbenazine effective treatments of tardive dyskinesia (level A) that must be recommended to patients (61; 16; 111). Valbenazine, a purified (+)-α-isomer of tetrabenazine that also acts as a VMAT2 inhibitor, was evaluated in a 6-week, randomized, double-blind, placebo-controlled study involving 102 subjects with moderate to severe tardive dyskinesia (113). There was a significant reduction in the AIMS scores with NBI-98854 compared to placebo (mean change -3.6 vs. -1.1, p < 0.0005). The most common side effects were fatigue and headache. A phase 3 (KINECT 3) study concluded that a once-daily administration of 40 or 80 mg of valbenazine provided significant improvement in tardive dyskinesia versus placebo (mean AIMS scores change from baseline -3.2 vs. -0.1 for 80 mg dose and -1.9 vs. -0.1 for 40 mg dose) (55). Medication was generally well tolerated, even in patients taking a wide range of concomitant medications (including antipsychotic agents). A phase 3, one year, open-label trial of valbenazine (KINECT 4) reported similar data on the safety and efficacy of the medication (96). Deutetrabenazine, a reversible VMAT2 inhibitor, is a deuterated formulation of tetrabenazine. The percentage of responders in the fixed-dose phase 3 study was 46% for deutetrabenazine versus 26% for placebo on the clinical global impression of change, and an improvement in AIMS severity score of 50% or more was 34% for the deutetrabenazine arm versus 12% for placebo (26; 111). A long-term open-label study of deutetrabenazine estimated the mean change in AIMS score -4.9 at week 54, -6.3 at week 80, and -5.1 at week 106 (39). A 3-year, single-arm, open-label extension study of 343 patients with tardive dyskinesia treated with deutetrabenazine (mean dose: 39.4 ± 0.83 mg/day) reported mean change from baseline in total motor AIMS score of 6.6 ± 0.37 at week 145 (54). Two thirds of patients achieved at least 50% improvement in total motor AIMS score. Deutetrabenazine was generally well tolerated, with low rates of mild-to-moderate adverse events.

A new compound, 9-trifluoroethoxy-dihydrotetrabenazine, was identified as a promising agent for treatment of tardive dyskinesia due to its stronger, faster, and longer-lasting inhibitory effect on VMAT2 than valbenazine at an equivalent dose in rats (154).

Reserpine, another dopamine depleter, has been shown to be effective in the treatment of tardive dyskinesia, but because of its peripheral side effects such as orthostatic hypotension and lack of availability, it is almost never used (59; 34). Alpha-methylparatyrosine, a competitive inhibitor of tyrosine hydroxylase, can be effective when used with other presynaptic medications (34). A small double-blind, placebo-controlled study showed improvement of tardive dyskinesia with amantadine (118). Other drugs have been suggested as possibly being useful in treating tardive dyskinesias, including propranolol, levetiracetam, clonidine, and benzodiazepines (74). Propranolol has been reported to be effective for tardive dyskinesia in open-label studies and case series with doses less than 80 mg/day (33; 53). Vitamin E reduced Abnormal Involuntary Movement Scale (AIMS) scores by more than 40% in a double-blind study (166). A metaanalysis of 21 randomized controlled studies including 854 patients with tardive dyskinesia reported decrease of 2.36 in the AIMS score after vitamin E treatment compared to the control group (161). Vitamin B6 (108), a combination of acetazolamide and thiamine (28), verapamil (117), melatonin (134), gabapentin (52), and donepezil (20; 13) all may help, but at this time they are not established and not considered standard treatments. A Cochrane review of the studies of pyridoxal 5 phosphate (vitamin B6 or pyridoxine or pyridoxal phosphate) for the treatment of tardive dyskinesia found weak evidence for improvement in tardive dyskinesia (01). In an open-label study with eight patients, levetiracetam reduced abnormal movements (as judged on the AIMS) by 44% (79). Anticholinergics may exacerbate tardive orofacial dyskinesias and should not be used as treatment. Anticholinergics and baclofen, however, can be beneficial for tardive dystonia (71). An open-label study involving 11 subjects with tardive dyskinesia demonstrated that add-on zonisamide (50 to 100 mg/day) significantly decreased Abnormal Involuntary Movement Scale (AIMS), and 36% of subjects had a 20% or greater improvement on AIMS (60). A randomized, double-blind, placebo-controlled trial involving 157 subjects with schizophrenia evaluated the efficacy of the antioxidant EGb-761 (extract of Ginkgo biloba) in tardive dyskinesia (165). Compared to placebo, Ginkgo biloba significantly reduced AIMS. Naringin, a bioflavonoid found in citrus fruits and thought to have potent antioxidative, anti-inflammatory, antiapoptotic, and neuroprotective properties, was evaluated as a potential protective agent against haloperidol-induced orofacial dyskinesia in rats (155). Coadministration of naringin with haloperidol significantly prevented orofacial dyskinesia. Additionally, naringin-treated animals demonstrated decreased markers of haloperidol-induced nitric oxide and lipid peroxide production as well as neuroinflammatory and apoptotic markers, but increased the antioxidation power and neurotransmitter levels in the striatum. Treatment with isoflavones, a subclass of flavonoids found mainly in soy, was able to reduce the number of vacuous chewing movements in rats when co-administered with haloperidol (106). Rats co-treated with isoflavones also had lower levels of proinflammatory cytokines compared to animals treated only with haloperidol. Other herbal antioxidants were reported to be effective in the treatment of tardive dyskinesia in animal and small human studies or case series, including crocin (carotenoid constituent of saffron), curcumin (principal substance in turmeric), rice bran oil, and Harpagophytum procumbens (grapple plant) (17; 68; 126; 128). Animal studies of nicotine and drugs targeting nicotinic acetylcholine receptors have also shown a reduction in tardive dyskinesia (122; 18). Of the different forms of tardive dyskinesia, tardive akathisia is the most difficult to treat, but in some cases, zolpidem may be helpful without causing daytime drowsiness (153).

In addition to oral medications, injections of botulinum toxin may be helpful for certain forms of tardive dyskinesia, particularly tardive dystonia but also orofacial stereotypy and lingual dyskinesia (151; 09).

Pallidotomy (156) and thalamotomy (57) have ameliorated orolinguofacial tardive dyskinesia in patients treated for severe tardive dystonia. Deep brain stimulation largely replaced stereotactical ablation surgeries, and it can be effective for medically refractive tardive dystonia according to a number of small studies and case reports (105; 42; 109). A metaanalyis of 14 studies with total of 134 patients with tardive dystonia reported 66.88 ± 11.96% overall estimate improvement in the Burke-Fahn-Marsden Dystonia Rating Scale (BFMDRS) after pallidal deep brain stimulation, whereas only seven patients had worsening or lack of improvement after surgery (50). Nineteen patients with tardive dyskinesia treated with bilateral globus pallidus deep brain stimulation were followed for one year, and 14 patients were followed for up to 11 years (6 to 11 years) after surgery (124). A greater than 40% reduction in the Extrapyramidal Symptoms Rating Scale (ESPS) was reported at six months follow-up, 58% reduction at 12 months, and 60% long-term reduction. Two case series reported a successful treatment of tardive dyskinesia and tardive dystonia with subthalamic deep brain stimulation (164; 141). Deep brain stimulation, however, is an off-label use for tardive dyskinesia. A placebo-controlled study of repetitive transcranial magnetic stimulation (rTMS) in 26 subjects with tardive dyskinesia reported improvement of dyskinesia with AIMS score reduction by 8.3 +/- 1.7 points in rTMS arm versus only 1.2 +/- 3.3 points in the placebo arm (73).

Outcomes

Although treatments listed above can be beneficial, all have the potential for side effects (88), and we will discuss a few. Potential side effects of tetrabenazine include sedation, depression, akathisia, parkinsonism, and QT prolongation (64). Similar but milder and less frequent side effects were reported with the use of deutetrabenazine and valbenazine. Botulinum toxin, which is used effectively to treat focal dystonia, can cause excessive weakness of targeted muscles. Potential complications of deep brain stimulation include lead misplacement, intracerebral hemorrhage, hardware malfunction, infection, stimulation-related side effects, and although uncommon, suicide (42). Thus, deep brain stimulation should be reserved for patients who have failed alternate therapies.

Special considerations

Pregnancy

No information is available regarding the risks of pregnancy in patients with tardive dyskinesia compared to the general population. However, among a population of obstetric schizophrenic patients in Ireland, those with tardive dyskinesia were more likely to have a family history of psychiatric disorder with higher risk of complications (114).

One infant has developed tardive dyskinesia following birth to a woman who was taking haloperidol during pregnancy (133); the infant showed repetitive tongue protrusion until six months of age.

Anesthesia

A single patient with tardive dyskinesia of 3.5 years’ duration experienced a complete remission following general anesthesia for an orthopedic procedure (67). The benefit was sustained for at least three years. There is obviously no strong evidence for this approach.