Dopamine and serotonin (5-HT) receptor antagonism

Although drugs that are labeled as atypical antipsychotic agents block dopamine D2 receptors, they also block serotonin 5-HT_2A receptors, and some investigators attribute their antipsychotic effect to this mechanism and recommend research and drug development on newer agents that block other 5-HT receptors (Meltzer, 1999).

How tightly drugs to bind to the D2 receptor (their occupancy rate) is currently considered to be important in the drug having neuroleptic potential in direct proportion. Seeman and Tallerico (1998) found that the atypical antipsychotic drugs bind more loosely than classic antipsychotic drugs (neuroleptics) to the D2 receptors. Kapur and colleagues (1999) measured D2 and 5-HT₂ receptor occupancies by positron emission tomography (PET) scan for patients receiving clozapine, risperidone, or olanzapine. Clozapine showed a much lower D2 occupancy (16–68%) than risperidone (63–89%) and olanzapine (43–89%); all three showed greater 5-HT₂ occupancy than D2 occupancy at all doses, although the difference was greatest for clozapine. In their PET study, Moresco and colleagues (2004) found that striatal D2 receptor occupancy was significantly higher with olanzapine than with clozapine. All these studies support the relationship between receptor occupancy and the clinical observations that clozapine and quetiapine are the most “atypical,” with the least propensity to cause parkinsonism, tardive syndromes, or other neuroleptic drug reactions.

Despite interest in the 5-HT_2A receptor, it is the D2 receptor that is associated with the antipsychotic effect (Seeman, 2010). Antipsychotic clinical doses correlate with their affinities for this receptor. The receptor has high- and low-affinity states. Clozapine and quetiapine do not elicit parkinsonism and rarely result in tardive dyskinesia because they are released from D2 within 12–24 hours. Traditional antipsychotics remain attached to D2 receptors for days, preventing relapse, but allowing accumulation that can lead to tardive dyskinesia. Glutamate agonists that treat schizophrenia have affinity for the dopamine D2(High) receptor and the D3 receptor (Seeman and Guan, 2009).

A comparison of the receptor-binding profile of quetiapine, clozapine, olanzapine, and risperidone is presented in Table 19.5. Both quetiapine and clozapine have poor affinity for the dopamine D2 receptor, which probably accounts for the low incidence of inducing parkinsonism and tardive syndromes. Risperidone has the highest affinity for the D2 receptor, resembling haloperidol (see Table 19.6), and is therefore more of a “typical” than an “atypical” antipsychotic.

In one PET study striatal D1 and D2 receptor occupancies were evaluated (Tauscher et al., 2004). D1 occupancy ranged from 55% with clozapine to 12% with quetiapine (rank order: clozapine > olanzapine > risperidone > quetiapine). The striatal D2 occupancy ranged from 81% with risperidone to 30% with quetiapine (rank order: risperidone > olanzapine > clozapine > quetiapine). The ratio of striatal D1/D2 occupancy was significantly higher for clozapine (0.88) relative to olanzapine (0.54), quetiapine (0.41), or risperidone (0.31). In an [¹²³I]epidepride single photon emission computed tomography (SPECT) study involving clozapine-, olanzapine-, and haloperidol-treated schizophrenia patients, as well as drug-naive patients and healthy controls, midbrain D(2/3) receptor occupancy was studied. Of those on medication, occupancy was least for those on clozapine (5%), next for those on olanzapine (28%), and greatest for those on haloperidol (40%); no significant differences were observed in the temporal poles (Tuppurainen et al., 2009). All imaging studies are compatible with the clinical results in that clozapine and quetiapine, with the least neuroleptic features, have the lowest D2 receptor occupancy rate.

Clozapine

Clozapine was the first agent to be labeled as an atypical antipsychotic and deservedly so, although rare case reports do exist of drug-induced parkinsonism (Kurz et al., 1995), but it has been reported not to induce rigidity, and it rarely induces parkinsonian tremor (Gerlach and Peacock, 1994). It has also caused rare cases of acute akathisia (Friedman, 1993; Safferman et al., 1993), acute dystonic reaction (Kastrup et al., 1994; Thomas et al., 1994), tardive syndromes (Dave, 1994), tardive dystonia (Bruneau and Stip, 1998; Molho and Factor, 1999a, van Harten et al., 2008b), tardive akathisia (Kyriakos et al., 2005), tardive oculogyria (Uzun and Doruk, 2007), and NMS (Sachdev et al., 1995; Benazzi, 1999; Lara et al., 1999; Gambassi et al., 2006). There are also parkinsonian features that are seen with clozapine. In a prospective study, seven out of 25 patients on clozapine developed TD (Bunker et al., 1996). In a retrospective study, comparison of clozapine and typical antipsychotics showed no lower prevalence of tardive dyskinesia in the clozapine group (Modestin et al., 2000). But it is not clear that the patients in the clozapine group had not been previously exposed to typical antipsychotics and had not developed TD while receiving them. What is clear from this study is that the conversion to using clozapine has not markedly reduced the prevalence of extrapyramidal syndromes; this finding supports the realization that clozapine does not effectively treat TD once it has developed. How effective its use would be in preventing TD in the first place still needs to be verified. For patients with PD, the low propensity to augment existing parkinsonism makes clozapine very useful in treating patients who have dopaminergic drug-induced psychosis (Factor and Friedman, 1997; Friedman et al., 1999).

A SPECT study measuring dopamine D2 receptor binding reveals lower binding with clozapine than with typical neuroleptics (Broich et al., 1998). In animal studies, rats that have been pretreated with haloperidol for 4 weeks develop vacuous chewing movements (VCMs) after treatment with a dopaminergic, whereas rats that have been pretreated with clozapine do not (Ikeda et al., 1999).

A problem with clozapine is that weekly blood counts are required because there is a 1% to 2% incidence of leukopenia, which is reversible if the drug is withdrawn within 1–2 weeks. Granulocyte colony-stimulating factor can be an effective means to treat the agranulocytosis (Sperner-Unterweger et al., 1998). Other common adverse effects are drowsiness, drooling, weight gain, and seizures. An unusual adverse effect with clozapine was a case of myokymia (David and Sharif, 1998). Myocarditis associated with clozapine has been reported (Hill and Harrison-Woolrych, 2008).

Quetiapine

After clozapine, quetiapine is the most favorable in being least likely to induce parkinsonism, NMS (Stanley and Hunter, 2000), or tardive syndromes. In a prospective, head-to-head comparison with risperidone treatment in schizophrenia, quetiapine had far fewer extrapyramidal adverse effects, but more orthostasis (Perez et al., 2008). D2 receptor antagonism is relatively selective for limbic than striatal receptors for clozapine and sertindole, followed by quetiapine, ziprasidone, olanzapine, and remoxipride, whereas risperidone in many respects has a profile that resembles that of haloperidol (Arnt and Skarsfeldt, 1998). In primates, quetiapine and clozapine were equally less likely to induce oral dyskinesias compared to standard antipsychotics (Peacock and Gerlach, 1999). Like clozapine, quetiapine easily induces drowsiness, so when it is used to treat dopa-induced psychosis, it should be taken at bedtime. The major advantage over clozapine is that it does not require blood tests because it does not induce agranulocytosis. However, there may be a rare risk that quetiapine can cause agranulocytosis (Ruhe et al., 2001). Since its introduction, there have been reports of acute dystonic reactions (Jonnalagada and Norton, 2000; Desarker and Sinha, 2006), acute akathisia (Prueter et al., 2003), neuroleptic malignant syndrome (El-Gaaly et al., 2009; Gortney et al., 2009), and also TD (Ghaemi and Ko, 2001). In 367 patients treated for 12 months with quetiapine (mean dose 720 mg/day), parkinsonism was seen in 10%, akathisia in 3%, dystonia in 1.4%, and “hyperkinesia” in 1.7% (Perez et al., 2008).

In a SPECT study using the D2/D3 ligand [I-123]-epidepride the percent occupancy of receptors in the limbic system (temporal lobe) and striatal receptors while patients were receiving quetiapine was 60% and 32%, respectively, which is similar to clozapine (Stephenson et al., 2000). In another SPECT study, quetiapine was shown to occupy 5-HT_2A receptors in the frontal and temporal cortex (Jones et al., 2001). A PET study using [18F]fallypride showed similar results (Vernaleken et al., 2010).

A patient with oculogyric crisis, who failed to improve after withdrawal of antipsychotic medication, was successfully treated with quetiapine (Gourzis et al., 2007). Drug-induced parkinsonism and acute akathisia due to other neuroleptics can be reduced by switching patients to quetiapine (Cortese et al., 2008). Drowsiness and weight gain are common; postural hypotension is not infrequent. A rare complication of quetiapine is hepatotoxicity (Shpaner et al., 2008).

Olanzapine

In contrast to clozapine and quetiapine, olanzapine more readily increases parkinsonian symptoms in patients with PD (Jimenez-Jimenez et al., 1998; Molho and Factor, 1999b; Granger and Hanger, 1999), but does so less readily than risperidone and conventional antipsychotics. Drug-induced parkinsonism, including the rabbit syndrome, is seen with olanzapine (Durst et al., 2000), but less than with haloperidol (Peuskens et al., 2009). It can induce acute akathisia (Kurzthaler et al., 1997; Jauss et al., 1998) and TD but less so than haloperidol (Tollefson et al., 1997; Wood 1998). It can also induce NMS (Filice et al., 1998; Moltz and Coeytaux, 1998; Burkhard et al., 1999; Levenson, 1999; Margolese and Chouinard, 1999; Sierra-Biddle et al., 2000; Abu-Kishk et al., 2004; Zaragoza Fernandez et al., 2006), tardive dyskinesia (Herran and Vazquez-Barquero, 1999), and tardive dystonia (Dunayevich and Strakowski, 1999). Acute dystonic reactions also occur (Beasley et al., 1997; Landry and Cournoyer, 1998; Vena et al., 2006).

A direct comparison with chlorpromazine showed similar parkinsonism and acute akathisia for the two drugs (Conley et al., 1998). A double-blind comparison with haloperidol by Beasley and colleagues (1999) showed a much lower risk of developing tardive dyskinesia with olanzapine. After 1 year of exposure, 0.52% of patients developed TD with olanzapine and 7.45% developed TD with haloperidol. An open-label comparison with conventional antipsychotics after a 9-month follow-up after discharge from the hospital favored olanzapine, with TD being present in 2.3% for olanzapine (2/87), and 16.7% (12/72) for the conventional treatment (Mari et al., 2004). There are case reports of agranulocytosis induced by olanzapine (Meissner et al., 1999; Naumann et al., 1999; Tolosa-Vilella et al. 2002), including in a patient who had previously had agranulocytosis with clozapine (Benedetti et al., 1999). This has been attributed to some similar structural and pharmacologic properties of clozapine. A case of restless legs syndrome with periodic movements in sleep has been attributed to olanzapine (Kraus et al., 1999).

Olanzapine has relative regional mesolimbic dopaminergic selectivity and a broad-based binding affinity for serotonin (all 5-HT₂ receptor subtypes and the 5-HT₆ receptor), dopamine (D2, D3, and D4 receptors), muscarinic, and α₁-adrenergic receptors (Bymaster et al., 1999). A PET study in schizophrenic patients being treated with olanzapine revealed that this drug is a potent 5-HT₂ blocker, but also a blocker of D2 dopamine receptors similar to risperidone and less so than clozapine (Kapur et al., 1998). Patients on olanzapine were also studied with [¹²³I]iodobenzamide (IBZM) SPECT; the D2 receptor was occupied 60% and 83% of the time at doses of 5 mg/day and 10 mg/day, respectively (Raedler et al., 1999). Such high rates of occupancy probably account for olanzapine’s tendency to worsen parkinsonism and induce DRBA complications, because, as was noted above, D2 receptor occupancy rates are directly related to neuroleptic potential.

Risperidone

With the success of clozapine, there is a commercial advantage for pharmaceutical companies to tout other antipsychotics as “atypical.” Such has been claimed for risperidone, but this drug can readily induce parkinsonism including the rabbit syndrome (Levin and Heresco-Levy, 1999), tardive syndromes (Haberfellner 1997; Silberbauer 1998; Hong et al., 1999; Ananth et al., 2000), and NMS (Newman et al., 1997; Norris et al., 2006). The prevalence of parkinsonism from risperidone is usually considered to be less than that with conventional neuroleptics, but it was observed in 42% compared to 29% in those on haloperidol (Knable et al., 1997). Tardive dyskinesia and tardive dystonia occurred in a patient who was exposed only to risperidone (Bassitt and Garcia, 2000). The annual incidence of TD in patients taking risperidone has been estimated to be 0.3%, compared to an annual incidence in patients taking conventional neuroleptics of 5–10% (Gutierrez-Esteinou and Grebb, 1997). However, in an open-label prospective study of 255 institutionalized patients with dementia who were treated with risperidone, the 1-year cumulative incidence of TD was 2.6% (Jeste et al., 2000).

Like conventional neuroleptics, risperidone induces acute dystonic reactions in marmosets, in contrast to clozapine, which does not (Fukuoka et al., 1997). In a report of an open-label comparison with haloperidol, Rosebush and Mazurek (1999) found the two drugs to have a similar side effect profile. In a prospective follow-up of first-episode schizophrenic patients treated with risperidone, movement disorders developed in more than one-third of these patients, who had previously never been exposed to antipsychotic drugs (Lang et al., 2001). When risperidone is compared with low-potency antipsychotics, such as thioridazine, no difference was discerned in the rates of developing movement disorders (Schillevoort et al., 2001).

In a SPECT study dopamine receptor binding with IBZM showed risperidone to have effects between those of haloperidol and clozapine, with a dose–response curve for risperidone showing greater similarity to that of haloperidol (Dresel et al., 1998). A PET study showed dopamine D2 receptors occupancies of about 70% and 60% in the striatum and extrastriatum, respectively (Ito et al., 2009). Clearly, risperidone is not an “atypical” antipsychotic agent. Risperidone’s occupancy of the 5-HT₂ receptors is about 90%, and its occupancy of the D2 receptors is between 50% and 80%, but the latter correlates with the extrapyramidal side effects (Yamada et al., 2002).

Ziprasidone

This benzisothiazole was approved by the Food and Drug Administration in 2001. There are already case reports of acute dystonic reactions (Weinstein et al., 2006; Yumru et al., 2006, Rosenfield et al., 2007), NMS, rhabdomyolysis, and pancreatitis (Murty et al., 2002; Yang and McNeely, 2002; Gray 2004). It has been reported to cause tardive dyskinesia (Mendhekar, 2005) and tardive dystonia (Papapetropoulos et al., 2005; Tsai et al., 2008). Although it is a potent 5-HT_2A antagonist (like risperidone, olanzapine, and clozapine), it is also a D2 antagonist in humans as detected by PET scan (Bench et al., 1993, 1996). But in-vitro studies reveal much lower affinity for the D2 receptor than for the 5-HT_2A receptor (Seeger et al., 1995), and ziprasidone also binds less tightly to the D2 receptor than dopamine (Seeman, 2002). Ziprasidone has two other unique features compared to other antipsychotic agents: (1) it is a potent 5-HT_1A agonist and thus inhibits dorsal raphe serotonergic cell firing (Sprouse et al., 1999) and increases cortical dopamine release (Rollema et al., 2000), and (2) it inhibits neuronal uptake of 5-HT and norepinephrine in a manner comparable to the antidepressant imipramine (Schmidt et al., 2001). What these unique actions might contribute to antipsychotic activity or to propensity for or against extrapyramidal reactions is unclear. Ziprasidone has been reported to induce a case of oculogyric crisis in an adult (Viana Bde et al., 2009).

Aripiprazole

This quinolinone has a novel mechanism of action. Like a number of other “atypical” antipsychotics, it is an antagonist at the 5-HT_2A receptors. It is also a 5-HT_1A partial agonist (Jordan et al., 2002). What is novel is that it is a partial agonist at the dopamine D2 receptor. It has a higher affinity for the presynaptic autoreceptor than for the postsynaptic receptor. Hence, it reduces dopamine synthesis and release through an agonist action at the dopamine autoreceptor (Tamminga and Carlsson, 2002). Because of its novel action as a partial D2 agonist, it was anticipated that it might cause fewer extrapyramidal adverse effects, and clinical trials reported favorable results (Kane et al., 2002). However, D2 ligand binding PET revealed a dose-related high occupancy state of 71.6% at 2 mg/day to 96.8% at 40 mg/day (Kegeles et al., 2008). With wider use, it has been shown to worsen PD (Friedman et al., 2006; Wickremaratchi et al., 2006), and there are already reports of it causing parkinsonism (López-Torres et al., 2008), NMS (Chakraborty and Johnston, 2004; Hammerman et al., 2006; Palakurthi et al., 2007), acute dystonic reactions (Desarkar et al., 2006; Fountoulakis et al., 2006), and tardive dyskinesia (Zaidi and Faruqui, 2008; Abbasian and Power, 2009; Friedman, 2010).

Amisulpride

Amisulpride, a substituted benzamide, is a highly selective antagonist for dopamine D2 and D3 receptors in the limbic region, which would predict potent antipsychotic activity with a low potential to cause extrapyramidal symptoms (Lecrubier, 2000). It binds less tightly to the D2 receptor than do the typical antipsychotics (Seeman, 2002). Using the full range of recommended amisulpride dosage, striatal occupancies up to 90% can be measured (Vernaleken et al., 2004). It is not yet commercially available in the United States. It has already been reported to induce tardive oculogyric crisis (Mendhekar et al., 2010).

Adverse effects from second-generation antipsychotics

Aside from the D2-blocking effects described above, the five drugs in Table 19.7 have a number of other adverse effects. Sedation is a particular problem seen with each of them, but particularly clozapine and quetiapine. Metabolic side effects are now widely recognized with the second-generation antipsychotics, with weight gain a common problem (Leucht et al., 2009; Patel et al., 2009). Table 19.7 lists the common adverse effects of these drugs, along with haloperidol, for comparison.

It is not likely that drug-induced tardive syndromes will disappear because use of both typical and atypical agents continues. In a survey in Lombardy (Italy), 35 363 individuals over the age of 65 were prescribed an antipsychotic prescription (2.18 subjects per 100 inhabitants, with two-thirds receiving first-generation agents) (Percudani et al., 2004). Although there may be a lower risk of developing these disorders with “atypicals,” these drugs can induce them (Table 19.8). It is even possible that if physicians consider “atypicals” safe to use, their use will increase, even in situations in which the risk from typicals might have precluded their use. In the United States, the prevalence of atypical antipsychotic use was found to be 267.1 per 100 000 subjects aged 19 years and younger and was more than twice as high for male patients as for female patients (Curtis et al., 2005).

In a large US randomized controlled trial (the CATIE study) comparing second-generation antipsychotics (olanzapine, quetiapine, risperidone, perphenazine, or ziprasidone) with the first-generation perphenazine, only olanzapine had a slightly more favorable outcome of both the patients’ ability to stay on this drug for a longer duration and with slightly less drug-induced parkinsonism, akathisia, and tardive dyskinesia than all other drugs including perphenazine (Table 19.9); perphenazine fared as well as the other so-called ‘atypicals” (Lieberman et al., 2005). A similar outcome was seen in a smaller British study (CUtLASS) (Jones et al., 2006). This was a randomized controlled clinical trial involving 57 centers, with flexible dosing up to 4 capsules/day in 1460 schizophrenic subjects (after excluding 33 from one center, integrity concern). The primary outcome was time to discontinuation, with a maximum observation of 18 months. TD was present in 231 subjects at baseline, and these subjects were excluded from the perphenazine arm. The outcome had multiple comparison, requiring a P value ≤0.017 to be significant.

In a more detailed assessment of extrapyramidal side effects in the CATIE study, there were no significant differences in incidence or change in rating scales for parkinsonism, dystonia, akathisia, or tardive dyskinesia when comparing the second-generation antipsychotics with perphenazine or comparing between second-generation antipsychotics (Miller et al., 2008). Secondary analyses revealed greater rates of concomitant antiparkinsonism medication among individuals on risperidone and lower rates among individuals on quetiapine.

Metoclopramide

After the introduction of metoclopramide for gastrointestinal disorders, reports of acute dystonic reactions in children and tardive dyskinesia in adults began to appear (Cochlin, 1974; Kataria et al., 1978; Lavy et al., 1978). Today, metoclopramide has become a common cause of TD, and the legal profession has found such cases ripe for lawsuits. The annual incidence of TD from metoclopramide has been reported to be between 4.5% and 23% per year (Grimes, 1981; Ganzini et al., 1993), about the same as that for haloperidol ~5.7%/year (Tollefson et al., 1997; Beasley et al., 1999; Csernansky et al., 2002), but higher than those for risperidone ~1.8%/year (Lemmens et al., 1999; Davidson et al., 2000; Jeste et al., 2000; Csernansky et al., 2002) and olanzapine ~1.2%/year (Tollefson et al., 1997; Beasley et al., 1999). Because antipsychotic drugs have an affinity for brain neuromelanin (Seeman, 1988), Chen and colleagues (2010) examined the amount of metoclopramide and antipsychotic drugs binding to postmortem human substantia nigra. They found that in clinical conditions the amount of metoclopramide that would bind to nigra is much higher than the amount of raclopride, haloperidol, or olanzapine that would be expected to be bound and they suggest that this binding might explain the higher annual incidence of TD induced by metoclopramide.

Acute dystonia

The earliest abnormal involuntary movement to appear after initiation of dopamine receptor antagonist therapy is an acute dystonic reaction. In about half of the cases, this reaction occurs within 48 hours, and in 90% of cases, it occurs by 5 days after starting the therapy (Ayd, 1961; Garver et al., 1976). The reaction may occur after the first dose (Marsden et al., 1975).

Dystonic movements are sustained muscle contractions, frequently causing twisting and repetitive movements, or abnormal postures (Fahn, 1988). In a series of 3775 patients, Ayd (1961) found that acute dystonia is the least frequent side effect, affecting about 2–3% of patients, with young males being most susceptible. The incidence rate increases to beyond 50% with highly potent dopamine receptor blockers such as haloperidol (Boyer et al., 1987). In one prospective study, Aguilar and colleagues (1994) reported that 23 of 62 patients developed acute dystonia after haloperidol was introduced, that anticholinergic pretreatment significantly prevented this, and that younger age and severity of psychosis were risk factors. In the prospective study by Kondo and colleagues (1999), 20 (51.3%) of 39 patients placed on nemonapride had dystonic reactions, onset occurring within 3 days after the initiation of treatment in 90%. As in other series, the incidence of acute dystonia was significantly higher in males than in females (77.8% vs. 28.6%, P < 0.05), and younger males (≤30 years) had an extremely high incidence (91.7%). The incidence is much lower with the so-called atypical antipsychotics; Raja and Azzoni (2001) observed only 41 cases out of 1337 newly admitted patients treated with antipsychotics, which included 8 treated with risperidone, 1 with olanzapine, and 1 with quetiapine. A meta-analysis compared intramuscular injections of second-generation antipsychotics (SGAs) to haloperidol injections (Satterthwaite et al., 2008). SGAs were associated with a significantly lower risk of acute dystonia (relative risk = 0.19) and acute akathisia (relative risk = 0.25), compared with haloperidol alone.

All agents that block D2 receptors can induce acute dystonic reactions, including risperidone (Brody, 1996; Simpson and Lindenmayer, 1997) and clozapine (Kastrup et al., 1994). One child developed an acute dystonic reaction after ingestion of a dextromethorphan-containing cough syrup (Graudins and Fern, 1996). Dextromethorphan has several different known pharmacologic actions, but D2 receptor blockade is not one of them. A case of acute dystonia following abrupt withdrawal of bupropion (DA reuptake blocker) was reported (Wang et al., 2007). A case of acute dystonia following abrupt withdrawal of dexamphetamine in a patient taking risperidone was also reported (Keshen and Carandang, 2007). Serotonergic agents have also been reported to induce acute dystonic reactions (Lopez-Alemany et al., 1997; Madhusoodanan and Brenner, 1997; Olivera, 1997). The mechanism could relate to inadequate release of dopamine from the nerve terminals in the striatum owing to the inhibitory effect of serotonin on dopamine neurons in the substantia nigra pars compacta. The opioid σ₁ and σ₂ receptors have also been implicated (Matsumoto and Pouw, 2000).

Acute dystonic reactions most often affect the ocular muscles (oculogyric crisis), face, jaw, tongue, neck, and trunk, and less often the limbs. Oculogyric crisis has previously been noted to occur as a common feature of postencephalitic parkinsonism (Duvoisin and Yahr, 1965). A typical acute dystonic reaction may consist of head tilt backward or sideways with tongue protrusion and forced opening of the mouth, often with arching of trunk and ocular deviation upward or laterally (Rupniak et al., 1986). The forcefulness of the muscle contractions can be extremely severe and led to auto-amputation of the tongue in one patient (Pantanowitz and Berk, 1999).

Mazurek and Rosebush (1996) studied the timing in the development of an acute dystonic reaction in 200 patients who were taking a neuroleptic medication for the first time. The neuroleptic was given twice daily, and over 80% of the episodes of acute dystonia occurred between 12 noon and 11 p.m.

Reserpine and α-methylparatyrosine, which deplete presynaptic monoamines, have not been associated with acute dystonic reactions (Duvoisin, 1972; Marsden et al., 1975; Walinder et al., 1976). However, another dopamine depletor, tetrabenazine (TBZ), has been reported to induce acute dystonic reactions (Burke et al., 1985). One possible explanation for this difference is that in addition to depleting dopamine, TBZ blocks dopamine receptors (Reches et al., 1983).

It is important to mention the case described by Wolf (1973) of a patient who developed an oral dyskinesia while taking reserpine. Figure 1 in his paper clearly shows a dystonic phenomenon and not the complex, rapid, stereotypic movements of classic tardive dyskinesia. This development appeared too late after initiation of reserpine to be considered an acute dystonic reaction, however. Whether this could be an example of tardive dystonia is not clear, since the phenomenology of tardive dystonia is identical to that of naturally occurring primary dystonia; therefore, the patient could have had a coincidental case of spontaneous, idiopathic oromandibular dystonia (Fahn, 1984b). Thus, there is no absolute evidence that reserpine induces acute dystonic reactions, classic tardive dyskinesia, or tardive dystonia. In fact, symptoms of these three tardive syndromes can be suppressed by reserpine, and eventually reserpine can be withdrawn successfully in many patients without exacerbation of the symptoms (Fahn, 1983). Furthermore, the long time that reserpine has been available without a clear-cut case of tardive syndromes can be compared to the much shorter duration of use of metoclopramide, in which there are already cases of acute dystonic reactions, classic tardive dyskinesia, and tardive dystonia as a consequence of its use (Casteels-Van Daele et al., 1970; Gatrad, 1976; Pinder et al., 1976; Reasbeck and Hossenbocus, 1979; Miller and Jankovic, 1989; Lang, 1990).

Of 452 patients given high-dosage oral metoclopramide to control emesis, Kris and colleagues (1983) observed 14 who developed acute dystonic reactions. However, there was a distinct preponderance of the reactions occurring in children (6/22) compared to adults (8/430). Intravenous metoclopramide is more likely to induce it than oral administration (Pinder et al., 1976). In a study at a Veterans Administration hospital, comparing patients treated with metoclopramide and controls, the relative risk for TD was 1.67, and the relative risk for drug-induced parkinsonism was 4.0 (Ganzini et al., 1993).

The available biochemical explanations for the acute dystonic reaction are unsatisfactory, but several observations contribute to the understanding of the phenomena, relating it to dopamine, muscarinic, and sigma receptors. Acute neuroleptic administration produces sudden increase of dopamine release and increased turnover lasting for 24–48 hours after a single dose (O’Keefe et al., 1970; Marsden and Jenner, 1980). This effect is blocked by anticholinergics consistent with their efficacy in treatment of acute dystonic reactions (O’Keefe et al., 1970). Moreover, in baboons that were pretreated with reserpine and α-methylparatyrosine, which markedly reduces presynaptic dopamine concentration, haloperidol-induced acute dystonic reaction is abolished or greatly reduced (Meldrum et al., 1977). On the other hand, blockade of the postsynaptic receptor fades in about 12 hours after a single dose of antipsychotics, and supersensitivity of receptors begins to develop. Therefore, presynaptic dopaminergic excess in combination with the emerging supersensitive postsynaptic dopamine receptors could result in markedly increased striatal dopaminergic activity at about 20–40 hours after a neuroleptic dose. This period corresponds to the critical time for acute dystonic reactions in human subjects who were given a single dose of butaperazine (Garver et al., 1976). However, extrapolating the data from experimental animals to the clinical situation has its limitations including the fact that rats do not develop acute dystonic reactions. Further data on human cerebrospinal fluid (CSF) dopamine metabolites as an indicator of presynaptic function might prove to be of value in pursuing the hypothesis. Jeanjean and colleagues (1997) suggested that the σ₂ receptors could be involved in the acute dystonic reaction. They found a correlation between the clinical incidence of neuroleptic-induced acute dystonia and binding affinity of drugs for the sigma receptor.

Another animal model of acute dystonia is the common marmoset that is treated with haloperidol (Fukuoka et al., 1997). But it takes at least 6 weeks of such treatment to develop this reaction. The dystonia subsides only to reappear when haloperidol treatment is restarted; other neuroleptics, including risperidone, can also make the dystonia reappear, but clozapine was without such an effect. In this animal model, the anticholinergic agent trihexyphenidyl inhibited the induction of acute dystonia.

In patients with acute dystonic reactions, symptoms can be relieved within minutes after parenteral anticholinergics or antihistaminics (Paulson, 1960; Waugh and Metts, 1960; Smith and Miller, 1961). Diphenhydramine 50 mg, benztropine mesylate or biperiden 1–2 mg is given intravenously and can be repeated if the effect is not seen in 30 minutes. Intravenous diazepam has also been shown to be effective and can be used as an alternative therapy (Korczyn and Goldberg, 1972; Gagrat et al., 1978; Rainer-Pope, 1979). If untreated, the majority of cases still resolve spontaneously in 12–48 hours after the last dose of the dopamine receptor antagonists. Dopamine receptor antagonists with high anticholinergic activities have low incidence rates of acute dystonic reactions (Swett, 1975). Therefore, prophylactic use of anticholinergics (Arana et al., 1988) and benztropine (Goff et al., 1991) has been studied and reported to be helpful in reducing the risk of acute dystonic reactions, especially in young patients on high-potency drugs. Three cases of recurrent episodes of acute dystonia and oculogyric crises despite withdrawal of the offending DRBAs (haloperidol in two cases, metoclopramide in one) have been reported, responding to anticholinergics each time (Schneider et al., 2009). The response of acute dystonic reactions to anticholinergics is so characteristic that it is difficult to explain the report of a few cases apparently due to amitriptyline, which has considerable anticholinergic activity and no known DRBA activity (Ornadel et al., 1992).

A case of an acute dystonic reaction occurring in an elderly person with bipolar disorder taking the serotonin uptake inhibitor paroxetine has been reported (Arnone et al., 2002). Speculation about the risk due to previous exposure to neuroleptics was raised. This class of drug can reduce firing rate of nigral dopaminergic neurons owing to their inhibition by serotonin.

Acute akathisia

The term akathisia, from the Greek, meaning unable to sit down, was coined by Haskovec in 1901 (cited by Mohr and Volavka, 2002), long before antipsychotic drugs were introduced. Akathisia was seen in some patients with advanced parkinsonism, and in others, it was frequently thought to be functional. Akathisia refers to an abnormal state of excessive restlessness, a feeling of the need to move about, with relief of this symptom on actually moving. Today, it is most frequently encountered as a side effect of neuroleptic drugs.

Two major issues of akathisia remain in confusion. First, there is no consensus about diagnostic criteria. Some authors consider akathisia to be an abnormal subjective state and regard the movements as an expression of the subjective state but not a necessary feature for the diagnosis (Van Putten, 1975). Others recognize the characteristic patterns of restless movements and consider presence of movements to be sufficient for the diagnosis (Munetz and Cornes, 1982; Barnes et al., 1983; Gibb and Lees, 1986). A second point of confusion is that akathisia occurs not only in an early-onset, self-limited form (acute akathisia), but also as a late-onset, persistent form (tardive akathisia). Much of the literature on akathisia does not distinguish between acute and tardive akathisia, which makes the interpretation of the literature difficult. The recognition of tardive akathisia as a distinct subsyndrome of tardive syndromes has been more recent (Fahn, 1978, 1983; Braude and Barnes, 1983; Weiner and Luby, 1983). For the discussion of the clinical features of akathisia, acute and tardive akathisia are lumped together, since they are similar, but their treatments and most likely their pathophysiologies are distinct.

The subjective aspect of akathisia is characterized by inner tension and aversion to remaining still. Patients complain of vague inner tension, emotional unease, or anxiety with vivid phrases such as “jumping out of my skin” or “about to explode.” Subjective descriptions, however, can be nonspecific. Inner restlessness and inability to remain still can be present in a significant number of psychiatric patients without akathisia and in control subjects without psychiatric problems (Braude et al., 1983). In an attempt to clarify the issue, Braude et al. (1983) systematically surveyed the frequency of various complaints and found that inability to keep the legs still was the most characteristic complaint and was present in over 90% of patients with akathisia in contrast to about 20% of those with other psychiatric disturbances. Others noted more conservative estimation of the frequency of complaints related to the legs from 27% (Gibb and Lees, 1986) to 57% (Burke et al., 1989). Various authors also described atypical features such as “acting out,” suicidal ideation, disruptive behaviors, homicidal violence, sexual torment, terror, and exacerbation of psychosis as akathitic phenomena (Van Putten and Marder, 1987). Evaluation of the subjective aspect also depends on patients’ ability to describe their feelings. Those with psychosis, dementia, or learning disability are often unable to provide useful descriptions for diagnosis. Although akathisia may manifest itself as subjective feeling alone, lack of specific subjective feeling and variable expression by patients pose a diagnostic dilemma. Therefore, the presence of the motor phenomenon is very helpful for the diagnosis.

Akathisia can present as focal pain or burning, usually in the oral or genital region (Ford et al., 1994). The symptom of moaning may be a verbal expression of the subjective feeling of akathisia. Some patients may moan as part of a generalized akathitic state and have other motor evidence of akathisia, such as marching in place, inability to sit still accompanied by associated walking about, inability to lie quietly with writhing and rolling movements, and making stereotypic caressing or rocking movements. The differential diagnosis of moaning includes parkinsonism, akathisia, levodopa usage (Fahn et al., 1996), dementia, pain, and other syndromes of phonations, such as tics, oromandibular dystonia, and Huntington disease, as discussed by Fahn (1991).

The motor aspect of akathisia (akathitic movements) is generally described as excessive movements that are complex, semipurposeful, stereotypic, and repetitive. Braude and colleagues (1983) found that rocking from foot to foot, walking on the spot, and coarse tremor and myoclonic jerks of the feet were characteristic of akathitic movements. Others agree that various leg and feet movements are more common in patients with akathisia than in those with TD (Gibb and Lees, 1986; Burke et al., 1989). However, they also noted that these did not distinguish akathisia from the group that did not meet criteria for akathisia (Gibb and Lees, 1986), and movements involving other parts of the body such as trunk rocking, respiratory grunting, face rubbing, and shifting weight while sitting were also frequent (Burke et al., 1989). Although there are not enough data and consensus on the diagnostic movements of akathisia, these movements seem to be characteristic enough to be recognized by different authors who have independently documented similar phenomena.

Akathisia is seen in patients with Parkinson disease (Lang and Johnson, 1987), in patients abusing cocaine (Daras et al., 1994), and as an adverse effect of selective serotonin uptake inhibitors (Poyurovsky et al., 1995). Acute akathisia occurs not only as an adverse effect of DRBAs but also fairly commonly as an acute adverse effect of the dopamine depletors – reserpine, TBZ, and α-methyl tyrosine (Marsden and Jenner, 1980).

In Ayd’s review (1961), half of the cases of acute akathisia occurred within 25 days of drug treatment and 90% occurred within 73 days. Acute akathisia was the most common side effect of DRBAs, occurring in 21.2% of patients in that study (Ayd, 1961). In a more recent study by Sachdev and Kruk (1994) of 100 consecutive patients placed on neuroleptics, mild akathisia developed in 41% and moderate to severe akathisia in 21%. In a literature review, Sachdev (1995b) reported that incidence rates for acute akathisia with conventional neuroleptics vary from 8% to 76%, with 20% to 30% being a conservative estimate. Sachdev stated that preliminary evidence suggests that the newer atypical antipsychotic drugs are less likely to produce acute akathisia. Using the criterion that both subjective and objective phenomena are required for the diagnosis of acute akathisia, Miller and colleagues (1997) found an incidence rate of 22.4%, 75% of which occurred within the first 3 days of exposure to a neuroleptic. Muscettola and colleagues (1999) found a prevalence rate of 9.4%.

The potency of neuroleptics has been associated with incidence of akathisia, ranging from 0.5% for reserpine (Marsden and Jenner, 1980) to 75% for haloperidol (Van Putten et al., 1984). Other risk factors are neuroleptic dose, the rate of dosage increase, and the development of drug-induced parkinsonism (Sachdev and Kruk, 1994). Akathisia has occurred with all second-generation antipsychotics as well as first-generation ones (Kane et al., 2009).

As with acute dystonic reactions, akathisia has also been induced by serotonergic agents (Chong, 1996; Lopez-Alemany et al., 1997). The mechanism was discussed in the above section on acute dystonia.

As was noted previously, akathisia needs to be distinguished from other conditions such as agitated depression; restless legs syndrome, in which similar subjective sensations may be described by patients but are mainly localized to legs and are present particularly at night (Blom and Ekbom, 1961); or complex motor tics with preceding aura, which show more variety of abnormal movements; and complex vocal tics (Jankovic and Fahn, 1986). Akathisia can also be obscured by other psychiatric disorders, or it could be mistaken for a psychiatric disease. For example, when patients with psychosis develop akathisia after withdrawal from antipsychotic drugs, it may be mistaken for recurrence of psychosis. Paradoxical dystonia (see Chapter 12) in which dystonic movements are relieved by movement can be mistaken for akathisia.

The pathophysiology of acute akathisia remains poorly understood. On the basis of the observation in rats that show increased locomotor activity after blockade of the mesocortical dopamine system (Carter and Pycock, 1978), reduction of this dopaminergic projection was suggested to be responsible for akathisia (Marsden and Jenner, 1980). However, tardive akathisia cannot be explained by this hypothesis because dopamine depletors can ameliorate those symptoms. The observation that acute akathisia can occur with a serotonin uptake inhibitor (Altshuler et al., 1994) indicates that inhibiting dopamine neurons in the substantia nigra by such drugs could link the dopamine system with akathisia. These types of drugs have been reported to increase parkinsonism in patients with PD (Meco et al., 1994). One attractive possibility is that akathisia might reflect an alteration of the dopaminergic mesolimbic system.

Because propranolol has been reported to be beneficial in treating acute akathisia, another suggestion is that acute akathisia results from alterations in the cingulate cortex, the piriform cortex, or area 1 of the parietal cortex based on effects of propranolol in these regions in haloperidol-treated rats (Ohashi et al., 1998).

Ayd (1961) noted that acute akathisia is self-limited, disappearing on discontinuation of neuroleptics, and is well controlled by anticholinergics despite continuation of neuroleptics. Others have noted that only patients with concomitant parkinsonism improve significantly with anticholinergics (Kruse, 1960; Braude et al., 1983). Amantadine may also help, but patients can develop a tolerance (Zubenko et al., 1984a). Beta-blockers at relatively low doses below 80 mg of propranolol per day have been noted to be effective in many studies including one with a double-blind design (Lipinski et al., 1984; Zubenko et al., 1984b; Adler et al., 1986; Dupuis et al., 1987). Nonlipophilic beta-blockers that have poor penetration to the central nervous system (CNS) are not as effective (Lipinski et al., 1984; Dupuis et al., 1987). Selective beta-blockers might not be as effective as nonselective ones (Zubenko et al., 1984b); however, when two equally lipophilic beta-blockers, propranolol and betaxolol, were compared, they were equally effective in treating acute akathisia (Dumon et al., 1992) although the former is a beta-2-blocker and the latter is a beta-1-blocker. In a rat model of acute akathisia (neuroleptic-induced defecation), a lipophilic beta-1-blocker was found to be effective in reducing the phenomenon (Sachdev and Saharov, 1997).

Clonidine also reduces central noradrenergic activity by stimulating central alpha-2 receptors and has been noted to be effective in a small number of studies. The sedating effect is pronounced, however (Adler et al., 1987). Nicotine patches have been reported to reduce akathisia (Anfang and Pope, 1997). Weiss and colleagues (1995) found cyproheptadine, an antiserotonergic agent, to be effective in ameliorating akathisia. In a small placebo-controlled trial, mianserin, a 5-HT₂ antagonist, was found to reduce the severity of acute akathisia (Poyurovsky et al., 1999). Trazadone has also been reported to be beneficial (Stryjer et al., 2003). Poyurovsky and Weizman (2001) discuss the potential of serotonin agents in akathisia. A literature search revealed mirtazapine to be effective and superior to propranolol (43.3% vs. 30.0%) (Hieber et al., 2008).

Parkinsonism

Neuroleptic-induced parkinsonism (usually referred to as extrapyramidal syndrome or EPS by psychiatrists and as drug-induced parkinsonism by neurologists) is a dose-related side effect and is indistinguishable phenomenologically from idiopathic PD, including high frequency of tremor and asymmetric signs (Hardie and Lees 1988). SPECT imaging of the dopamine transporter, however, may be helpful in determining whether the neuroleptic-induced parkinsonism is entirely drug-induced or an exacerbation of subclinical PD (Lorberboym et al. 2006). It develops with use of both DRBAs and dopamine-depleting drugs such as reserpine and TBZ. Some authors have noted perioral tremor and termed this rabbit syndrome, which is a localized form of parkinsonian tremor (Decina et al., 1990). The incidence of parkinsonism varies. Korczyn and Goldberg (1976) found 61%, and Muscettola and colleagues (1999) found 19.4%. Women are affected almost twice as frequently as men, which is the reverse of the ratio in idiopathic PD. Neuroleptic-induced parkinsonism also occurs increasingly with advanced age (Ayd, 1961; Hardie and Lees 1988) in parallel with the incidence of idiopathic PD.

Blockade of dopamine receptors by antagonists or depletion of presynaptic monoamines by drugs such as reserpine mimics the deficient dopamine state in PD. All DRBAs can induce parkinsonism, except clozapine (there are only rare reports with clozapine) (Factor and Friedman, 1997). Risperidone can do so (Gwinn and Caviness, 1997; Simpson and Lindenmayer, 1997), as well as olanzapine and only rarely quetiapine. Parkinsonism from neuroleptics is typically reversible when the medication is reduced or discontinued. Sometimes, the reversal can take many months; an interval of up to 18 months has been noted in the literature (Fleming et al., 1970).

Some patients show persisting parkinsonism despite prolonged discontinuation of neuroleptics (Stephen and Williamson, 1984; Hardie and Lees, 1988), giving rise to consideration of a proposed condition of tardive parkinsonism. A study of 8-week exposure of rats to haloperidol found a highly significant 32–46% loss of tyrosine hydroxylase (TH) immunoreactive neurons in the substantia nigra, and 20% contraction of the TH-stained dendritic arbor (Mazurek et al., 1998). Perhaps such pathologic changes account for some cases of prolonged drug-induced parkinsonism in humans. Several cases in the literature had initial resolution of parkinsonism and later reappearance of the symptoms without re-exposure to neuroleptics (Hardie and Lees, 1988). Two cases that had complete resolution of drug-induced parkinsonism after withdrawal of neuroleptics showed evidence of mild PD at autopsy (Rajput et al., 1982). Although one assumes that these patients had subclinical PD, the effect of neuroleptics on the disease progression is unknown. The use of the term “tardive parkinsonism” to refer to cases of persistent parkinsonism remains an enigma. Some are due to concurrent development of progressive PD, and there are no autopsied proven examples of non-PD in any example. Therefore, this suggests that there is as yet no evidence of tardive parkinsonism.

With the introduction of selective serotonin reuptake inhibitors (SSRIs) to treat depression, it has been noticed that these drugs can sometimes worsen parkinsonism in patients with PD (Meco et al., 1994) and occasionally can induce parkinsonism in patients who never had symptoms of PD (Coulter and Pillans, 1995; DiRocco et al., 1998). In an intensive monitoring program in New Zealand of the SSRI drug fluoxetine over a 4-year period, there were 15 reports of parkinsonism in 5555 patients who were exposed to the drug (Coulter and Pillans, 1995). Four of these 15 patients were also on a neuroleptic and one was on metoclopramide. In a literature search, parkinsonism was found with different classes of antidepressants, is not dose related, and can develop with short-term or long-term use (Madhusoodanan et al., 2010). The explanation for inducing or enhancing parkinsonism is that increased serotonergic activity in the substantia nigra will inhibit dopamine-containing neurons, thus causing functional dopamine deficiency in the nigrostriatal pathway (Baldessarini and Marsh, 1992).

The possibility of the existence of tardive parkinsonism comes up from time to time because some patients have continued parkinsonism despite long-term discontinuation of the DRBA (Melamed et al., 1991). However, there is always the possibility that the patient had preclinical Parkinson disease prior to developing drug-induced parkinsonism. Then, when the DRBA is withdrawn, the parkinsonism persists because of progressively worsening PD. One would need to show that there are no Lewy bodies or that the PET scan shows no loss of fluorodopa uptake in patients believed to have tardive parkinsonism.

Anticholinergics can be effective in reducing the severity of the parkinsonism induced by DRBAs, whereas dopaminergic drugs (that activate the dopamine receptors) are ineffective, probably because they are not able to displace the DRBA from its binding to the receptor. Levodopa up to 1000 mg in combination with a peripheral dopa decarboxylase inhibitor had no significant effect (Hardie and Lees, 1988), nor did apomorphine, a dopamine receptor agonist (Merello et al., 1996). On the other hand, levodopa can effectively reverse parkinsonism induced by dopamine depletors, such as reserpine. In fact, the discovery of the dopamine hypothesis for parkinsonism was based on this observation (Carlsson et al., 1957; Carlsson 1959). Treatment is usually initiated with anticholinergics or amantadine (Mindham et al., 1972; Johnson, 1978; Konig et al., 1996).

Neuroleptic malignant syndrome

NMS is an idiosyncratic reaction that can sometimes be life-threatening. The clinical triad consists of (1) hyperthermia, usually with other autonomic dysfunctions such as tachycardia, diaphoresis, and labile blood pressure; (2) extrapyramidal signs, usually increased muscle tone of rigidity or dystonia, often with accompanying elevation of muscle enzymes; and (3) alteration of mental status, such as agitation, inattention, and confusion. Fever is not an essential symptom (Peiris et al., 2000), and it can be delayed (Norris et al., 2006). The syndrome begins abruptly while the patient is on therapeutic, not toxic, dosages of medication. In a review of 340 clinical reports of NMS in the literature (Velamoor et al., 1994), changes in either mental status or rigidity were the initial manifestations of NMS in 82.3% of cases with a single presenting sign. All the symptoms are fully manifest within 24 hours and reach a maximum within 72 hours. There does not seem to be any relationship with the duration of therapy. NMS can develop soon after the first dose or at any time after prolonged treatment. Recovery usually occurs within 1 to several weeks, but can be fatal in 20–30% of cases (Henderson and Wooten, 1981; Gute and Baxter, 1985). Even with awareness of the potential of fatality in modern medicine, death still occurs (van Maidegem et al., 2002). Muscle biopsies have shown swelling and edema, with 10–50% of fibers involved with vacuoles but scanty mononuclear infiltration (Behan et al., 2000).

All agents that block D2 receptors can induce NMS, including risperidone (Raitasuo et al., 1994; Webster and Wijeratne, 1994; Dave, 1995; Singer et al., 1995; Levin et al., 1996), clozapine (Miller et al., 1991; Amore et al., 1997; Dalkilic and Grosch, 1997), amisulpride (Bottlender et al., 2002), olanzapine (Kontaxakis et al., 2002; Kogoj and Velikonja, 2003), and phenothiazines with antihistaminic activity, such as alimemazine (van Maidegem et al., 2002). A case of NMS associated with bupropion has been reported (Kasantikul and Kanchanatawan, 2006). TBZ has been reported to cause NMS; this seems likely to be due to its D2-blocking activity (Reches et al., 1983) rather than to its dopamine-depleting action (by blocking the vesicular dopamine transporter). Reserpine has no known dopamine receptor antagonism, only dopamine-depleting activity (also by blocking the vesicular dopamine transporter), and has not been reported to cause NMS.

In a Japanese study, 10 of 564 (1.8%) patients who received antipsychotics developed NMS (Naganuma and Fujii, 1994), many more than the 12 of 9792 patients (0.1%) reported previously (Deng et al., 1990). Risk factors that were found were psychomotor excitement, refusal of food, weight loss, and oral administration of haloperidol at 15 mg/day or above (Naganuma and Fujii, 1994). Young males appear to be more predisposed to NMS (Gratz and Simpson, 1994), but the reason for this is uncertain. In a case-control study searching for risk factors, Sachdev et al. (1997) found that patients with NMS were more likely to be agitated or dehydrated, often needed restraint or seclusion, had received larger doses of neuroleptics, and more often had previous treatment with electroconvulsive therapy (ECT) before the development of the syndrome.

The pathophysiologic mechanism of NMS is not well understood. Autopsies failed to show any consistent findings (Itoh et al., 1977). A similar syndrome has been reported following abrupt withdrawal of levodopa (Friedman et al., 1985; Hirschorn and Greenberg, 1988; Keyser and Rodnitzky, 1991), suggesting a common mechanism of acute dopamine deficiency (Henderson and Wooten, 1981). IBZM SPECT in one patient showed the dopamine receptor to be blocked in the acute phase of NMS, but the patient had been receiving a D2 blocker, which would be expected to result in this finding (Jauss et al., 1996). There is a report of a patient who developed the NMS syndrome following abrupt withdrawal of the combination of a long-acting neuroleptic and an anticholinergic agent (Spivak et al., 1996). Because it responded to procyclidine administration, it implicates a muscarinic overactivity. There are also reports of an NMS-like syndrome following the sudden withdrawal of amantadine (has dopaminergic and antimuscarinic activity) (Ito et al., 2001), and following withdrawal of baclofen, with recovery after reintroduction of baclofen (Turner and Gainsborough, 2001).

The idiosyncratic nature and rarity of the syndrome remain unexplained. Ram and colleagues (1995) evaluated the structure of the D2 receptor gene in 12 patients who had a history of NMS. One patient was found to have a nucleotide substitution of an exon of the D2 gene. The A1 allele of the TaqI A polymorphism of the dopamine D2 receptor gene appears to occur more commonly in patients who developed NMS (Suzuki et al., 2001). Kishida et al. (2004) found that patients with NMS had a higher association with a polymorphism in the D2 receptor gene.

Treatment of NMS consists of discontinuing the antipsychotic drugs and providing supportive measures. Rapid relief of symptoms has been reported with the use of dantrolene, bromocriptine, or levodopa (Granato et al., 1982; Gute and Baxter, 1985). Nisijima and colleagues (1997) found levodopa to be more effective than dantrolene, but Tsujimoto and colleagues (1998) found intravenous dantrolene plus hemodialysis to be effective. Subcutaneous apomorphine has been found to be effective as a solo treatment (Wang and Hsieh, 2001). Gratz and Simpson (1994) recommended using anticholinergics in an attempt to reverse rigidity prior to utilizing bromocriptine. Carbamazepine was dramatically effective in two patients (with recurrence on withdrawal of the drug) (Thomas et al., 1998). Steroids added to standard therapy have been reported to speed recovery time (Sato et al., 2003). Re-exposure to dopamine receptor antagonists does not necessarily lead to recurrence of NMS (Singh and Albaranzanchi, 1995; Singh and Hambidge, 1998). Residual catatonia that can last weeks to months has been reported, with some patients responding to ECT (Caroff et al., 2000). Hyponatremia can sometimes occur; it has been attributed to inappropriate secretion of antidiuretic hormone and also to cerebral salt-wasting syndrome associated with the NMS (Lenhard et al., 2007). When present, salt replacement is necessary.

Withdrawal emergent syndrome

Withdrawal emergent syndrome was first described in children who had been on antipsychotic drugs for a long period of time and then were withdrawn abruptly from their medication (Polizos et al., 1973). The movements are choreic and resemble those of Sydenham disease (Videos 19.1 and 19.2). The abnormal movements are brief and flow from one muscle to another in a seemingly random way. They differ from the movements of classic tardive dyskinesia, which are brief, but stereotypical and repetitive. The movements in withdrawal emergent syndrome involve mainly the limbs, trunk, and neck, and rarely the oral region, which is the most prevalent site in classic tardive dyskinesia. The dyskinetic movements disappear spontaneously within several weeks after withdrawal of the DRBA. For immediate suppression of movements, dopamine receptor antagonists can be reinstituted and withdrawn gradually without recurrence of the withdrawal emergent syndrome (Fahn, 1984a). A withdrawal reaction from melatonin with O-B-L dyskinesia and akathisia was reported by Giladi and Shabtai (1999) and was described earlier in the chapter.

Withdrawal emergent syndrome is analogous to the classic tardive dyskinesia seen in adults, except that the course is more benign and movements are more generalized, resembling the choreic movements of Sydenham disease. In fact, most cases of tardive dyskinesia that have been reported in children have a benign course and the phenomenology has been reported to be more generalized choreic movements rather than stereotypic repetitive movements of oral, buccal, and lingual distribution. Acute withdrawal of chronic antipsychotic drugs in adults can also lead to transient tardive dyskinesia, which disappears within 3 months. These types of movements have been labeled withdrawal dyskinesia (Gardos et al., 1978; Schooler and Kane, 1982).

On the other hand, acute withdrawal of DRBA can precipitate a persistent akathisia (Poyurovsky et al., 1996; Rosebush et al., 1997) or dyskinesia, i.e., tardive akathisia and other tardive syndromes. Acute withdrawal should be avoided because of the propensity to induce TD, and a slow taper and withdrawal should be substituted. Abrupt withdrawal of risperidone therapy in one elderly person resulted in a near-fatal development of respiratory dyskinesia (Komatsu et al., 2005).

Classic tardive dyskinesia

Dyskinesia is a general term referring to abnormal involuntary movements. The term tardive dyskinesia has been used to refer to abnormal movements that are seen as a complication of long-term dopamine receptor antagonist therapy, mainly the type that presents with rapid, repetitive, stereotypic movements involving the oral, buccal, and lingual areas. However, over the years, other types of movements have been noted as complications of dopamine receptor antagonist therapy. These movements have more specific terminologies, such as dystonia and akathisia. Therefore, some authors refer to the type of movements that were originally described as “classic tardive dyskinesia” for lack of a more specific and distinct name for the movements (Fahn, 1989). Some have called them tardive stereotypy (Stacy and Jankovic, 1991; Jankovic, 1994) because of their repetitive, rather than random, nature. However, the stereotypical movements in classic tardive dyskinesia are so characteristic and resemble those seen in almost all patients with this disorder, in contrast to other types of stereotypies that are seen in patients with learning disability, autism, and psychosis, that the term tardive stereotypy does not convey this uniqueness and therefore seems unsatisfactory. Stereotyped hand clasping appears to be a rare form for the presentation of classic tardive dyskinesia (Kaneko et al., 1993).

On the other hand, some have used the term tardive dyskinesia as equivalent to any oral dyskinesia. It therefore needs to be emphasized that the term “tardive” has become synonymous with chronic neuroleptic complications and should be reserved for these disorders. Sustained dystonic movements of the lower face must be distinguished from classic tardive dyskinesia. Frequently, patients on anticholinergics or other medications develop oral dyskinesia with dryness of mouth. Other movements such as myokymia, myoclonus, and tics must be distinguished. The differential diagnosis of oral dyskinesia is summarized in Table 19.10, whereas Table 19.11 compares clinical features of tardive dyskinesia, Huntington disease, and oromandibular dystonia, the three most common forms of oral dyskinesias.

Table 19.10 Differential diagnosis of oral dyskinesias

Clinical features of classic tardive dyskinesia

The clinical features of classic TD are quite distinct from the features of other movement disorders (Fahn, 1984a). The principal site is the face, particularly around the mouth, typically called oral-buccal-facial (O-B-L) dyskinesias. The limbs and trunk are affected less often than the mouth. Even when they are involved, it is usually in addition to involvement of the mouth. The forehead and eyebrows are seldom involved unless tardive dystonia is also present; this is in contrast to Huntington disease, in which chorea of the forehead and eyebrows is more common than choreic movements of the oral musculature. In TD, the mouth tends to show a pattern of repetitive, complex chewing motions (Video 19.3; see also Video 1.73), occasionally with smacking and opening of the mouth, tongue protrusion (flycatcher tongue) (Video 19.4), lip pursing, sucking movements, and fishlike lip puckering movements. The rhythmicity and coordinated pattern of movement are striking. This stereotypic pattern is in contrast to the dyskinesias that are seen in Huntington disease, in which the movements are without a predictable pattern. Usually, the limb involvement is limited to the distal part. Like the mouth region, the movements of the distal limbs show a repetitive pattern, earning the label of piano-playing fingers and toes. When the patient is sitting, the legs often move repeatedly, with flexion and extension movements of the toes and foot tapping. When the patient is lying down, flexion and extension of thighs may be seen. Rhythmic truncal rocking can be seen when the patient is lying, sitting, or standing (Video 19.5). The respiratory pattern can be involved with dyskinesia, causing hyperventilation at times and hypoventilation at other times (Yassa and Lai, 1986). In a study of the breathing pattern in patients with TD, patients had an irregular tidal breathing pattern, with a greater variability in both tidal volume and time of the total respiratory cycle (Wilcox et al., 1994). The presence of respiratory dyskinesia never causes a medical problem, although it might look alarming. Esophageal (associated with lingual) dyskinesias have also been reported, resulting in increased intraesophageal pressure and death due to asphyxiation in one patient (Horiguchi et al., 1999).

The involuntary movements of the mouth in classic tardive dyskinesia are readily suppressed by patients when asked to do so. Furthermore, the movements cease as the patient is putting food in the mouth, when talking, or when a finger is placed on the lips. Since the movements do not interfere with basic functions, patients are often unaware of their movements. When the patient is asked to keep the tongue at rest inside the mouth, the tongue tends to assume a continual writhing motion of athetoid side-to-side and coiling movements. The constant lingual movements might lead to tongue hypertrophy, and macroglossia is a common clinical sign. On command, however, most patients with TD can keep the tongue protruded out without darting back into the mouth for more than half a minute; patients with the chorea of Huntington disease typically cannot maintain a protruded tongue. This inability to sustain a voluntary contraction is called motor impersistence, which is not seen in TD.

Abnormal movements in schizophrenia in the absence of exposure to DRBAs

Tardive dyskinesia as an entity induced by dopamine receptor blockers has been challenged by reports claiming that spontaneous movements are sometimes encountered in patients with schizophrenia who have never been exposed to neuroleptic agents (see the review by Boeti et al., 2003). McCreadie and colleagues (2002) evaluated 37 schizophrenic patients never treated with antipsychotics and followed for 18 months. Nine (24%) had dyskinesia on both occasions, 12 (33%) on one occasion, and 16 (43%) on neither occasion. Twenty-one (57%) had dyskinesia on at least one occasion. Thirteen patients (35%) had parkinsonism on at least one occasion. It is critical that the quality of the dyskinesia be reported, because the classic O-B-L dyskinesias are very distinct and almost specific for tardive dyskinesia, whereas many other types of movements could represent a different disorder. The presence of parkinsonism, though, suggests that the patients had been exposed to neuroleptics, but that the investigators were clueless as to the exposure.

Many authors noted the existence of spontaneous oral dyskinesia occurring in untreated populations and in untreated schizophrenics. McCreadie and colleagues (1996) examined 308 elderly individuals in Madras, India, looking for abnormal movements and found them in 15%. The prevalence of spontaneous dyskinesia has been reported to be as high as 20% in psychiatric or nursing home patients (Brandon et al., 1971), but the average rate is about 5% (Jeste and Wyatt, 1982a; Kane and Smith, 1982). Others noted increasing prevalence of spontaneous dyskinesia with age (Klawans and Barr, 1982). Some studies looking at healthy elderly populations estimate the prevalence rate to be about 1% (Lieberman et al., 1984; D’Alessandro et al., 1986). Blanchet and colleagues (2004) evaluated 1018 (69.3% women) noninstitutionalized, frail elderly subjects attending day care centers to document the prevalence and phenomenology of spontaneous oral dyskinesia. The prevalence rate for spontaneous oral dyskinesia was 3.7% (4.1% for women and 2.9% for men). They reported more frequent ill-fitting dental devices, oral pain, and a lower rate of perception of good oral health compared to nondyskinetic subjects.

The true prevalence rate of spontaneous oral dyskinesia that can mimic TD may be even lower, considering the fact that many other identifiable oral dyskinesias listed in Table 19.10 such as oromandibular dystonia can be difficult to distinguish from classic tardive dyskinesia and may be counted as spontaneous dyskinesia in epidemiologic studies. On the other hand, spontaneous oral dyskinesia resembling the stereotypical O-B-L dyskinesia of classic tardive dyskinesia has been reported in an aged cynomologus monkey (Rupniak et al., 1990). Some patients with oral dyskinesias resembling those of TD and unexposed to DRBAs would seemingly be idiopathic in origin, but careful workup might reveal other etiologies (Table 19.10), including treatment with lithium (Meyer-Lindenberg and Krausnick, 1997) and brainstem infarcts (Fahn et al., 1986).

Reports of oral dyskinesias occurring in schizophrenic patients who have never been exposed to neuroleptics raise the question as to how specifically the O-B-L movements should be attributed to DRBAs and whether or not they could be due to the schizophrenia. Fenn and colleagues (1996) examined 22 never-medicated schizophrenics in Casablanca, Morocco. Three had abnormal movements that are said to be characteristic of TD. Fenton and colleagues (1997) compared the prevalence of spontaneous oral dyskinesias among drug-naive schizophrenics and patients with other psychiatric disorders. They found that dyskinetic movements were more common in the former group. Gervin and colleagues (1998) reported 6 (7.6%) out of 79 first-episode schizophrenics to have spontaneous dyskinesias. Until movement disorder experts can evaluate the movement phenomenology, it is possible that these individuals have some disorder other than TD. Moreover, it is possible that the history of nondrug exposure is faulty. The overwhelming evidence is that the abnormal movements that are seen with DRBAs are due to these drugs.

As was mentioned earlier, a major concern in diagnosing spontaneous oral dyskinesias is the distinction between the stereotypic movements of TD and the movements of oromandibular dystonia. Authors should publish the videotape demonstrations of the abnormal movements so that the medical community can judge whether the movements do indeed fit the phenomenologic criteria of movements seen in TD.

Epidemiology, risk factors, and natural history

Epidemiologic studies looking into the prevalence of TD have been confounded by the factors that affect the detection of the abnormal involuntary movements as well as the variables that affect the prognosis of the movements. Therefore, it is not surprising to find a wide range of prevalence estimations from 0.5% to 65% in the literature (Jeste and Wyatt, 1982a; Kane and Smith, 1982). The prevalence of TD has been noted to have increased from 5% before 1965 to 25% in the late 1970s (Jeste and Wyatt, 1982a; Kane and Smith, 1982). Mean prevalence rates calculated by two different reviewers, however, agree well at around 20% (Jeste and Wyatt, 1982a; Kane and Smith, 1982).

The prevalence rates of spontaneous dyskinesias and dyskinesias in people who have been exposed to neuroleptics have been compared. In the study by Woerner and colleagues (1991) the overall prevalence of spontaneous dyskinesias was 1.3% in 400 healthy elderly people and 4.8% in elderly inpatients, with a range of 0–2% among psychiatric patients who had never been exposed to neuroleptics. These investigators reported prevalence of TD of 13.3% and 36.1% in voluntary and state psychiatric hospitals, respectively. There was an interplay between age and gender. Among younger patients, men had higher rates; among subjects over age 40 years, rates were higher in women. Van Os and colleagues (1999) also found an increased risk for men in the younger population.

In one prospective study examining development of TD with low-dose haloperidol, the 12-month incidence of probable or persistent tardive dyskinesia was 12.3% (Oosthuizen et al., 2003). In a much larger study, the annual incidence rates range from 5% in a younger population (mean age 28) (Kane et al., 1986) to 12% in an older group (mean age 56) (Barnes et al., 1983). Kane and colleagues’ data (1986) also show that the cumulative incidence of TD increases linearly with increasing duration of neuroleptic exposure at least for the first 4–5 years of such exposure. In a subsequent study by these authors, Chakos and colleagues (1996) studied prospectively 118 patients in their first episode of psychosis who were treatment-naive and were then followed for up to 8.5 years while they were on neuroleptics. The cumulative incidence of presumptive TD was 6.3% after 1 year of follow-up, 11.5% after 2 years, 13.7% after 3 years, and 17.5% after 4 years. Persistent TD had a cumulative incidence of 4.8% after 1 year, 7.2% after 2 years, and 15.6% after 4 years. Thus, the earlier findings of Kane and colleagues (1986) of about 5% a year cumulative incidence seem to have been confirmed.

Jeste et al. (1999) looked at the cumulative incidence of tardive dyskinesia after exposure to neuroleptics in patients over the age of 45 years. The mean cumulative incidence was 3.4%, 5.9%, and 22.3% at 1, 3, and 12 months, respectively. Woerner and his colleagues (1998) studied patients over the age of 54 when first exposed to neuroleptics. The cumulative rates of tardive dyskinesia were 25%, 34%, and 53% after 1, 2, and 3 years of cumulative antipsychotic treatment. A greater risk of tardive dyskinesia was associated with a history of ECT treatment, higher mean daily and cumulative antipsychotic doses, and the presence of extrapyramidal signs early in treatment. From both these studies, the incidence rates for patients beginning treatment with conventional antipsychotics in their fifth decade or later are three to five times what has been found for younger patients, despite treatment with lower doses.

Studies have been carried out in other countries. A prospective study of 11 psychiatric facilities in Japan found the prevalence of TD to be 7.6%, the annual incidence rate to be 3.7%, and an annual remission rate to be 28.7% (Inada et al., 1991). On the other hand, Hayashi and colleagues (1996) reported a prevalence of 22.1% in 258 patients receiving neuroleptics. A study in Austria reported the follow-up of the 270 patients still in a psychiatric hospital after 10 years out of an original population of 861 patients; the prevalence rate of TD was 3.7% in 1982 and 12.7% in 1992; the major risk factor for TD was advanced age (Miller et al., 1995). Jeste and colleagues (1995) followed 266 patients over the age of 45 years to determine incidence and prevalence rates of tardive dyskinesia following exposure to neuroleptics, using electromechanical sensors to detect the presence of movements. Cumulative incidence of TD was 26%, 52%, and 60% after 1, 2, and 3 years, respectively. These rates, which are higher than those found by Kane and his colleagues, might reflect the sensitive sensors used, and these could possibly give false-positive findings. The same group compared the development of abnormal movements in the orofacial and limb-truncal areas in these 266 middle-aged and elderly patients treated with neuroleptics and found that the cumulative incidence of orofacial TD was 38.5% and 65.7% after 1 and 2 years, respectively, whereas that of limb-truncal TD was 18.6% and 32.6% after 1 and 2 years, respectively (Paulsen et al., 1996).

Host and treatment factors affect the development, severity, and persistence of TD, thereby resulting in different prevalence rates. Age of the patient has been the most consistent factor that adversely affects the incidence, prognosis, and severity of tardive dyskinesia (Smith and Baldessarini, 1980; Jeste and Wyatt, 1982a; Kane and Smith, 1982; Kane et al., 1986). Possibly the youngest individual who developed O-B-L tardive dyskinesia was a 2-month-old girl after a 17-day treatment with metoclopramide for gastroesophageal reflux (Mejia and Jankovic, 2005). The movements persisted for at least 9 months after the drug was discontinued. This patient is the first documented case of tardive dyskinesia in an infant.

Female sex has been associated with increased prevalence of TD, especially in the older population (Jeste and Wyatt, 1982a; Kane and Smith, 1982; Kane et al., 1988). Several authors noted increased incidence and prevalence among patients with affective disorders compared to schizophrenia or schizoaffective disorders (Gardos and Casey, 1983). Other host factors such as presence of previous brain damage as noted by increased ventricular size remain controversial (Jeste and Wyatt, 1982a; Kane and Smith, 1982). Some noted poor treatment response of schizophrenia to drug treatment as a risk factor for development of tardive dyskinesia (Chouinard et al., 1986). The parameters of drug exposures such as dose, duration, type of neuroleptics, and drug-free intervals have been very difficult to correlate with risk of TD partly because accurate history concerning the drug treatment is usually not available and the drug itself can affect the detection of TD by masking or uncovering it.

Ethnicity has been found to be an important risk factor in both the development and prognosis of tardive dyskinesia, African-Americans being more susceptible than European-Americans (Wonodi et al., 2004a). In a meta-analysis of the literature, the only statistically significant risk factors for TD were non-white ethnicity and having developed drug-induced parkinsonism; the association with older age was suggestive but inconclusive (Tenback et al., 2009).

Prospective studies have noted that the total cumulative drug exposure correlates with incidence of withdrawal tardive dyskinesia (Kane et al., 1985), and development of TD 5 years later in patients who did not have TD initially (Chouinard et al., 1986). Continued use of DRBAs after the appearance of TD also adversely affects subsequent prognosis (Kane et al., 1986). Drug holidays were once advocated to decrease the risk of TD, but an increased number of drug-free intervals was found to worsen the prognosis after withdrawal (Jeste et al., 1979). The risk of developing TD is three times as great for patients with more than two neuroleptic interruptions as for patients with two or fewer interruptions (van Harten et al., 1998). Other than reduced risk with clozapine, olanzapine, and quetiapine, no other particular type of neuroleptic, including depot preparations, has been clearly identified as a risk factor (Yassa et al., 1988). The effect of other drugs such as anticholinergics on the incidence and prevalence of TD is controversial, although anticholinergics may increase its severity (Yassa, 1988). Lithium may decrease the chance of TD development (Kane et al., 1986; van Harten et al., 2008a). Development of parkinsonism tends to predispose patients to TD (Chouinard et al., 1986; Kane et al., 1986). Substance abuse has also been reported to be a risk factor (Bailey et al., 1997). Lacking the cytochrome P450 enzyme required for metabolism exogenous toxins because of nonfunctional alleles of the CYP2D6 gene appears to be a risk factor for developing both TD and drug-induced parkinsonism (Andreassen et al., 1997). Another genetic risk factor is polymorphism in the ATP13A gene, especially for drug-induced parkinsonism (Kasten et al., 2011).

A review of the literature by Correll and colleagues (2004) found a lower incidence with the newer generation of antipsychotics (Table 19.12). A European consortium of investigators looking at 6-month results came to a similar conclusion (Tenback et al., 2005). In their review of the literature, Tarsy and Baldessarini (2006) also suggest that there may be a declining incidence of TD as the second-generation antipsychotics are being more regularly used. On the other hand, in a survey on elderly patients with dementia treated with older or newer generation antipsychotics, no statistical difference was found in the development of nonparkinsonian movement disorders in relation to treatment class (Lee et al., 2005). Children appear to be increasingly treated with “atypical” antipsychotics and in one study, 9% developed TD (Wonodi et al., 2007).

Table 19.12 Annual incidence of tardive dyskinesia comparing the second generation antipsychotics with haloperidol

Data from Correll CU, Leucht S, Kane JM. Lower risk for tardive dyskinesia associated with second-generation antipsychotics: a systematic review of 1-year studies. Am J Psychiatry 2004;161(3):414–425.

Although the majority of TD occurs while patients are on chronic treatment with DRBAs or shortly after discontinuing them, many cases have been reported in which the TD occurred after only a short interval of treatment and persisted. Some authors define TD arbitrarily as dyskinesias that occur after a minimum of 3 months of dopamine receptor antagonist therapy (Schooler and Kane, 1982), but it appears that there is no safe low-incidence period right after the initiation of treatment with DRBAs nor is there any particularly high-risk period. The overall risk appears to accumulate as time goes on, although whether it continues to increase linearly even after the first several years of exposure is not clear from available data. Kang and colleagues’ (1986) retrospective data also show that the cumulative number of patients who developed tardive dystonia increased almost linearly from the first few months.

Probably most patients with TD have mild cases that could improve and disappear over time if the offending DRBAs were withdrawn. In an attempt to look at risk factors for the severe forms of TD, Caligiuri and colleagues (1997) conducted a longitudinal prospective study to determine the incidence of severe TD in middle-aged and elderly psychiatric patients. The cumulative incidence of severe TD was 2.5% after 1 year, 12.1% after 2 years, and 22.9% after 3 years. Risk factors for severe TD were higher daily doses of neuroleptics at study entry, greater cumulative amounts of prescribed neuroleptic, and greater severity of worsening negative psychiatric symptoms.

Once established, TD does not frequently become more severe even though the DRBA is continued (Labbate et al., 1997). Fernandez and colleagues (2001) evaluated the dyskinesia of TD and the clinical features of parkinsonism in 53 patients residing in a state psychiatric hospital over a 14-year period. TD improved and parkinsonism worsened in patients who continued to receive neuroleptic drugs. But the natural history of TD is not easy to determine because the DRBAs that cause the movement disorders also tend to suppress the movements, while they can cause drug-induced parkinsonism.

Movements that disappear with increasing dose or resumption of DRBAs have been called masked TD (Schooler and Kane, 1982). Conversely, movements of TD typically increase after discontinuation of the DRBAs. Transient dyskinesias that appear after withdrawal of DRBAs have been called withdrawal dyskinesias (Gardos et al., 1978). The term covert dyskinesia has also been used for dyskinesia that is not detectable during drug administration and is first noted during drug reduction or withdrawal and becomes permanent (Gardos et al., 1978). Therefore, the difference between withdrawal and covert dyskinesia is in duration. In fact, the spectrum of TD ranges from mild transient withdrawal dyskinesia to severe irreversible TD, which can occur during or after discontinuation of the antipsychotic drugs. These examples illustrate the complexity in the detection of TD in the presence of DRBAs. In an attempt to estimate the false-negative rate of detection for TD in patients taking DRBAs, Kane and colleagues (1988) withdrew the drug from patients who showed no evidence of TD while taking DRBAs. Withdrawal TD was seen in 34% of 70 subjects, with persistence beyond 3 months in 7 of them.

Pathophysiology

Although tardive dyskinesia is an iatrogenic condition with an established etiologic connection to dopamine receptor antagonism, the pathophysiology is not well understood. Any proposed mechanism for tardive dyskinesia will have to explain (1) the cumulative increase of risk with increasing duration of treatment, (2) frequent persistence of the condition despite discontinuation of the treatment, (3) increasing risk with age and other individual dispositions that leave only a portion of the population affected, (4) increased risk with intermittent exposure to DRBAs, and (5) clinical pharmacology of tardive dyskinesia (i.e., improvement of the condition by drugs that depress dopaminergic activity and worsening by increasing dopaminergic activity with levodopa or amphetamine).

Treatment of classic tardive dyskinesia (TD)

Conclusions on classic tardive dyskinesia

By blocking D2 receptors or depleting dopamine from synaptic terminals, neuroleptics are effective in reducing the abnormal involuntary movements of TD. Depletors do not cause TD and are therefore preferred; their use, while controlling symptoms, allows the brain to heal spontaneously. Avoiding side effects, particularly drug-induced parkinsonism, is a major limiting factor. DRBAs, while controlling symptoms, are the culprits causing TD; the brain does not heal, and TD may continue to worsen in the presence of these agents. If one needs to utilize an antipsychotic, it seems safer to use one that is suspected of having less ability to induce TD, such as clozapine and quetiapine. Olanzapine comes closer to being a typical, rather than an atypical, antipsychotic. When clozapine and quetiapine occasionally reduce dyskinesias, the doses that are used suggest the response is from their D2-blocking activity.

Figures 19.1 and 19.2 are flow charts that provide a useful algorithm for treating TD.

Figure 19.1 Flowchart for treating tardive dyskinesia in the absence of psychosis. AMPT, α-methyl-para-tyrosine; DBS, deep brain stimulation; ECT, electroconvulsive therapy; TBZ, tetrabenazine.

Figure 19.2 Flowchart for treating tardive syndromes in the presence of psychosis.

Tardive dystonia

As with classic TD, the most effective medications for tardive dystonia are also antidopaminergic drugs (Kang et al., 1986), but the percentage of patients who improve is smaller. Reserpine and TBZ each produce improvement in about 50% of patients. Some patients who do not respond or have intolerable side effects to one might respond to the other. DRBAs are more effective in suppressing the movements (77%). Symptomatically, those who remained on DRBAs after the onset of tardive dystonia and those who were withdrawn from them do not have a significant difference in their improvement rate. This again is in agreement with the data in classic tardive dyskinesia, in which continued use of DRBAs does not necessarily lead to aggravation of their movements (Casey et al., 1986; Gardos et al., 1988). The atypical antipsychotic clozapine has been helpful in some patients with tardive dystonia (Lieberman et al., 1989, 1991; Van Putten et al., 1990; Friedman, 1994; Trugman et al., 1994; Wolf and Mosnaim, 1994; Raja et al., 1996; van Harten et al., 1996a). There are reports of quetiapine’s effectiveness as well (Gourzis et al., 2005). It is likely that its treatment of tardive dystonia in some situations is due to its D2 receptor-blocking activity resulting in a masking of the symptoms, because withdrawal would exacerbate the dystonia (Krack et al., 1994). The combination of clozapine and clonazepam has been effective in some patients when either drug alone was much less satisfactory (Shapleske et al., 1996).

In tardive dystonia, antimuscarinics are almost as effective as antidopaminergic drugs. This is different from classic tardive dyskinesia, which may get worse with antimuscarinics (Yassa, 1988). Kang and colleagues (1986) reported a 46% improvement rate on antimuscarinics such as trihexyphenidyl and ethopropazine. CNS side effects of the antimuscarinics include forgetfulness, lethargy, psychosis, dysphoria, and personality changes; elderly patients are more susceptible to these. Peripheral side effects include blurred vision, dry mouth, constipation, urinary retention, and orthostatic dizziness. Those who develop side effects to one anticholinergic drug may tolerate another anticholinergic better. Although there is no evidence that one anticholinergic is more efficacious than the others, ethopropazine may produce fewer CNS side effects in elderly patients. Although peripheral pharmacokinetics show relatively short half-lives, their central effects have a very slow onset of action and several weeks are often required before benefit is noticed. Therefore, the medications are started at a low dose, 2.5 mg of trihexyphenidyl or 25 mg of ethopropazine, and increased slowly by 2.5 mg of trihexyphenidyl or 25 mg of ethopropazine weekly until sufficient control of dystonia or intolerable side effects are achieved. As in idiopathic dystonia, many patients respond only to a high dose of anticholinergics. Therefore, every attempt must be made to control side effects so that high-dose anticholinergics may be tried. Kang and colleagues (1986) reported use of a maximum of 450 mg ethopropazine or 32 mg of trihexyphenidyl, and higher doses may be tolerated in young patients if judiciously used. Peripheral side effects are often controlled by peripheral cholinergic drugs, such as oral pyridostigmine and pilocarpine eye drops.

The clinical pharmacology of tardive dystonia indicates two subtypes: one group that responds to antidopaminergic drugs like the other tardive syndromes and one that responds to anticholinergics like idiopathic dystonia. Analysis of clinical characteristics of patients who respond to antidopaminergic drugs and those who respond to anticholinergic drugs has not shown any significant difference (Kang et al., 1986). However, the data from Kang and colleagues (1986) are retrospective and the treatment choice between the two classes of drugs was rather arbitrary, partly based on anticipated side effects. For example, patients who were elderly or who had dementia were treated with dopamine-depleting drugs first, because these patients have greater risk for anticholinergic side effects such as memory loss and confusion. Patients with depression might have relapse of their symptoms on dopamine-depleting drugs and were treated preferentially with anticholinergics. This is a reasonable clinical approach, but characterization of subtypes will require a controlled study with random assignment of the treatment. If either a dopamine depletor or antimuscarinic is ineffective by itself, the combination should be tried.

Benzodiazepines are mainly helpful as adjunctive therapy with dopamine-depleting or anticholinergic drugs, and occasionally can be quite beneficial (Yamamoto et al., 2007). Minimal success with propranolol, levodopa, carbamazepine, and baclofen has been noted. Bromocriptine, deanol, clonidine, lisuride, amantadine, and valproate were reported with mixed results. The calcium channel blocker verapamil was reported to be effective in one patient (Abad and Ovsiew, 1993). Opioids do not have lasting value in suppressing tardive dystonia (Berg et al., 2001). But one study found the combination of naltrexone and clonazepam to offer some benefit (Wonodi et al., 2004b).

If any residual dystonia remains that is localized to one or a few parts of the body, injections of botulinum toxin into the affected parts, such as the orbicularis oculi, masseters, or cervical muscles, might be useful (Chatterjee et al., 1997; Tarsy et al., 1997; Kanovsky et al., 1999).

Electroconvulsive therapy (ECT) might be effective in intractable cases (Yoshida et al., 1996), and deep brain stimulation in the globus pallidus interna is often effective (Franzini et al., 2005; Sako et al., 2008; Gruber et al., 2009), as is intrathecal baclofen (Dressler et al., 1997). Surgery is not always effective and it does pose risks of complications (Trottenberg et al., 2001). However, deep brain stimulation in the pallidum can be safe and effective (Capelle et al., 2010; Chang et al., 2010); it is worth considering when medications fail.

A centrally acting muscle relaxant, eperisone, was successful in treating one patient with tardive dystonia (Nisijima et al., 1998). Eperisone is a beta-aminopropiophenone derivative.

Figures 19.1 and 19.2 depict flowcharts for therapy that can be applied to tardive dystonia, as well as to classic TD.

Tardive akathisia

As was noted earlier, variants of akathisia have been the source of confusion in the literature. Most of the papers do not distinguish between acute and tardive akathisia or usually refer to acute akathisia and not tardive akathisia. Very few studies have focused on the treatment of tardive akathisia.

Tardive akathisia is difficult to treat and does not respond to anticholinergics, which have been reported to help acute akathisia. It and tardive dystonia are the most distressing and disabling features of the tardive symptoms, and their treatment is important. In the study of tardive akathisia by Burke et al. (1989), all of the patients noted that the subjective sensations were distressing. The same study reported that 87% of patients improved on reserpine up to 5 mg/day and 58% on TBZ up to 175 mg/day. In one-third of these patients, the movements were completely suppressed. In this respect the clinical pharmacology is more like that of classic tardive dyskinesia than that of acute akathisia. Opioids were reported to be beneficial (Walters et al., 1986), but the effect has not been persistent (Burke et al., 1989). Electroconvulsive treatment can be effective in those patients whose akathisia has proved to be intractable (Hermesh et al., 1992).

Figures 19.1 and 19.2 depict flowcharts on therapy that can also be applied to tardive akathisia.

Treatment summary

Table 19.14 summarizes a useful approach to treat tardive syndromes in general. The approach to treatment should be earmarked according to whether the patient is psychotic and requires antipsychotic medication, or is not psychotic. For detailed, step-by-step decisions in the treatment of the tardive syndromes, the flowcharts of Figures 19.1 and 19.2 should be helpful.

Table 19.14 Treatment for tardive dyskinesia syndromes

In summary, since the initial description of tardive dyskinesia as an iatrogenic complication of dopamine receptor antagonists, considerable progress has been made in understanding the risks, epidemiology, and clinical subtypes of the condition. However, pathogenesis and characteristics of distinct subtypes are still poorly understood. Management of tardive syndromes must be based on proper understanding of these aspects of the disorder. The only effective and safe antipsychotic agents that do not produce, or rarely produce, tardive dyskinesia appear to be clozapine and quetiapine. Meanwhile, prevention is possible by avoiding indiscriminate use of DRBAs and limiting their use to disorders for which no other type of medication is available or effective. When prevention and withdrawal of the offending drugs for eventual remission fail, symptomatic suppression can be achieved in many patients with pharmacologic treatments. New approaches are needed, and prospective controlled clinical trials with particular attention to clinical subtypes are necessary for better management of these conditions.