Clinicians should carefully evaluate patients to determine whether subjective cognitive complaints reflect true cognitive deficits, psychiatric symptoms, or other deficits (e.g., learning disabilities). Clinicians should recognize common cognitive adverse effects of specific psychotropic agents and reduce dosages or eliminate likely offending agents when possible. Clinicians also should address other exogenous factors that may contribute to cognitive problems (e.g., alcohol or substance abuse, sleep disorders). Psychostimulants (amphetamine, methylphenidate, and modafinil or armodafinil) may aid attentional processing and verbal fluency in patients who are suitable candidates. Dopamine agonists (e.g., pramipexole) also may help attention in some patient groups.
Numerous factors in addition to psychotropic medications can contribute to subjective cognitive complaints and include affective or psychotic symptoms, anxiety disorders or symptoms, substance abuse, attention-deficit disorder, and a variety of comorbid medical conditions. Furthermore, at least in some conditions (e.g., bipolar disorder), subjective cognitive complaints often correlate poorly with objective deficits in cognitive performance, making it essential for clinicians to discern the nature of cognitive complaints beyond patients’ superficial reports of their presence.
Any psychoactive drug may impair judgment, thinking, or motor skills, although demonstrated adverse cognitive effects related to psychiatric medications are relatively circumscribed. Cognitive dulling that is plausibly attributable to sedating psychotropics (e.g., anticholinergic drugs, benzodiazepines, FGAs, SGAs, and certain anticonvulsants [in particular, topiramate]) poses difficult obstacles that are not easily overcome short of dosage reductions, or when necessary, eliminating a causal agent. In the case of topiramate, adverse cognitive effects may be dose-related and typically include psychomotor slowing, word-finding problems, impaired working memory, poor attention and concentration, and decreased verbal and nonverbal fluency (reviewed in Goldberg 2008). In some instances, agents with a high potential for cognitive dulling or disorganization can be replaced with other, more cognitively neutral medications (e.g., replacing the anticholinergic drug benztropine with the dopamine agonist amantadine as an alternative method to counteract extrapyramidal side effects of antipsychotic drugs).
Few if any rigorous data implicate antidepressants as having adverse cognitive effects, with the exception of anticholinergic effects related to TCAs (especially tertiary-amine TCAs such as amitriptyline or imipramine) or paroxetine. In fact, given their potential neuroprotective effects as demonstrated from preclinical studies (e.g., increasing brain-derived neurotrophic factor), nonanticholinergic antidepressants conceivably may have some potential benefit for enhancing neuronal viability and function. Most notably, the novel serotonergic antidepressant vortioxetine has been shown to improve aspects of cognitive functioning in major depression patients independent of its antidepressant effects (McIntyre et al. 2014), potentially mediated in part through its antagonism of the 5-HT7 receptor. Among SGAs, 5-HT7 antagonism may also contribute to the observed improvement in global cognition, visual memory, and visuospatial working memory seen in euthymic bipolar patients who were randomly assigned to receive adjunctive lurasidone (added to existing pharmacotherapy, vs. receiving only treatment as usual) (Yatham et al. 2017).
A summary of known adverse cognitive effects associated with psychotropic medications is provided in Table 17–1.
Anticonvulsants and lithium
In epilepsy studies, reports of subtle adverse effects on learning, delayed visuospatial processing, and impaired visual memory (reviewed in Goldberg 2008).
Subtle dose-related attentional and memory deficits, impaired verbal memory, and delayed decision time (reviewed in Goldberg 2008).
Minimal adverse cognitive effects.
“Concentration disturbance” noted as an adverse event in 2% of epilepsy patients during premarketing studies; clinical trials in bipolar disorder suggest improvement (rather than worsening) of working and verbal memory, verbal fluency, and immediate recall (reviewed in Goldberg 2008); lamotrigine toxicity may manifest with diffuse cognitive impairment (Bouman et al. 1997).
Diminished creativity, associative fluency, and verbal memory; no significant effects on visual memory, attention, executive function, processing speed, and psychomotor performance (Wingo et al. 2009).
No adverse effects on memory seen acutely in healthy volunteers (taking 300–600 mg/day); possible enhanced attention and motor speed, which may decline with dosages ≥1,200 mg/day (reviewed in Goldberg 2008).
Possible marked impairment (not clearly dose related) of global cognitive functioning, including attention, concentration, verbal and nonverbal fluency, processing speed, language skills, working memory, and perception; in epilepsy patients, cognitive deficits were shown to attenuate with time in some studies but not others (reviewed in Goldberg 2008).
Dose-dependent increased reaction time (van Laar et al. 1995).
Possible impaired attention, psychomotor speed and coordination, and memory.
No known adverse effects; may improve processing speed and recall.
Possible impaired attention, arousal, and verbal or nonverbal memory, as well as motor speed and reaction time; anterograde amnesia may be common (reviewed by Buffett-Jerrott and Stewart 2002).
Possible impaired attention, spatial working memory and visuospatial function, processing speed, verbal memory, verbal fluency, set shifting, and other executive functions—all independent of sedative effects or severity of psychopathology in bipolar or schizophrenia patients (reviewed by Goldberg 2008).
Note. MAOI=monoamine oxidase inhibitor; SGA=second generation antipsychotic; SNRI=serotonin-norepinephrine reuptake inhibitor; SSRI=selective serotonin reuptake inhibitor; TCA=tricyclic antidepressant.
Few strategies have been formally studied to address remedies for adverse cognitive effects caused by psychotropic agents. Interest and curiosity surround the potential for cognitive enhancement associated with procholinergic agents (e.g., donepezil or galantamine) or glutamate antagonists (e.g., memantine), although the efficacy of these agents has not been examined for purposes of counteracting cognitive deficits caused by psychotropic agents. In our clinical experience, it is unlikely for procholinergic drugs or glutamate antagonists to ameliorate the cognitive dulling caused by benzodiazepines, anticholinergic drugs, or sedating antipsychotics.
In 2012, a large, 15-year prospective naturalistic study of elderly French patients found an approximate 50% increased risk for developing dementia among those who had taken a benzodiazepine several years before being assessed (Biliotti de Gage et al. 2012). Similarly, a second larger, age- and sex-matched case-control study by the same investigators using Canadian insurance claims data found a 1.3-fold increased risk for Alzheimer’s dementia among patients taking benzodiazepines (particularly longer-acting agents, such as diazepam or flurazepam) for 3–6 months, and a 1.8-fold increased risk after >6 months’ exposure (Biliotti de Gage et al. 2012). While these findings are provocative, the nonrandomized study designs from which they were derived make it impossible to attribute a causal relationship between benzodiazepine use and development of dementia. It is equally plausible that adults who are at greater risk for dementia may be more likely to be prescribed benzodiazepines prior to manifesting clear signs of dementia (such as undiagnosed mood or anxiety disorders, which in themselves may predispose to eventual dementia).
By contrast, a later, 8.3-year follow-up study of 3,433 older adults found only a modest increased risk for developing dementia among ever- versus never-users of benzodiazepines, but no significantly in-creased risk among those with greatest cumulative exposure (Gray et al. 2016). Findings from this latter study also indicated that chronic use of anticholinergic drugs (including tricyclic antidepressants) separately increased the risk for developing dementia by 19%–54% (increasing with duration exposure) (Gray et al. 2015). On the other hand, a separate report from the Baltimore Epidemiologic Catchment Area (ECA) study found no evidence of gross cognitive decline, using the MMSE, among adults (mean age=54 years) who took TCAs over an 11.5-year follow-up period (Podewils and Lyketsos 2002).
Our interpretation of these studies is that benzodiazepines and anticholinergic drugs may indeed pose cognitive hazards, particularly among older adults, but it is difficult to attribute to either drug class an indisputable independent risk alongside the many other known contributing risks for dementia, such as smoking status, untreated hypertension, concomitant medications, substance misuse, and preexisting psychiatric disorders. Cautious observation, rather than dogmatic avoidance, would seem prudent in risk-benefit decision making regarding the long-term use of either drug class.
With respect to other psychotropic drugs linked with adverse cognitive effects, all SGAs carry an FDA warning label identifying the potential for cognitive and motor impairment; however, the absence of head-to-head comparison studies focusing on changes in cognitive function makes it difficult to judge whether some SGAs may be more cognitively sparing than others.
The novel stimulant modafinil, as well as its enantiomer armodafinil, has been the subject of interest as a strategy to counteract sedation and possible adverse cognitive effects, although few formal studies have addressed that specific purpose. The dopamine D2/D3 agonist pramipexole has been suggested to improve attentional processing in euthymic bipolar disorder patients and could potentially be of value for iatrogenic cognitive dysfunction (Burdick et al. 2011).
The oral hypoglycemic agent liraglutide (1.8 mg/day administered subcutaneously) was found in one preliminary open-label trial to improve multiple dimensions of (presumably noniatrogenic) cognitive dysfunction in mood disorder patients (Mansur et al. 2017).
The root and leaf extract of the herb Withania somnifera (also known as ashwagandha, or Indian ginseng), dosed at 500 mg/day, has been reported to improve working memory, reaction time, and social cognition better than placebo in patients with bipolar disorder (Chengappa et al. 2013) and may offer broader utility for safely addressing cognitive complaints in other patient groups.
Delirium and Encephalopathy
Acute mental status changes that involve disorientation, a waxing-and-waning level of arousal, and other focal neurological signs rarely result from psychotropic drugs apart from specific toxicity states (e.g., anticholinergic delirium, serotonin syndrome). Underlying etiologies (including systemic infections or neurological processes and toxic-metabolic states) should be thoroughly evaluated in the assessment of delirium or related acute changes in mental status.
Relatively rare reports exist of individual psychotropic agents or combinations of agents that may cause delirium or encephalopathy. Among the best known of these is the possible adverse interaction between lithium and FGAs, based on a handful of anecdotal case observations beginning in the mid-1970s. Some authors of retrospective case reviews subsequently suggested that dose-dependent neurotoxicity (notably, delirium with extrapyramidal signs, and more rarely, cerebellar signs) could result from lithium, antipsychotics, or their combination. Modern interpretations of this anecdotal literature have been tempered by vast experience for several decades with the safe and effective use of combination therapy involving lithium with FGAs and SGAs. The possibility of idiosyncratic neurotoxic interactions between these drugs is considered remote.
Clinicians should be aware, however, of more common causes of delirium or acute encephalopathic states; these include drug toxicities, acute cerebral events, hypoxia, infection, and neoplasm. Anticholinergic delirium typically manifests with associated signs of antimuscarinic (M1) anticholinergic toxicity, commonly described as follows:
“Mad as a hatter” (i.e., altered mental states),
“Dry as a bone” (i.e., dry skin and mucous membranes),
“Red as a beet” (i.e., flushing),
“Blind as a bat” (i.e., mydriasis with loss of accommodation),
“Full as a flask” (i.e., urinary retention), and
“Hot as a hare” (i.e., fever).
Anticholinergic delirium must be recognized in patients receiving drugs such as diphenhydramine, TCAs, antispasmodics (e.g., dicyclomine), low-potency FGAs, histamine H2 blockers, and some calcium channel blockers (e.g., nifedipine). Certain systemic neurological drug reactions—notably, serotonin syndrome—also may manifest with delirium.
Rare reports (approximately 2%) exist of profound somnolence following intramuscular administration of the long-acting injectable formulation of olanzapine (termed “post-injection delirium/sedation syndrome [PDSS]”), characterized by sedation or somnolence, confusion, dysarthria, dizziness, and disorientation. The phenomenon typically occurs within 1 hour of injection and characteristically resolves spontaneously within 72 hours.
Fatigue and Sedation
Sedation from psychotropic drugs, particularly those that are antihistaminergic, may or may not tolerize with time or dosage reductions. Effective management involves identifying and eliminating or managing additional potential causes of sedation (e.g., sleep apnea, alcohol abuse, concomitant pharmacotherapies, comorbid medical conditions). Adjunctive stimulants, including modafinil, armodafinil, amphetamine, or methylphenidate, may help to alleviate persistent or significant sedation in patients for whom stimulants may be safe and appropriate short-term adjunctive treatments.
Fatigue encompasses a wide range of phenomena that may include the soporific effects of a drug (e.g., drowsiness, somnolence, sedation), loss of energy, and physical weakness independent of level of arousal. In premarketing FDA registration trials, MedDRA terms such as sedation or somnolence are differentiated by a matter of relative degree. Most randomized trials do not report information on the longitudinal trajectory of sedation (i.e., whether it is an early or late phenomenon, whether it plateaus over time), variations in severity, or timing during the day (e.g., sedation arising shortly after administration vs. excessive daytime sleepiness following nighttime administration).
A summary of reported incidence rates of sedation or somnolence across major psychotropic drug classes in FDA registration trials is presented in Table 17–2. For comparative purposes, Table 17–2 also includes information regarding drug-associated insomnia, a topic further discussed in the section “Insomnia” in Chapter 19, “Sleep Disturbances.”
Histamine H1 antagonism is considered to be one of the most common causes of sedation or somnolence caused by psychotropic agents. Some authors have proposed that tolerance may develop to the sedating effects of H1 blockade over time during continued treatment with psychotropic drugs that have antihistaminergic properties; however, in our experience this often does not routinely occur.
Adjunctive psychostimulants likely represent the most obvious pharmacological strategy to counteract sedation from other psychiatric medicines. Many clinicians consider the wakefulness-promoting agent modafinil (or armodafinil) to be among the safest and best tolerated therapeutic options for treating somnolence or cognitive dulling. However, randomized trials have generally failed to demonstrate an advantage for adjunctive modafinil over placebo in diminishing either fatigue or cognitive functioning in patients with schizophrenia. In studies of patients with major depression, open-label trials have suggested some value for adjunctive modafinil in reducing associated symptoms of fatigue and sleepiness, although subsequent randomized placebo-controlled trials have failed to replicate earlier open-label findings. Some studies also suggest an advantage for adjunctive modafinil over placebo during the first few weeks of antidepressant treatment, but these effects may attenuate over time.
Somnolence or sedation
32% (Equetro in bipolar mania; 12% in open-label extension)
≤5% (Equetro in bipolar mania)
17% (migraine), 19% (bipolar mania), 27% (epilepsy)
≤5% (all indications)
21% (postherpetic neuralgia)
>1% (but ≤placebo)
9% (bipolar maintenance), 14% (epilepsy)
6% (epilepsy), 10% (bipolar maintenance)
20%–36% (epilepsy; apparent dose relationship)
2%–4% (epilepsy; no apparent dose relationship)
15% (epilepsy; no apparent dose relationship)
4% (epilepsy; no apparent dose relationship)
2%–3% (MDD, 300–400 mg/day)
11%–16% (MDD, 300–400 mg/day); 20% (seasonal affective disorder, 150–300 mg/day)
4%–12% (MDD, 50–400 mg/day)
9%–15% (MDD, 50–400 mg/day)
6% (MDD), 13% (GAD)
9% (MDD), 12% (GAD)
13%–17% across disorders
16%–33% across disorders
32%–35% across disorders
26%–27% across disorders
19%–24% across disorders
18%–24% across disorders
13%–15% across disorders
12%–28% across disorders
15%–24% across disorders
11% (across acute adult indications)
18% (across acute adult indications)
13% (schizophrenia), 24% (bipolar disorder) in acute trials
6% (bipolar mania), 16% (schizophrenia)
2%–3% (schizophrenia), 4%–6% (MDD)
<2% (all indications)
7%–8% (bipolar mania), 5%–10% (schizophrenia)
8%–9% (bipolar mania), 11%–13% (schizophrenia)
9%–15% (schizophrenia, across doses)
22% (combined data for 20–120 mg/day over 6 weeks), 13%–15% (adolescent schizophrenia)
8% (combined data for 20–120 mg/day over 6 weeks)
29% (across acute trial indications)
12% (across acute trial indications)
6%–11% (acute schizophrenia, across doses); 9%–26% (adolescent schizophrenia); 12% (schizoaffective disorder, across doses)
<2% (across indications)
Quetiapine XR: 25% (acute schizophrenia), 50% (bipolar mania), 52% (bipolar depression)
Quetiapine XR: ≥1% (acute mania or schizophrenia), 9% (long-term placebo-controlled schizophrenia trials)
2%–7% (acute schizophrenia, across doses); 5% (1–6 mg/day, as monotherapy for bipolar mania)
25%–32% (acute schizophrenia. across doses); 4% (adjuvant therapy in acute bipolar mania)
14% (schizophrenia acute trials), 31% (bipolar acute mania trials)
Note. GAD=generalized anxiety disorder; MDD=major depressive disorder; SGA=second-generation antipsychotic; XR=extended release.
aRefers specifically to excessive daytime sedation.
In our experience, initial sedation that persists from antihistaminergic medications can be effectively managed with adjunctive modafinil dosed at 100–300 mg/day (in one or two divided doses) or armodafinil dosed at 150–250 mg/day (in one or two divided doses). Traditional psychostimulants such as methylphenidate or amphetamine are also sometimes considered as viable adjunctive treatments to counteract iatrogenic sedation from other medications, although concerns about abuse potential as well as tolerance and dose-related sympathomimetic or psychotomimetic effects from traditional stimulants sometimes limit enthusiasm for their use for such purposes. One must also recognize that because (ar)modafinil potently inhibits CYP2C19, it may increase blood levels of other psychotropic drugs that are substrates for this enzyme, such as clozapine, citalopram, escitalopram, levomilnacipran, amitriptyline, and nortriptyline (see Chapter 2, Table 2–7).
Headaches are common, nonspecific side effects that often occur initially and transiently with many psychotropic agents and are best treated, if necessary, with over-the-counter analgesics as needed. Persistent headaches merit careful evaluation for other (i.e., noniatrogenic) etiologies. Psychotropic agents that are suspected of causing persistent headaches (notably, lamotrigine or some SSRIs) may warrant discontinuation for the purposes of diagnostic clarification as well as relief of the presumed side effect.
Headache is among the most commonly reported adverse effects in both active drug and placebo arms in randomized trials of psychotropic compounds. The often nonspecific nature of headache can pose some difficulties in determining its iatrogenic from noniatrogenic etiologies, particularly in the absence of other neurological or systemic symptoms. Incidence rates of headache from controlled trials of common psychotropic drugs are reported in Table 17–3.
Importantly, because headache is among the most frequently occurring of all adverse effects from placebo (see the section “The Nocebo Phenomenon and Proneness to Adverse Effects” in Chapter 1, “The Psychiatrist as Physician”), the incidence rates with active drug reported from controlled trials in Table 17–3 likely highly overestimate true drug effects. For example, in the case of escitalopram for GAD, the active drug incidence rate of 24% is counterbalanced by a nocebo incidence rate for headache of 17%.
Incidence of 19% in acute (≤18-week) child and adolescent ADHD studies (cf. 15% with placebo).
Incidence of 6% in FDA registration trials.
Incidence of 22% in manufacturer’s 6-month open-label trial of Equetro in bipolar disorder.
Incidence of 31% in FDA registration trials as adjunctive therapy for complex partial seizures (cf. 21% incidence with placebo); reported incidence >5% but no different from placebo in trials for acute mania.
Incidence of 3.3% in postherpetic neuralgia (monotherapy); >1% (but no different from placebo) as add-on therapy in epilepsy.
Incidence of ~30% of migraine-like headaches in clinical trials for bipolar disorder; however, case reports also support efficacy of lamotrigine to prevent migraine with aura.
Incidence of 26%–32% (no clear dose relationship) in add-on therapy studies for epilepsy.
Incidence of 25% (bupropion SR 400 mg/day)–26% (bupropion SR 300 mg/day) (cf. 23% incidence with placebo) and 34% (bupropion XL across doses; cf. 26% with placebo).
Incidence of >2% but less than seen with placebo in FDA registration trials for major depression.
Incidence of 20%–22% (doses of 50–100 mg/day) in FDA registration trials for major depression (no different from placebo).
Incidence of 14% across indications in FDA registration trials (no different from placebo).
Incidence of 24% in FDA registration trials for GAD; in FDA registration trials for MDD, incidence of ≥2% but comparable or higher with placebo than escitalopram; specific incidence rate in MDD trials not reported.
Approximate incidence of 20% in FDA registration trials for all indications.
Incidence of 22% in FDA registration trials for combined adult OCD and major depression.
Not reported (aside from a general risk for transient headaches with abrupt cessation).
Incidence of ≥1% in FDA registration trials.
Approximate incidence of 18% in FDA registration trials for GAD or major depression.
Approximate incidence of 25% in FDA registration trials for all indications.
Incidence of 38% in social anxiety disorder (but no different from placebo across other indications).
Migraine reported in ≥1% in FDA registration trials for major depression.
Not reported (aside from a general risk for transient headaches with abrupt cessation).
Note. ADHD=attention-deficit/hyperactivity disorder; FDA=U.S. Food and Drug Administration; GAD=generalized anxiety disorder; MDD=major depressive disorder; OCD=obsessive-compulsive disorder; SR=sustained release; XL=extended release.
In patients taking lithium, chronic headaches have been associated, rarely, with pseudotumor cerebri, a syndrome that also involves bilateral papilledema and increased intracranial pressure on lumbar puncture but no localized neurological signs or structural abnormalities visible on neuroimaging.
Akathisia and Extrapyramidal Adverse Effects
Centrally acting β-blockers (e.g., propranolol) or benzodiazepines remain the most evidence-based adjunctive strategies to counteract antipsychotic-induced akathisia. Preliminary studies also support the adjunctive use of gabapentin, trazodone, and mirtazapine, as well as amantadine, although benefits with amantadine may be transient. The use of anticholinergic drugs such as benztropine is less well-established for akathisia than for pseudoparkinsonism.
Extrapyramidal adverse effects broadly encompass abnormal motor movements that originate outside of the motor cortex, including akathisia and parkinsonian symptoms, as well as bradykinesia, akinesia, tremor, choreiform movements, and dystonias. Akathisia—either the objective manifestation or the subjective experience of physical restlessness—is a common, usually dose-related problem related to treatment with most dopamine antagonists. It has in some instances been linked with violence and suicidal thinking or behavior. Tardive akathisia describes chronic akathisia, lasting for at least 1 month and often persisting for months or even years after antipsychotic discontinuation. A handful of case reports also suggest that true akathisia may also be inducible by lithium, mirtazapine (with chronic use), and some SSRIs, possibly in dose-related fashion (presumably via striatal dopamine antagonism resulting from the inhibitory effects of serotonin). Pseusoparkinsonism describes EPS resembling Parkinson’s disease, usually caused by antipsychotic-induced disruption of dopaminergic tracts in the nigrostriatal pathway (which, when intact, normally modulate motor coordination through the extrapyramidal system). Rare case reports of de novo parkinsonian symptoms have also been described among epilepsy patients taking either divalproex or lamotrigine, while other reports suggest efficacy for lamotrigine to treat parkinsonian symptoms in bipolar disorder patients. Sertraline is thought to be the most potent dopamine reuptake inhibitor among the SSRIs, and as such, its potential for causing adventitious movements may be minimal.
Incidence rates of akathisia and other forms of EPS from antipsychotics have been reportedly lower with agents that dissociate rapidly and are thus considered “loose” binders of the dopamine D2 receptor (Kapur and Seeman 2001), such as quetiapine and clozapine, and higher among agents with “tight” D2 receptor binding affinities, such as risperidone and aripiprazole (Table 17–4). Patient-specific risk factors for akathisia vary across studies and in some, but not all, may include older age, female sex, negative symptoms, cognitive deficits, and affective symptoms. Akathisia also may be more likely to occur in patients taking two or more antipsychotics simultaneously (Berna et al. 2015), or during combination therapy with mood-stabilizing drugs or antidepressants in bipolar disorder or major depression, respectively, than might otherwise occur as monotherapy. EPS from high-potency antipsychotics might be less likely to occur when significant initial improvement is evident within the first 2 weeks of treatment (Rasmussen et al. 2017).
Physiological mechanisms to explain the pathogenesis of akathisia are complex and not fully understood, but prevailing theories include hypodopaminergic activity in the ventral striatum, with compensatory upregulation of noradrenergic terminals from the locus coeruleus to the shell of the nucleus accumbens and prefrontal cortex. In other words, striatal dopamine blockade may lead to diminished noradrenergic tone, likely accounting for the potential benefits of noradrenergic agents.
Centrally acting lipophilic β-blockers such as propranolol have become traditional cornerstones of treatment for akathisia, alongside benzodiazepines. Propranolol nonselectively antagonizes both β1 and β2 CNS receptors and is usually dosed from 30 to 90 mg/day in two or three divided doses. Betaxolol, another centrally acting antagonist that is selective for β1 receptors, dosed from 10 to 20 mg/day, has shown efficacy comparable to that of propranolol (20–40 mg/day), suggesting possible specificity of β1 receptors in the mechanism of akathisia (Dumon et al. 1992; Dupuis et al. 1987). By contrast, non–centrally acting hydrophilic β-blockers that do not cross the blood-brain barrier (e.g., atenolol, metoprolol, nadolol, sotalol) show little efficacy in akathisia (Dumon et al. 1992; Dupuis et al. 1987). One must obviously avoid β-blockers (particuarly β2 blockers) in patients with asthma, bradycardia, or sick sinus syndrome, although β1-blockers (which are cardioselective) pose no bronchopulmonary hazard.
10%–13% in acute monotherapy trials across indications; 19%–25% when added to lithium or divalproex (bipolar disorder) or antidepressants (major depression)
4%–11% across indications
4%–7% in pooled schizophrenia trials, dosed 1–4 mg/day; 4%–14% in pooled MDD adjunctive trials, dosed 1–3 mg/day
9%–14%in acute trials for schizophrenia (dosed 1.5–6 mg/day); 20% or 21% in acute trials for bipolar mania (dosed at 3–6 mg/day or 9–12 mg/day, respectively)
1.7%–2.3%; parkinsonism in 0.2%–0.3%
6%–22% in acute trials for schizophrenia
3% in acute trials across indications
6%–9% in acute schizophrenia trials
1%–4% across indications
5%–9% (across indications); parkinsonism in 12%–20% across indications
8%–10% (across indications); other extrapyramidal symptoms in 14%–31% across indications
Note. BD=bipolar depression; BM=bipolar mania; MDD=major depressive disorder; NNH=number needed to harm; NR=not reported; SZ = schizophrenia.
aBased on manufacturers’ product information.
Increased noradrenergic tone also can arise from α1β agonism, which likely accounts for the extremely low incident rate of akathisia (but also high rate of orthostatic hypotension) with iloperidone.
Among dopamine agonists, amantadine, dosed from 100 to 200 mg bid, has been shown to rapidly improve akathisia, but its effects may dissipate within several weeks of initiation. Its value may be more robust in pseudoparkinsonism than in akathisia, and as an alternative to benztropine, diphenhydramine, or trihexyphenidyl, it spares the cognitive and adverse effects of anticholinergic drugs. Preliminary controlled data also support use of rotigotine (mean dose=~3 mg/day; range= 2–8 mg/day) for akathisia and extrapyramidal signs.
Case reports have suggested potential value with adjunctive gabapentin up to 1,200 mg/day for antipsychotic-induced akathisia, as well as instances of akathisia induced by abrupt cessation of gabapentin.
Benzodiazepines are also commonly used to manage acute akathisia. A 2002 Cochrane Database analysis based on two small randomized trials (N=27) found clonazepam superior to placebo for reducing symptoms of akathisia within 7–14 days of initiation (Resende Lima et al. 1999).
Akathisia associated with SSRIs as well as some antipsychotics has been hypothesized to result from undesired 5-HT2A agonism, prompting interest in the use of 5-HT2A antagonists such as trazodone (50–100 mg/day) or mirtazapine (15 mg/day); however, case reports also link mirtazapine with causing akathisia.
Some investigators have reported low serum iron levels in association with chronic akathisia, with case reports of improvement following intravenous iron therapy.
Use of the antiserotonergic drug cyproheptadine (16 mg/day) also has demonstrated efficacy comparable to that of propranolol (up to 80 mg/day) for treating akathisia in preliminary randomized trials.
Preliminary randomized controlled data show high tolerability and greater improvement in antipsychotic-associated parkinsonism with adjunctive modafinil (50–200 mg/day) than placebo (Lohr et al. 2013).
Vitamin B6 dosed at 600 mg twice daily also has been reported in small, preliminary randomized trials to ameliorate subjective, but not objective, restlessness and distress associated with akathisia (Lerner et al. 2004).
Finally, anticholinergic drugs such as benztropine are sometimes used in the treatment of akathisia, although their value for this intended purpose appears less well established than in the amelioration of antipsychotic-induced pseudoparkinsonism. Indeed, a 2006 Cochrane Database review found no relevant randomized controlled trials from which to draw broad recommendations or that support or refute the efficacy of anticholinergic drugs to treat akathisia (Rathbone and Soares-Weiser 2006).
Acute dystonic reactions should be treated promptly with an oral or intramuscularly dosed anticholinergic drug such as diphenhydramine 50–100 mg or benztropine 1–2 mg. Dystonias typically occur due to excessive dosing of high-potency antipsychotics but may also reflect motor sensitivity to usual dosages in some patients.
Dystonia refers to “sustained or intermittent muscle contractions causing abnormal, often repetitive movements, postures or both” that are “typically patterned, twisting, and (possibly) tremulous… often initiated or worsened by voluntary action” (Albanese et al. 2013, p. 866). Involuntary spasmodic “pulling” of large muscle groups also is often a descriptive term. Acute dystonic reactions to antipsychotic drugs most often occur within the first week of starting (or raising the dose of) a dopamine-blocking drug. In some studies, younger age and male sex have been reported as possible risk factors. Dystonic reactions can, rarely, be life-threatening (e.g., due to airway obstruction from laryngospasm) and require urgent attention.
Administration of anticholinergic medications is the usual treatment of choice. For acute dystonic reactions, this may involve benztropine 1–2 mg or diphenhydramine 1–2 mg/kg (up to 100 mg); in children, dosing is typically 0.02 mg/kg. Alternatively, biperiden 1–5 mg IM typically renders relief within 20 minutes of administration. Intramuscular or slow intravenous administration of anticholinergic drugs renders relief faster than oral administration. Dosing may be repeated after 10–30 minutes if no signs of response are evident. Parenteral (IV but not IM) diazepam 5–10 mg is sometimes used if the interventions described above fail to relieve symptoms promptly. Continued oral dosing of an anticholinergic drug (e.g., benztropine 1–2 mg twice daily) for several days is then usually recommended to prevent recurrence.
Types of acute focal dystonic reactions to antipsychotic drugs include the following:
Spasticity: dystonic reaction in which truncal and sometimes limb muscles develop prolonged increased tone (becoming tight or stiff) that may be painful
Torticollis and retrocollis: movements involving turning the head to one side or backward, respectively
Laryngospasm: sudden contraction and spasm of the vocal cords.
Opisthotonos: often painful hyperextension of the neck, and possibly the back, causing a marked overarching posture
Oculogyric crises: dystonic reactions involving a sustained and involuntary upward deviation of the eyes due to spasm of the extraocular muscles. It may be caused by FGAs or SGAs or by abrupt cessation of antipsychotic agents, and more rarely of other psychotropic agents, including lithium, carbamazepine, oxcarbazepine, amantadine, SSRIs, and benzodiazepines, among other medications. Antipsychotic-induced oculogyric crises have been reported to occur, often in association with autonomic features (e.g., flushing, sweating) and transient exacerbations of psychotic symptoms (e.g., hallucinations, delusions, catatonia). Oculogyric crises are typically self-limited, but their resolution may be hastened by anticholinergic medications such as benztropine or diphenhydramine. Case reports also suggest that oculogyric crises caused by FGAs or some SGAs may not necessarily resolve solely by discontinuation of the antipsychotic but may improve after the initiation of quetiapine. It is unknown whether recurrent or chronic antipsychotic-induced oculogyric crises increase the probability of developing long-term movement disorders such as tardive dyskinesia. Relatively long-acting benzodiazepines, such as clonazepam, may also provide benefit for alleviating antipsychotic-induced oculogyric crises, with clonazepam being among the best studied in case reports (Horiguchi and Inami 1989; Viana Bde et al. 2009).
Identification and elimination of the causal agent are advisable if the symptom produces distress. Alpha agonists such as guanfacine or clonidine or low-dose antipsychotics are generally considered the treatment of choice to counteract tics.
Tics are repetitive, intermittent hyperkinetic movements that characteristically can be voluntarily suppressed and are usually associated with an urge to perform the movement. They rarely may be caused by certain psychotropic medications. Obviously, the clinician’s first task is to affirm whether the development of a new, sudden repetitive movement or vocalization likely represents a tic (as opposed to a compulsive behavior or motor neuron disease) and whether other more primary etiologies (e.g., Tourette’s syndrome, usually in individuals under age 18; head trauma; stroke; infection) are plausible explanations. Tics commonly include coughing, throat clearing, grunting, sniffing, blinking, and head jerking, and they may include more complex behaviors such as shouting or touching objects or people. Rare neurological conditions such as gelastic seizures (sudden paroxysms of laughter) or chorea may also pose unusual symptoms that require differentiation from tics and determination of a primary neurological etiology versus a secondary iatrogenic phenomenon.
Psychotropic drugs that have been reported to cause (or exacerbate) tics most recognizably include stimulants (i.e., amphetamine and methylphenidate) but may also more rarely include bupropion, sertraline, fluoxetine, imipramine, and certain anticonvulsants (notably, carbamazepine and lamotrigine). Although the manufacturer’s product information for atomoxetine states that the drug does not worsen tics in patients with ADHD and comorbid Tourette’s syndrome, such cases have been reported. In the great majority of case reports involving the emergence or exacerbation of tics with each of the aforementioned medications, subjects were usually those with preexisting tic disorders. Increased dopaminergic tone is thought to contribute to pharmacologically induced or exacerbated tics; mechanisms by which some anticonvulsants may cause tics are less well understood but are thought to involve antiglutamatergic effects that may affect motor control. In general, the recommendation is to discontinue a medication if it is believed to cause or exacerbate a tic.
Alpha agonists such as clonidine (0.1–0.3 mg/day) or guanfacine (0.5–3.0 mg/day) have demonstrated efficacy in tic disorders in children and adolescents, although extended-release guanfacine may show only a modest effect. Outcomes with alpha agonists for tic treatment in adults are less well established. Data also exist to support the utility of both FGAs (e.g., haloperidol, fluphenazine) and SGAs (e.g., aripiprazole) for treatment of tics, again with a stronger evidence base in youth than in adults.
Pharmacotherapies intended to counteract or treat iatrogenic motor tics are generally not undertaken, although medications used to suppress tics include α2-adrenergic agonists such as clonidine or guanfacine, benzodiazepines such as diazepam, antipsychotic agents (haloperidol and pimozide being among the best studied), and possibly donepezil (2.5–10 mg/day). In children with ADHD and chronic tics, desipramine suppresses tics while reducing ADHD symptoms better than placebo (Spencer et al. 2002), suggesting a broader role for TCAs in managing tics. Of note, discontinuation of any of these agents (or poor adherence) could lead to the reemergence of tics that are no longer being suppressed.
Restless Legs Syndrome
Adjunctive benzodiazepines or dopamine agonists (e.g., pramipexole, ropinirole) may help to curtail restless legs caused by psychotropic agents. Restless legs caused by FGAs or SGAs may be a manifestation of dose-related akathisia, which may improve with dosage reductions or, if necessary, a change to an alternative within-class agent that may be associated with a lower incidence of akathisia.
Restless legs syndrome (RLS) is sometimes regarded as a form of dyskinesia that reportedly may be provoked by a number of medications in addition to dopamine antagonists, including mirtazapine (particularly when coadministered with tramadol or dopamine antagonists; S. W. Kim et al. 2008), escitalopram, and citalopram. Some authors have suggested that RLS induced by SSRIs may reflect the consequences of SSRI-induced downregulation of dopamine tone in the basal ganglia, for which replacement with bupropion (as a dopaminergic alternative antidepressant) may be useful. Some SGAs have been reported to cause RLS or periodic limb movements during sleep that occur independently of other motor abnormalities and potentially in a dose-related fashion. RLS has been linked with olanzapine and risperidone. Treatment with benzodiazepines or dopamine agonists (e.g., pramipexole or ropinirole) may not be efficacious to counteract RLS induced by SGAs, although changing to an alternative SGA with loose D2 receptor binding affinity, such as quetiapine, may be ameliorative. Anticholinergic drugs are not known to ameliorate RLS despite the likely pathogenic role of the basal ganglia and related extrapyramidal structures. Gabapentin enacarbil, a pro-drug with nearly twofold greater bioavailability than gabapentin, is an established treatment for primary RLS but has not been studied as a remedy to counteract RLS secondary to psychotopic medications.
Tardive dyskinesia (TD) is a potentially severe and sometimes irreversible movement disorder caused by long-term use of antipsychotic drugs. The emergence of early signs of TD (e.g., involuntary oral-buccal movements) should prompt the limited use—if not complete cessation—of antipsychotics if clinically feasible, balanced against the risk for worsening underlying psychosis or other psychopathology. A lower incidence of TD may occur with clozapine, quetiapine, or olanzapine than with other antipsychotics. Valbenazine and deutetrabenazine are both human vesicular monoamine transporter type 2 inhibitors that are FDA approved for the treatment of TD. Data from preliminary randomized controlled trials also suggest safety and potential benefit with vitamin B6, Ginkgo biloba, levetiracetam, melatonin, and amantadine.
Tardive dyskinesia is a hyperkinetic, often complex and irregular involuntary movement disorder involving symptoms developing after >3 months’ cumulative exposure to antipsychotics or >1 month if older than 60 years (American Psychiatric Association 2013). In a majority of cases, TD involves oral-buccal regions (characterized by tongue thrusts and lip smacking or pursing) and also may encompass abnormal involuntary movements of the head and neck, trunk, and upper and lower extremities. TD movements may be described as
Choreiform (i.e., abrupt, irregular, nonrepetitive, and nonrhythmic) in all regions, including the tongue when protruded. Fingers may move nonrhythmically as if in a “piano player” fashion.
Athetotic, involving slow, continuous, sinewy or serpentine-like writhing.
Focal dystonias, including blepharospasm (i.e., frequent blinking or sustained eyelid closure) and torticollis of the head, neck, and shoulders; botulinum toxin injections may be indicated for blepharospasm if its severity essentially renders a patient functionally blind.
Tardive dystonia is considered a subtype of TD; it manifests with prolonged, nonrhythmic contractions of specific muscle groups that involve increased tone and spasmodic contortions. (In contrast, dyskinesias involve more rhythmic movements of large muscle groups, without increased motor tone.)
The incidence of TD has been estimated at 3%–5% per year during treatment with FGAs; the risk during treatment with SGAs is thought to be somewhat lower but still observable (with a weighted annual mean incidence of 0% in children, 0.8% in adults, 6.8% in adults plus older adults combined, and 5.4% in adults over age 54, according to some estimates; Correll et al. 2004). Although TD risk is widely thought to be lower with newer-generation antipsychotics, meta-analyses suggest that point prevalence rates are substantial with both FGAs (~30%) and SGAs (~21%), independent of age (Carbon et al. 2017). Clinicians must also keep in mind that lifetime risk is cumulative and may be higher in patients taking SGAs as a result of past exposure to FGAs.
Established risk factors for TD include the following:
Duration of antipsychotic exposure
Diagnosis of an affective disorder
African, African American, or Afro-Caribbean race has been linked with a higher risk for developing TD during antipsychotic use; however, these reported associations may be artifacts of higher antipsychotic dosages and durations. Diabetes mellitus and cigarette smoking also have been reported in some, but not all, studies as TD risk factors among antipsychotic recipients. Some studies also suggest that akathisia or EPS in themselves may be a risk factor for developing TD. HIV+ patients have anecdotally been reported to be at higher risk for developing TD from antipsychotics, presumably due to basal ganglia viral penetration involvement.
TD is often irreversible. It can sometimes be masked at least temporarily by increased doses of an antipsychotic. Proposed mechanisms behind TD include dopamine supersensitivity as well as neurotoxic effects related to oxidative stress. A key pharmacological management approach is to reduce the availability of presynaptic central dopamine. One such strategy involves tetrabenazine, which reversibly inhibits vesicular monoamine oxidase 2 (VMAT2), the intracellular apparatus for transporting cytosolic dopamine quanta to axonal terminals for presynaptic release. Less available dopamine in the synaptic cleft is thought to diminish postsynaptic supersensitivity to striatal dopamine release, thereby reducing symptoms of TD. Reserpine, known for its antihypertensive and potential depressogenic effects, is another VMAT2 inhibitor that has been shown to improve TD symptoms in small preliminary trials. Unlike tetrabenazine, reserpine acts irreversibly and binds to both VMAT1 and VMAT2, and also causes peripheral as well as central monoamine depletion as well as orthostatic hypotension. In principle, VMAT2 inhibitors could also treat psychosis—as is, in fact, the case with reserpine—but dosages needed to potentially exert antipsychotic effects would likely pose problems with psychotropic tolerability.
Tetrabenazine was FDA approved in 2008 to treat Huntington’s chorea and is occasionally used off-label to treat TD. However, its short half-life requires multiple daily doses, and it carries nontrivial risks for causing or exacerbating depression and suicidal thinking or behavior (■) in Huntington’s disease, although this labeled warning has not been extended to patients with TD. Tetrabenazine also has been shown to prolong the QTc by about 8 msec on ECG.
In April 2017, the FDA approved the reversible VMAT2 inhibitor valbenazine for the treatment of TD. Tetrabenazine is metabolized by carbonyl reductase to four distinct stereoisomers of dihydrotetrabenazine, of which only one (R,R,R-dihydrotetrabenazine, or [+]-α-dihydrotetrabenazine) has very high binding affinity for VMAT2. Valbenazine is the valine-esterified analogue of tetrabenazine; peripheral hydrolysis of the valine moiety yields R,R,R-dihydrotetrabenazine, the aforementioned highly active [+]-α-dihydrotetrabenazine isomer (see Figure 17–1). Valbenazine itself has only modest VMAT2 binding affinity (Ki=~150 nM), but its [+]-α-dihydrotetrabenazine metabolite has high VMAT2 selectivity and binding affinity (Ki=~4 nM) (Grigoriadis et al. 2017). As compared with tetrabenazine, valbenazine’s longer half-life (15–22 hours) permits once-daily dosing, and the molecule carries a much lower risk for inducing depression or suicidality, or cardiovascular effects, presumably because of its extremely weak binding affinity for D1, D2, 5-HT1A, α1 or α2 adrenergic receptors, or for the serotonin, dopamine, or norepinephrine transporters. Valbenazine can prolong the QTc interval if coadministered with a strong CYP2D6 or 3A4 inhibitor, or in poor metabolizers of CYP2D6.
Valbenazine, dosed at 40 mg/day for 1 week, then 80 mg/day, yielded a response rate (defined as >50% reduction in global symptom severity) after 6 weeks of 23.8% (40 mg/day) and 40.0% (80 mg/day), both significantly better than the reduction seen with placebo (8.7%). Its most common adverse effect is sedation (about 10%; NNH=15). Although valbenazine demonstrates a moderate effect size to reduce TD symptoms (NNT=~4), those symptoms often promptly recur if and when valbenazine is stopped.
A second reversible VMAT2 inhibitor, deutetrabenazine (a deuterated form of tetrabenazine; see Figure 17–2), received FDA approval in August 2017 for the treatment of TD. Deuterium, a stable isotope of hydrogen, extends the half lives of tetrabenazine’s active metabolites, and deutetrabenazine appears to have fewer adverse effects than tetrabenazine (including less risk for depression, akathisia, sedation, insomnia, and parkinsonism).
Two 12-week placebo-controlled randomized trials have demonstrated the safety and efficacy of deutetrabenazine for TD when dosed at 24 or 36 mg/day (Anderson et al. 2017; Fernandez et al. 2017). Deutetrabenazine carries a boxed warning (n) because of its risk for inducing suicidal thoughts or behaviors, based on clinical trials in patients with Huntington’s disease, and presumably by virtue of its potential for depletion of catecholamines and indoleamines, analogous to tetrabenazine. In FDA registration trials of deutetrabenazine for TD, the most commonly observed adverse effects were nasopharyngitis (NNH=50) and insomnia (NNH=34), yielding a likelihood of being helped or harmed of 27 (Citrome 2017). Predicted Fridericia-corrected QTc interval prolongation with deutetrabenazine in registration trials for TD (across dosages of 18 mg bid and 24 mg bid) revealed a QTc of 4.69 msec (90% CI=2.03–7.85 msec).
QTc prolongation is listed in the manufacturer’s product labeling for deutetrabenazine even though this phenomenon was not observed in its FDA registration trials for TD (likely reflecting the potential for QTc prolongation associated with tetrabenazine). Per the manufacturer, use of deutetrabenazine in patients who are poor metabolizers of CYP2D6, or when coadministered with strong CYP2D6 inhibitors, could cause clinically relevant QTc prolongation.
Among other, non-FDA-approved treatment options for TD, adjunctive vitamin E (dosed from 800 to 1,600 IU/day) was historically regarded as among the few viable strategies for either its treatment or its prevention. However, contemporary reports demonstrate less robust acute efficacy than was once believed and only modest value from taking 800 IU twice daily for delaying the onset of TD during antipsychotic therapy. The U.S. recommended daily allowance of vitamin E is 22 IU/day. A 2001 Cochrane Database review of 10 randomized studies found no evidence for reduction of TD symptoms with vitamin E but less progression of TD symptoms with vitamin E use as compared with placebo (Soares-Weiser et al. 2011). (Of note, only 3 of the 10 studies included in that review involved treatment durations of 5 months or longer.)
High-dose vitamin E in the form of alpha-tocopherol has separately been implicated as a risk factor for prostate cancer, leading some authors to advocate the use of mixed (alpha and gamma) tocopherol formulations of vitamin E along with other antioxidants (e.g., vitamin C) to protect against its possible pro-oxidative effects. Among other antioxidants, vitamin B6 (dosed at 1,200 mg/day) was found to be superior to placebo in reducing parkinsonian and dyskinetic movements as well as EPS during a 26-week randomized double-blind study involving 50 patients with schizophrenia or schizoaffective disorder and TD (Lerner et al. 2007). Extract of gingko biloba (a putative free-radical scavenger) was studied at a dosage of 240 mg/day (n=78) versus placebo (n=79) over 12 weeks in Chinese inpatients with schizophrenia and yielded a significantly greater reduction in Abnormal Involuntary Movement Scale (AIMS) scores, with response (defined as ≥30% reduction from baseline AIMS) seen in 51% of the patients who were taking active drug versus 5% of the patients who were taking placebo (Zhang et al. 2011). Similarly, the antioxidant piracetam (orally dosed at 4,800 mg/day) was superior to placebo during a 9-week randomized crossover study in 40 patients with schizophrenia or schizoaffective disorder and TD (Libov et al. 2007). Relatedly, a randomized controlled study using the ethylated congener of piracetam, levetiracetam (dosed from 500 to 3,000 mg/day; mean final dosage=2,156 mg/day), was found to be superior to placebo over 12 weeks in reducing moderately severe TD symptoms in 50 patients with schizophrenia, with a mean reduction in AIMS scores of 43.5% (as compared with 18.7% for patients given placebo) (Woods et al. 2008). Consistent with these observations, an open case series of adjunctive levetiracetam (mean dose=2,290 mg/day) found significant improvement from baseline of abnormal involuntary movements in 16 patients with TD after 1–3 months (Meco et al. 2006). It has been hypothesized that levetiracetam may improve motor function via free radical oxidative scavenging, enhancement of GABA function, enhancement of nitric oxide production, or reduction of neuronal hypersynchrony in the basal ganglia. Lastly, the hormone melatonin also possesses antioxidant properties; during one 6-week double-blind crossover study in 22 schizophrenia patients with TD, melatonin, dosed nightly (8:00 P.M.) as a single 10-mg tablet, was associated with greater reductions in abnormal involuntary motor movements and high tolerability (Shamir et al. 2001).
Other agents have been studied on the basis of the rationale of presumptive damage to striatal cholinergic neurons associated with TD. Favorable preliminary open-label data have been reported with the procholinergic agent donepezil (5–10 mg/day for 6 weeks) (Caroff et al. 2001), although findings were negative from a larger, controlled trial of the cholinesterase inhibitor galantamine (dosed at 8–24 mg/day for 12 weeks) (Caroff et al. 2007). By contrast, the M1 selective anticholinergic drug biperiden (2 mg bid) improved both parkinsonian symptoms and abnormal involuntary motor movements better than did placebo during a double-blind crossover study of 32 schizophrenia inpatients with TD (Silver et al. 1995).
Certain dopamine agonists have demonstrated value in treating TD symptoms in patients with schizophrenia. The most notable is amantadine (100 mg bid), which in a study of 32 schizophrenia inpatients with TD yielded better efficacy than placebo and comparable reductions in AIMS scores as compared with biperiden during successive 2-week single-blind randomized crossover trials (Silver et al. 1995).
GABAergic drugs such as baclofen, divalproex, progabide, and tetrahydroisoxazolopyridine also have been studied preliminarily in patients with TD. In a Cochrane Database meta-analysis of eight short-term studies, Alabed et al. (2011) described the evidence for each of these agents as “inconsistent and unconvincing,” adding that any potential benefits were typically outweighed by substantial adverse effects involving sedation or the exacerbation of psychotic symptoms.
Eicosapentaenoic acid dosed at 2 g/day was no different from placebo in a study of 84 subjects with schizophrenia or schizoaffective disorder and TD over 12 weeks (Emsley et al. 2006).
Optimistic results were reported from a pilot study of open-label high-dose buspirone (180 mg/day for 12 weeks) in a small group (N=8) of schizophrenia patients with TD (Moss et al. 1993), although these findings have not been replicated, and supratherapeutic doses of buspirone have been associated with seizures, gastrointestinal problems, sedation, paresthesias, and blurred vision. Coadministration of buspirone with other serotonergic agents also may increase the risk for serotonin syndrome.
Branched-chain amino acids received interest as a potential strategy for treating TD on the basis of observations that the large neutral amino acid phenylalanine appears to be associated with TD, whereas ingestion of branched-chain amino acids (i.e., valine, leucine, isoleucine) may correspondingly diminish brain phenylalanine availability, which in turn may reduce TD symptoms (Richardson et al. 2003). A proprietary powdered drink mix containing large branched-chain amino acids (Tarvil) became commercially available in 2002, but in 2007 it was discontinued by its manufacturer, Nutricia North America.
A number of studies have examined the potential for some atypical antipsychotics to diminish the symptoms of TD (in contrast to the masking of TD, which can be observed using increased doses of FGAs). Among these drugs, low-dose clozapine has perhaps the most robust evidence, with global improvement rates in TD symptoms reportedly in the range of 70% to 80% over periods of up to 18 weeks (Spivak et al. 1997). Some case reports suggest that the D2 partial agonist aripiprazole may improve (Karabulut et al. 2008) or cause (Abbasian and Power 2009) TD, and other reports suggest improvement in TD symptoms after quetiapine is substituted for other antipsychotics (Abbasian and Power 2009). The substitution of olanzapine for other atypical antipsychotics or FGAs in schizophrenia patients with TD has been associated with reductions in TD symptoms, with improvements sustained over an 8-month period, and no rebound worsening of TD symptoms during imposed dosage reduction periods (Kinon et al. 2004).
A long-acting form of acamprosate, SNC-102, has been in Phase II development trials as a possible novel TD treatment strategy. Other possible therapeutic treatment options under development for TD include Tardoxal (MC-1; pyridoxal-5’-phosphate), noninvasive transcranial focused ultrasound surgery guided by magnetic resonance imaging (the transcranial ExAblate system [MRgFUS]) and the Activa PC neurostimulator.
In our experience, none of the available pharmacological options to reduce TD symptoms or curtail their progression yield a dramatic or substantial benefit, with the notable exception of valbenazine. TD remains a difficult and often intransigent problem for which there is no reliable or well-proven remedy. Realistic expectations should be established with patients before embarking on experimental strategies such as those described in this section. In some cases, despite the presence of TD, the severity of psychotic symptoms may be sufficiently great to outweigh the risks of potential worsening of TD and warrant continued dopamine antagonist therapy. In such instances, the use of antipsychotic agents with relatively loose D2 receptor affinities, such as clozapine or quetiapine, may at least theoretically help to minimize the known and sometimes accepted risk for TD worsening.
To minimize the potential for withdrawal dyskinesias, the preferable action is to gradually taper off or cross-taper an existing antipsychotic rather than to abruptly discontinue it.
Abrupt cessation of antipsychotics may provoke motor adverse effects suggestive of a withdrawal dyskinesia. Some clinicians believe this phenomenon to reflect more accurately the unmasking of an underlying tardive movement disorder rather than a true process induced by withdrawal of an antipsychotic. The clinician should consider this possibility as being especially likely with antipsychotics that have short half-lives (cf. aripiprazole [t½=75 hours] or brexpiprazole [t½=91 hours]) or agents with especially tight D2 binding affinities in the basal ganglia (e.g., risperidone, haloperidol). Withdrawal dyskinesias are sometimes self-limited and may require no intervention apart from continued monitoring. If the withdrawal dyskinesia is significantly discomforting, clinicians might 1) reintroduce the discontinued antipsychotic, followed by a slower taper; 2) substitute an alternative antipsychotic with lesser known risk for disturbing extrapyramidal movements; 3) introduce a benzodiazepine; or 4) use adjunctive clonidine on a short-term basis. Withdrawal dyskinesias pose no known increased risk for future development of TD.
Clinicians should properly evaluate the characteristics and likely primary versus secondary etiology of a newly observed tremor, reduce dosages when feasible, and assure the absence of other signs of neurotoxicity. Propranolol begun at 10 mg tid or primidone dosed at 100–300 mg/day (maximum of 750 mg/day) in two or three divided doses may help to reduce or ameliorate a drug-induced tremor.
Tremor is defined as “rhythmical, involuntary oscillatory movement of a body part” (Deuschl et al. 1998, p. 3). It may be caused by numerous psychotropic agents, and its presence should signal the need for careful evaluation before changing treatments or dosages or initiating adjunctive pharmacotherapies. Tremor sometimes is a sign of neurotoxicity that should be evaluated in the context of other features that could suggest supratherapeutic dosing of a given agent (e.g., a coarse postural tremor in the setting of gastrointestinal upset and ataxia would be consistent with probable lithium toxicity). A careful history should include the assessment of caffeine intake, use of sympathomimetic agents, risk for alcohol or benzodiazepine withdrawal, history of familial or essential tremor, and other pertinent factors that may contribute to the emergence of tremor. Physiologic or benign tremors can be exacerbated by some psychotropic agents (e.g., carbamazepine, divalproex, lithium, psychostimulants, SSRIs, tricyclic antidepressants).
The evaluation of a tremor should include 1) localization of affected regions (i.e., upper and/or lower extremities; head and neck or trunk; unilateral or bilateral presence), 2) characteristics (e.g., occurring at rest or with action; qualitative dimensions such as a pill-rolling or parkinsonian tremor vs. resting tremor), 3) pupillary examination and assessment of nystagmus, 4) assessment of deep tendon reflexes, and 5) identification of the presence of cerebellar signs (e.g., ataxia, dysdiadochokinesia, Romberg sign). The differential diagnosis of tremor is vast, and its comprehensive discussion falls beyond the scope of this work.
Tremors are often characterized as occurring at rest (resting tremor) or only during voluntary muscle movement (action tremors). Action tremors are often further subclassified as
Postural: occurring with outstretched hands against gravity.
Intention: occurring when making purposeful movements, such as touching finger to nose.
Kinetic: occurring during any voluntary movement.
Isometric: occurring during a sustained voluntary muscle contraction, such as holding a heavy object.
Task-specific: occurring during repetitive specific motor tasks, such as writing.
Major distinguishing characteristics among tremors that may be useful for the clinician to consider are summarized in Table 17–5.
In the absence of other signs of frank neurotoxicity, tremor can be an otherwise benign but disruptive adverse drug effect that may be remediable either by dosage reductions or by the use of adjunctive medications. β-Blockers that exert peripheral effects such as propranolol likely are the most commonly undertaken medication strategy to counteract drug-induced tremors. Metoprolol, which is moderately lipophilic, also may treat lithium-induced tremor (Gaby et al. 1983). Hydrophilic β-blockers such as atenolol, which do not appreciably cross the blood-brain barrier, also have shown efficacy in treating lithium-induced tremor (Davé 1989).
Rapid (>8 Hz) coarse tremor at rest, mainly in hands; coarse tremors in general are more suggestive of toxicity states.
Fine (8–12 Hz) tremor, usually postural but may occur at rest, often affecting limbs as well as head, mouth, and tongue. Incidence ~6%–45% (Rinnerthaler et al. 2005).
An intention/action tremor varying from 4 to 12 Hz, most often affecting the hands but may also affect head, voice, tongue, legs, and trunk. Tremor is absent during sleep. Onset is usually in mid-adulthood, but treatment-seeking may peak in older adulthood. Course may be progressive. A positive family history can be identified in a majority of individuals.
Fine (8–12 Hz) postural tremor (i.e., evident when upper extremities are held forward outstretched against gravity); incidence ~4%–65% (Gelenberg and Jefferson 1995).
6–12 Hz postural or action tremor, with incidence up to 20%, typically arising 1–2 months after treatment is started (Morgan and Sethi 2005); may be an early sign of serotonin syndrome during SSRI therapy.
Bradykinesia; low frequency (3–6 Hz) high-amplitude resting tremor. Tremor in Parkinson’s disease is often asymmetric and diminishes with voluntary movement.
Note. SSRI=selective serotonin reuptake inhibitor; TCA=tricyclic antidepressant.
For tremors caused by lithium, divalproex, or most other psychotropics, propranolol dosing is typically begun at 10 mg bid or tid and may increase to 20–40 mg bid or tid prn, with a maximum reported efficacious dosage of 320 mg/day in a single (long-acting) or divided dosing schedule. Clinicians must monitor for bradycardia and hypotension, although the latter is less common at relatively low doses of propranolol. Long-acting or once-daily preparations of β-blockers such as propranolol (Inderal LA) may sometimes be substituted for the immediate-release formulation, provided that the higher doses in which Inderal LA is formulated pose no risk for hypotension in a given patient. As noted in earlier sections of this chapter, β-blockers are relatively contraindicated in the presence of hypotension, significant bradycardia, sick sinus syndrome, second- or third-degree atrioventricular block, and forms of chronic obstructive pulmonary disease such as asthma.
Alternatively, the pyrimidinedione anticonvulsant primidone is commonly used in off-label fashion for treatment of essential tremor, with efficacy comparable to that with propranolol (Zesiewicz et al. 2005). Dosing is typically begun at 50–100 mg at bedtime for the first few days, and the dosage then increased to 100 mg bid for an additional 2–3 days, and then titrated further upward if necessary (based on response) to 100 mg tid or as high as 250 mg tid. In studies of primidone in essential tremor, high-dose primidone (i.e., 750 mg/day) has not demonstrated superior efficacy to a daily dose of 250 mg (Serrano-Dueñas 2003). Use of primidone as an anecdotal remedy to counteract psychotropic-induced tremor is an extrapolation from data involving its use in essential tremor; formal studies of primidone for this specific purpose have not been reported. One must keep in mind that primidone is a highly potent inducer of CYP3A4 and CYP1A2 isoenzymes and can hasten the metabolism of drugs that are substrates for them (see Chapter 2, Table 2–7).
Other anticonvulsants that have shown at least preliminary value in the treatment of essential tremor include topiramate, gabapentin, levetiracetam, and oxcarbazepine.
The carbonic anhydrase inhibitors acetazolamide (begun at 125 mg/day and increased by 125 mg weekly to a maximum of 500 mg/day; usual mean dose=250–300 mg/day) and methazolamide (begun at 25 mg/day and increased to a maximal dose of about 200 mg/day) have been reported to improve essential tremor (particularly in the presence of head tremor), but the literature contains only a handful of small case reports using these agents to counteract tremor induced by lithium, anticonvulsants (e.g., divalproex), or other agents. Additionally, one may also then need to contend with adverse effects from carbonic anhydrase inhibitors (e.g., headache, paresthesias, sedation). Lastly, open data also support the possible efficacy of vitamin B6 (900–1,200 mg/day) to reduce lithium tremor, while case reports suggest that amantadine (100 mg bid) may improve divalproex-associated tremor.
Muscle Twitching, Fasciculations, and Myoclonus
Medication-induced fasciculations or muscle twitches are generally benign phenomena that require no intervention. Clinicians should assure the absence of noniatrogenic and remediable causes of fasciculations (e.g., dehydration, hypomagnesemia, hypocalcemia) that may occur as phenomena that arise coincidentally during pharmacotherapy.
Fasciculations refer to small, involuntary contractions of a skeletal muscle fascicle visible under the skin, whose movement may or may not be rhythmic. Fasciculations that occur in periorbital muscle groups are a focal dystonia described as blepharospasms. Although most instances of muscle twitching or fasciculations are benign in nature, they can also result from mineral deficiencies or other medical or neurological causes that may warrant consideration depending on clinical circumstances. Etiologies of particular relevance to psychiatry include dehydration, hypomagnesemia, hypocalcemia, hypoparathyroidism, excessive caffeine intake, benzodiazepine withdrawal, serotonin syndrome, Lyme disease, myasthenia gravis, amyotrophic lateral sclerosis, and lower motor neuron disease (e.g., denervation). Stress or anxiety may exacerbate existing fasciculations. In the case of blepharospasm, fatigue (or poor sleep), dry eyes, and local irritants should be considered during clinical evaluation. Psychotropic medications known to be associated with benign fasciculations include anticholinergic agents, diphenhydramine, amphetamine, dopamine agonists (e.g., ropinirole and similar antiparkinsonian drugs), trazodone, selegiline, phenelzine, tranylcypromine, theophylline, and depolarizing muscle blockers such as succinylcholine (as used during electroconvulsive therapy).
Myoclonic movements are sudden jerklike contractions or twitches of whole muscle groups (e.g., shoulder twitch, elbow flexion, wrist extension) that are usually benign. They may occur during sleep onset (known as hypnic jerks) or while awake, although sometimes they may reflect underlying CNS disease. Myoclonus can arise from numerous causes, including seizures (e.g., myoclonic epilepsy), infection or encephalopathy (e.g., Lyme disease, HIV, syphilis), autoimmune disorders (e.g., systemic lupus erythematosus), metabolic disorders (e.g., hyperthyroidism, hepatic or renal failure), neurodegenerative diseases (e.g., Huntington’s disease, multiple sclerosis), drug toxicities, vascular origins (e.g., poststroke), inflammatory processes, and paraneoplastic syndromes. In juvenile myoclonic epilepsy, the anticonvulsant levetiracetam is FDA approved as an adjunctive therapy.
Myoclonus rarely results from psychotropic drugs, apart from its presence as part of serotonin syndrome, or sometimes as an early sign of lithium toxicity. Medications that have been associated with myoclonus include SSRIs, TCAs, morphine, midazolam, and tramadol. Myoclonic jerks that become intrusive or persistent can be treated with clonazepam, baclofen, fluoxetine, and propranolol.
Nystagmus may indicate neurotoxicity from a medication. A careful review should be made of all medications, their dosages, and patient adherence, as well as illicit substances. Additionally, the clinician should assess the patient for other neurological or systemic signs of toxicity. Laboratory assessments (e.g., serum drug levels) to determine supratherapeutic doses may sometimes provide corroborative information. When nystagmus is thought to represent neurotoxicity, dosage reductions or the elimination of a suspected causal agent may be necessary.
Nystagmus is a focal neurological sign with a wide differential diagnosis that may include neurotoxicity from a number of psychotropic drugs. It may occur as part of alcohol or benzodiazepine withdrawal or intoxication states, as well as intoxication from phencyclidine, phenobarbital, and organic solvents. Nystagmus can variably be described as downbeat, upbeat, rotary, horizontal, pendular, and gaze evoked. Distinct types of nystagmus may reflect localized CNS lesions: upward nystagmus often reflects cerebellar or medullary disease, whereas horizontal nystagmus is the most common type of drug-induced nystagmus, involving low-amplitude beating with slow-velocity movements.
Drug-related nystagmus may occur without other clinical signs of toxicity. Lithium, for example, has been reported to cause downbeat nystagmus even at dosages producing nontoxic blood levels, and dosage reductions do not necessarily ameliorate or diminish the phenomenon. Other medications known to cause nystagmus include carbamazepine, divalproex, lamotrigine, MAOIs, and propranolol. Gabapentin has been used successfully to treat congenital nystagmus and is not known to cause drug-related nystagmus. Nystagmus may occur as part of serotonin syndrome (but not NMS), although it has not otherwise been described in association with the use of serotonergic antidepressants. It may be a rare adverse effect of SNRIs. Nystagmus has not been reported in conjunction with the use of antipsychotics, apart from one report of coarse horizontal nystagmus after an ultimately fatal overdose of olanzapine.
The presence of nystagmus signals the need for a review of all existing medications and dosages, with an awareness of pertinent pharmacokinetic interactions (e.g., lamotrigine toxicity from unadjusted dosing when coadministered with divalproex). A basic neurological examination should target cranial nerve abnormalities and neurotoxic signs (e.g., slurred speech, tremor, ataxia), and laboratory measures should include a toxicology screen and assessment of pertinent drug levels (see Table 1–2 in Chapter 1, “The Psychiatrist as Physician”). Suspected causal drugs should be discontinued if dosage reductions alone do not resolve symptoms or if other systemic manifestations of toxicity are evident.
Paresthesias and Neuropathies
Paresthesias attributable to psychotropic medications are usually benign phenomena that may result from drug properties such as carbonic anhydrase inhibition or other direct effects on sensory nerve endings. They generally pose no medical concern and require no intervention, other than the ruling out of other possible etiologies unrelated to a suspected pharmacotherapy. Neuropathies rarely result from psychotropic drugs, benignly arising most often from lithium or phenelzine.
Paresthesias commonly occur during treatment with carbonic anhydrase inhibitors (e.g., topiramate was associated with paresthesias in 35%–49%of migraine patients at doses of 50 mg/day and 200 mg/day, respectively) and are identified in package insert labels as a rare event reported with a wide range of psychotropic drugs, including divalproex, most SGAs and SSRIs, venlafaxine, desvenlafaxine, duloxetine, mirtazapine, stimulants, and some nonbenzodiazepine sedative-hypnotics. Other nonpsychotropic medications that are known to cause paresthesias include certain antibiotics (e.g., doxycycline), acetazolamide, antineoplastic agents, and a number of antiretroviral agents, among other compounds. In addition, potential medical causes of paresthesias include hyperventilation (e.g., secondary to panic attacks), local trauma or nerve entrapment (e.g., carpal tunnel syndrome, disk herniation), diabetic neuropathy, migraine, peripheral vascular disease, vitamin B12 deficiency, toxic exposures, malignancy, infections, or connective tissue diseases. Withdrawal from SSRIs or SNRIs may be associated with paresthesias, remediable by undertaking more protracted drug tapers or by allowing for the passage of time alone, inasmuch as paresthesias in this context are medically benign. Iatrogenic paresthesias that occur in the absence of other focal neurological signs require no intervention.
Psychotropic drugs are rarely associated with peripheral neuropathies. Anticonvulsants other than phenytoin are not known to cause neuropathy (and in fact are often used in the treatment of diabetic neuropathy or neuropathic pain). Neuropathy has been reported in connection with lithium toxicity. The literature contains several dozen case reports of peripheral or optic neuropathy associated with disulfiram, typically at doses >250 mg/day, usually (but not always) resolving up to several months (but as long as 14 months) after drug discontinuation. Antipsychotics are not associated with neuropathy, although some case reports identify its development in the course of NMS. Antidepressants are not associated with the development of peripheral neuropathy, with the exception of the MAOI phenelzine—a compound that falls within a chemical class known as hydrazines, which have been shown in preclinical and human studies to reduce levels of vitamin B6 (pyridoxine)—which, in turn, can lead to peripheral neuropathy that may be remediable within several weeks’ time by the administration of supplemental pyridoxine dosed 150–300 mg/day (Stewart et al. 1984). Measurement of serum pyridoxine levels is not necessary to justify an empirical trial of supplemental therapy in this dosing range, which is a relatively benign intervention.
The seizure threshold may be lowered by antidopaminergic drugs and bupropion, often in dose-related fashion. Coadministration of anticonvulsants may be advisable in patients for whom high doses of antipsychotics (particularly clozapine) or bupropion are deemed necessary.
Among antidepressants, seizure risk during FDA registration trials with most SSRIs has been reported as approximately 0.2%, with higher rates associated with TCAs. Risk for seizures with antidepressants in general also appears related to high doses or frank toxicity states. Manufacturer’s product information for bupropion identifies a dose-dependent risk for seizures in about 4 in 1,000 patients when bupropion is prescribed up to 450 mg/day. Among anticonvulsants, long-term use raises the seizure threshold, which can, at least in theory, pose an increased risk for seizures following abrupt drug discontinuation. There-fore, anticonvulsant cessation should ideally occur over about a 2-week period. However, rapid anticonvulsant cessation mainly poses a concern for seizure induction in epilepsy patients, and for practical purposes (e.g., prior to ECT initiation), lengthy taper-offs may not always be necessary.
Clozapine-induced seizures (■) are more likely to occur at high clozapine dosages (incidence rate of 5% above 600 mg/day per the manufacturer), during rapid dose increases, during concomitant use of other medications that lower the seizure threshold or inhibit its metabolism, or in patients with a neurological deficit (Toth and Frankenburg 1994). Clinicians sometimes coadminister an anticonvulsant with clozapine when the latter is dosed >500–600 mg/day to help minimize the potential for seizures. (Co-administration of clozapine with lamotrigine represents an especially attractive strategy in that it could simultaneously provide seizure prophylaxis and possible antipsychotic synergy, as has been suggested from several early reports.) All antipsychotics have the potential to lower an individual’s seizure threshold, although risks appear highest with clozapine, loxapine, and chlorpromazine, and appear lowest with haloperidol, pimozide, thiothixene, and most SGAs other than clozapine (<1%–2% incidence). Risk factors include a past history of seizures, concomitant medications that also lower the seizure threshold, and rapid dose escalations. Decisions about whether to stop, reduce, or continue an otherwise efficacious drug after seizure occurrence depend on the presence of an underlying seizure diathesis, the risk-benefit analyses for continuing the drug in question, and viable methods to manage the risk for possible subsequent seizures.
Yawning is a generally medically benign, uncommon side effect of some antidepressants. It requires no intervention. It is not clearly dose related. If yawning is sufficiently distressing to the patient, the presumed causal agent might be discontinued and substituted with another medication with a presumptive different mechanism of action (e.g., switching from an SSRI or SNRI to bupropion or a TCA).
Excessive daytime yawning has been described in postmarketing reports with the SSRIs fluoxetine, sertraline, and citalopram, among others, as well as the SNRIs duloxetine and venlafaxine. Yawning has been hypothesized to occur as a reflex that is modulated by catecholaminergic, serotonergic, and other transmitter systems (including cholinergic, glucocorticoids, nitric oxide) in the paraventricular nucleus of the hypothalamus (De Las Cuevas and Sanz 2007) and that may become altered during antidepressant therapy. Usually, yawning is a benign phenomenon, although reports exist that its frequency may occasionally pose a disabling effect (e.g., yawning-associated orgasms during clomipramine therapy [McLean et al. 1983]). Iatrogenic yawning sometimes may be diminished by adjunctive cyproheptadine.
The symbol ■ is used in this chapter to indicate that the FDA has issued a boxed warning for a prescription medication that may cause serious adverse effects.