Journal of the Neurological Sciences 389 (2018) 61–66

Contents lists available at ScienceDirect

Journal of the Neurological Sciences

journal homepage: www.elsevier.com/locate/jns

Review Article

Reprint of: Clinical management of tardive dyskinesia: Five steps to T

success☆,☆☆
⁎
Leslie Citrome
New York Medical College, Valhalla, NY, USA

A R T I C L E I N F O A B S T R A C T

Keywords: Tardive dyskinesia (TD) has long been thought to be a generally irreversible consequence of the use of dopamine
Abnormal Involuntary Movement Scale receptor blocking agents. There is now an opportunity to successfully manage this condition with agents ap-
Antipsychotic proved by the US Food and Drug Administration. This is important because TD has not been eliminated with the
Deutetrabenazine use of second-generation antipsychotics, and the expansion of antipsychotics to treat conditions other than
Dopamine
schizophrenia has resulted in millions of additional individuals at risk for developing TD. Recognition of TD
Tardive dyskinesia
requires careful observation; a structured approach using the Abnormal Involuntary Movement Scale is en-
Valbenazine
VMAT2 couraged. Harm reduction can be achieved by using antipsychotics judiciously when possible and by paying
attention to other modifiable risk factors such as drug-induced parkinsonian symptoms and the use of antic-
holinergic medication. Once TD has emerged and is associated with dysfunction or distress, treatment with a
VMAT2 inhibitor such as deutetrabenazine or valbenazine is well supported by several controlled clinical trials.

1. Introduction by involuntary, repetitive, purposeless movements, typically of the

Tardive dyskinesia (TD) has long been thought to be a generally
irreversible consequence of the use of dopamine receptor blocking
agents. Dopamine receptor blocking agents not only include anti-
psychotic medications used for the treatment of psychiatric disorders,
but also antitussive agents such as promethazine and antiemetic med-
ications such as metoclopramide used for symptomatic gastro-
esophageal reflux in patients who fail to respond to conventional
therapy and for diabetic gastroparesis. Warnings regarding TD are in-
cluded in US product labeling for all antipsychotics [1]. Within the
product label for metoclopramide, the warning for TD is boxed and
bolded and reads “…Metoclopramide therapy should be discontinued in
patients who develop signs or symptoms of tardive dyskinesia. Treat-
ment with metoclopramide for longer than 12 weeks should be avoided
in all but rare cases where therapeutic benefit is thought to outweigh
the risk of developing tardive dyskinesia” [2]. In the treatment of
psychotic disorders such as schizophrenia, treatment with anti-
psychotics is often indefinite, and advice to discontinue this interven-
tion is usually not clinically feasible.
Although TD is sometimes quite challenging to tease apart from
other movement disorders that may be comorbid, TD is distinguished
tongue, jaw, lips, face, trunk, upper extremities, and lower extremities,
and these movements can be associated with significant functional
impairment and can be socially stigmatizing [3]. TD movements can
vary during the day and disappear during sleep.
Unfortunately, TD is not rare. Fortunately, there are therapeutic
options that have recently become available for the effective manage-
ment of TD. This review will provide an overview of the management of
TD in five steps: 1) Recognition; 2) Assessment; 3) Harm reduction; 4)
Interventions; and 5) Follow-up.
2. Step 1: recognition - admit the possibility that some of your
patients have tardive dyskinesia
If you use antipsychotics in your clinical practice, chances are you
have patients with TD. The availability of second-generation anti-
psychotics has not eliminated TD. In a meta-analysis of 41 studies
published since 2000, the overall mean TD prevalence among 11,493
patients exposed to antipsychotics was 25.3%. Although a lower pre-
valence for those receiving second-generation antipsychotics vs. first-
generation antipsychotics was noted (prevalence rates of 20.7% vs.
30.0%, respectively), and the lowest prevalence (7.2%) was observed

DOI of original article: http://dx.doi.org/10.1016/j.jns.2017.11.019

☆
This article is part of the Special Issue Tardive Syndromes and Their Management edited by Robert A. Hauser and Daniel D. Truong.
☆☆
A publisher's error resulted in this article appearing in the wrong issue. The article is reprinted here for the reader's convenience and for the continuity of the special issue. For
citation purposes, please use; Journal of the Neurological Sciences, 383C, pp. 199-204.
⁎
11 Medical Park Drive, Suite 106, Pomona, NY 10970, USA.
E-mail address: citrome@cnsconsultant.com.

https://doi.org/10.1016/j.jns.2018.02.037
Received 13 November 2017; Accepted 15 November 2017
Available online 28 February 2018
0022-510X/ © 2017 Elsevier B.V. All rights reserved.
L. Citrome
among patients receiving second-generation antipsychotics who had
never been exposed to first-generation antipsychotics [4], the pre-
valence rates are high enough that TD is a relatively common entity.
Illustrating the above is a well-executed cohort study conducted at
the Connecticut Mental Health Center that compared the incidence of
TD for second- vs. first-generation antipsychotics using methods like
those from a previous prospective cohort study at that site in the 1980s
[5]. Among the 352 initially TD-free outpatients who were examined
for a new diagnosis of TD every 6 months for up to 4 years, the adjusted
TD incidence rate-ratio for subjects treated with second-generation
antipsychotics alone vs. first-generation antipsychotics alone was 0.68,
suggesting a small advantage for the newer agents. However, the 95%
confidence interval of 0.29 to 1.64 includes the possibility that there
really is no difference. Importantly, the incidence rates did not differ
substantially from previous findings at this site in the 1980s when TD
was a prominent concern. TD rating scale scores were only slightly
lower among incident cases of TD appearing after recent second-gen-
eration antipsychotic exposure vs. recent first-generation antipsychotic
exposure. Of note, staff involved in this study were well trained to
identify TD, and systematically looked for it. What would the yield be in
your practice if you were to screen your own patients for TD on a
regular basis?
It is likely that with the expansion of the use of antipsychotics be-
yond psychotic disorders the number of people at risk to develop TD
will continue to increase. Many persons with major depressive disorder
receive adjunctive antipsychotic medications in an effort to better
control their depressive symptoms. This increased use of dopamine
receptor blocking agents in routine practice makes admitting the pos-
sibility that your patients may have TD even more important.
3. Step 2: assessment - understanding how to effectively use the
Abnormal Involuntary Movement Scale and making a differential
diagnosis
Clinically assessing patients for TD requires careful observation.
Although screening for TD can be done without the use of a specific
scale, quantifying the severity of the TD will necessitate the use of an
agreed-upon measurement. The 12-item Abnormal Involuntary
Movement Scale (AIMS), developed by the US National Institute of
Mental Health, is a convenient way to structure a brief evaluation of a
patient for TD ([6,7]; see also Supplementary Appendix). Although
other scales exist, the AIMS is arguably the best known, and has also
served as the primary outcome measure for recent studies of agents
recently approved by the US Food and Drug Administration (FDA) for
the treatment of TD [8–11]. The AIMS alone is not sufficient to diagnose
TD, but can serve as a comprehensive screen and can be used to follow a
person's progress over time.
When will a complete AIMS examination be required? If the patient
or caregiver initially complains of abnormal movements or if the clin-
ician observes new abnormal movements, a full evaluation is desirable.
Since TD can vary over time and can worsen with emotional stress,
repeat examinations can be helpful. The AIMS examination begins by
observing the patient at rest, which initially can be done unobtrusively
while the patient is in the waiting area. Once in the office or ex-
amination room it is ideal to have two identical hard, firm chairs with
no arms placed facing each other – one for the patient and one for the
examiner. In this manner the clinician can easily model what is being
asked of the patient: sitting with hands on knees, hands hanging un-
supported between the legs or over the knees, opening the mouth,
protruding the tongue, tapping the thumb with each finger, and then
standing up, extending both arms in front with palms down, and lastly,
walking. The thumb-finger tapping, arm extension and walking are
“activation” maneuvers used to elicit abnormal movements in other
body areas. An additional activation maneuver that can be used is a
cognitive task such as asking the patient to count backwards from 100
or to recite the months of the year in reverse order.
62
Journal of the Neurological Sciences 389 (2018) 61–66
During the entire AIMS examination the clinician remains vigilant
for any dyskinetic movements, which then get rated using the first 7
items of the AIMS, representing 7 muscle groups or body areas, using a
5-point scale (0 = none, 1 = minimal/extreme normal, 2 = mild,
3 = moderate, 4 = severe). For each item the rating that is given
should reflect the most severe movement that was witnessed, regardless
of any activation maneuvers. The total sum of these 7 items constitutes
the AIMS dyskinesia score. However, clinicians should exercise caution
in interpreting this total score in the absence of knowing the individual
item scores. For example, when treating individual patients, an AIMS
dyskinesia score of 7 has very different implications when it is based on
a score of 1 (i.e., minimal/extreme normal) across all 7 items vs. a score
of 3 (moderate) on 2 items and a 1 on a third item. In the latter patient,
more urgent efforts will need to be made to get those individual item
scores of moderate reduced to as low as possible.
To avoid confusing oral movements due to the presence of gum or
candy, these will need to be removed from the mouth. The patient will
also need to be asked about any dental problems – these may also lead
to oral movements that can be confusing to differentiate from TD. After
completing the first 7 items of the AIMS, the next 3 items document the
overall severity of the abnormal movements (0–4; based on the highest
single score in the first 7 items), the degree of incapacitation due to
abnormal movements (0–4; the patient will need to be asked to what
extent any movements interfere with activities such as eating, drinking,
speaking, breathing, dressing oneself, writing, working, leisure activ-
ities, being with others, etc.), the patient's awareness (and distress
level) of the abnormal movements (0–4, with 0 noting no awareness, 1
noting being aware with no distress, and 2–4 noting awareness and
distress rating from mild to moderate to severe). It is not unusual for
persons with schizophrenia to have little insight into their dyskinetic
movements; however, patients with mood disorders may be better able
to articulate their distress. The last 2 items of the AIMS are yes/no
questions regarding dentition status and the use of dentures.
The instructions for the AIMS also include an assessment of upper
extremity rigidity by flexing and extending the patient's left and right
arms, as well as observation of gait, but these are not rated.
Nevertheless, findings from these actions may be helpful when de-
termining if the patient has drug-induced parkinsonian side effects.
An AIMS score of at least 2 (i.e., mild) in 2 or more body regions or a
score of 3 (moderate) or 4 (severe) in at least one body region in a
patient with at least 3 months of cumulative antipsychotic drug ex-
posure equates to a probable diagnosis of TD according to the Schooler-
Kane criteria [12], provided that other causes of the abnormal move-
ments have been ruled out. An alternate (but similar) definition of TD is
the Glazer-Morgenstern criteria [13], where the detection of TD cases is
based on two applications of the AIMS at the beginning and end of each
visit; the criteria for diagnosing new persistent cases of TD at follow-up
visits are a total AIMS dyskinesia score of 3 or more on both applica-
tions at two successive visits and at least 1 body area score of 2 or more
on both applications at the same two visits.
The definition of TD contained in the American Psychiatric
Association Diagnostic and Statistical Manual of Mental Disorders, Fifth
Edition, notes that symptoms may develop after a shorter period of
medication use in older persons [14]. Moreover, if the abnormal
movements appear after discontinuation of an antipsychotic, or after a
change or reduction in the dosage of the antipsychotic medication, the
condition is called withdrawal-emergent dyskinesia. However, if the
withdrawal-emergent dyskinesia persists beyond 8 weeks, the patient
would be considered as having TD.
In the American Psychiatric Association's practice guideline for the
treatment of patients with schizophrenia [15], patients taking an anti-
psychotic medication should be regularly monitored for movement
disorders. The recommended frequency for the assessment of TD is
every 6 months for persons taking a first-generation antipsychotic or
every 12 months for those receiving second-generation antipsychotics.
However, if patients have additional risk factors for TD, then the
L. Citrome
recommended monitoring is more frequent: every 3 months for first-
generation antipsychotics and every 6 months for second-generation
antipsychotics. Although the AIMS is mentioned as a possible means of
assessing TD, the complete AIMS examination was not explicitly noted
as being the only means of clinically assessing TD. It may be possible to
conduct an “abbreviated” AIMS consisting of examination of the face
and arms, and with activation maneuvers; this abbreviated version of
the AIMS is intended for screening purposes and has not been system-
atically compared with the utility of doing the complete examination.
The most common obstacle in making a correct diagnosis of TD is
confusion with drug-induced parkinsonism. Parkinsonian tremor is
rhythmic and faster (3–6 Hz) compared to the slow arrhythmic move-
ments typical of TD [7]. Drug-induced parkinsonism typically occurs
soon after an antipsychotic is initiated or when the dose is raised, and is
often lessened when the antipsychotic dose is reduced [16]. TD occurs
later (“tardive” means delayed) and is often made worse (at least
temporarily) when the antipsychotic dose is reduced, and the dyskinetic
movements appear improved (“masked”) when the antipsychotic dose
is increased; this pattern of apparent worsening or improvement of TD
with a decrease or increase of antipsychotic dose, respectively, is the
reverse of what is expected with drug-induced parkinsonism. Moreover,
anticholinergic medications, such as benztropine, will treat drug-in-
duced parkinsonian tremor but may make TD worse [17].
Other tardive syndromes may be present, for example tardive dys-
tonia and tardive akathisia [18]. Conditions that may resemble TD in-
clude spontaneous dyskinesias not associated with antipsychotic ex-
posure, as well as conditions such as Huntington disease and Wilson
disease that are generally accompanied by a variety of other signs and
symptoms. Some drugs or substances can also be associated with ab-
normal movements; these include stimulants or excessive caffeine use.
4. Step 3: harm reduction – reducing risk
Ideally it is best to prevent TD. There are many different risk factors
for TD. Older age and antipsychotic medication exposure (dose/dura-
tion) are well established risk factors and other risk factors include
female sex, African American ethnicity, pre-existing mood disorder,
cognitive disturbance, alcohol or substance abuse, treatment with first-
generation antipsychotics, use of lithium or antiparkinsonian agents,
early occurrence of drug-induced parkinsonian symptoms, diabetes,
and HIV positivity [19].
Avoiding the use of dopamine receptor blocking agents is not fea-
sible when managing patients with schizophrenia; for schizophrenia,
antipsychotic medication is currently the only foundational treatment
available. Antipsychotic medication may also be difficult to avoid in
persons with bipolar disorder who cannot be stabilized or maintained
on mood stabilizers alone; again, attention to antipsychotic dose is re-
quired. For adjunctive use in major depressive disorder, antipsychotic
medications have been studied principally as an acute treatment and
thus there are few data available regarding long-term exposure in the
maintenance phase of treatment. Other uses of antipsychotic medica-
tions such as for sleep, post-traumatic stress disorder, or anxiety, are
currently “off-label,” and thus risk for TD makes this choice of treat-
ment for these conditions problematic, especially in the face of alter-
native interventions.
Minimizing drug-induced parkinsonian symptoms by selecting
agents with lower risk for this problem may be helpful in reducing the
risk of eventually developing TD. This is a rationale for using second-
generation antipsychotics, and for switching antipsychotic medications
should drug-induced parkinsonian symptoms occur. The use of antic-
holinergics such as benztropine to address drug-induced parkinsonian
symptoms can place the patient at additional risk for developing TD,
and can worsen existing TD [17]. Moreover, anticholinergic medication
has a deleterious effect on cognition and can further disadvantage a
person with schizophrenia who already is struggling with cognitive
impairment [20].
63
Journal of the Neurological Sciences 389 (2018) 61–66
Because age is a strong risk factor for TD, if antipsychotics are to be
used in the elderly, low but effective doses of single drugs should be
employed together with regular and specific assessments for early TD.
5. Step 4: interventions
Up until recently there were few options to manage TD, none were
approved by regulatory authorities, and this lack of effective inter-
ventions led to substantial therapeutic nihilism [3,18,21]. A common
management strategy for TD would be to gradually switch a patient
from a first-generation antipsychotic to a second-generation anti-
psychotic (other than clozapine) and discontinue any anticholinergic
medications. If the TD did not improve, a switch to another second-
generation antipsychotic (other than clozapine) was attempted, and if
that failed, a switch to clozapine was recommended. A last choice
would be to consider so-called “suppression therapy” where the TD
would be masked by adding a potent first-generation antipsychotic
medication; this was generally only contemplated for severe TD that
substantially impaired function or was potentially life-threatening (as
when TD can interfere with breathing because of involvement of the
diaphragm muscle). A nutritional intervention was also commercia-
lized: a combination of the branched-chain amino acids L-leucine, L-
valine, and L-isoleucine. This intervention was approved by the FDA as
a medical food for the dietary management of TD in male patients based
on a double-blind study where one-third of the treatment group had a
reduction of 60% or more in dyskinetic movements, compared with
none of the subjects receiving placebo [22]; the branded product is no
longer available but compounding pharmacies can make it with the
same ratio of ingredients tested in the clinical trial.
In 2013, guidelines from the American Academy of Neurology [23]
opined that data were insufficient as to whether stopping the medica-
tion associated with TD would suppress symptoms and the strategy of
switching from a first- to second-generation antipsychotic was also
determined to have insufficient evidence. Other strategies that have
come and gone because of insufficient data to support (or refute) their
use include acetazolamide, bromocriptine, thiamine, baclofen, vitamin
E, vitamin B6, selegiline, melatonin, nifedipine, levetiracetam, bus-
pirone, yi-gan san, biperiden discontinuation, botulinum toxin type A,
electroconvulsive therapy, α-methyldopa, reserpine, and pallidal deep
brain stimulation. The guidelines explicitly did not recommend dil-
tiazem, galantamine, or eicosapentaenoic acid.
According to the American Academy of Neurology guidelines, 4
interventions could be recommended, based on evidence of varying
quality: clonazepam, Ginkgo biloba, amantadine, and tetrabenazine.
Clonazepam is a benzodiazepine with indirect GABA agonist activity
and has been explored as a treatment for TD since the 1970s [24].
Unfortunately, although in one study clonazepam reduced dyskinesia
scores by approximately 37% (using the Maryland Psychiatric Research
Center Movement Disorder Scale), tolerance developed in patients who
received clonazepam in the long-term extension after 5 to 8 months of
use [25]. Gingko biloba, an antioxidant, appeared effective in reducing
TD as measured by the AIMS in a meta-analysis of three 12-week,
randomized trials from China [26]. Amantadine, a blocker of N-methyl-
D-aspartate glutamate receptors, has a modest effect in reducing TD
(reduction of AIMS dyskinesia score by 15–22%) [27,28]; an extended-
release formulation of amantadine recently received approval for the
treatment of dyskinesia in patients with Parkinson's disease receiving
levodopa-based therapy, with or without concomitant dopaminergic
medications [29]. Tetrabenazine was originally developed in the 1950s
and was approved in the US in 2008 as an orphan drug for the treat-
ment of choreiform movements associated with Huntington disease
[30]. Tetrabenazine is a reversible and specific inhibitor of vesicular
monoamine transporter 2 (VMAT2), a protein in the brain that trans-
ports monoamine neurotransmitters such as dopamine, norepinephrine,
serotonin and histamine into vesicles for release into the synapse when
the neuron fires [31,32]. With less dopamine in the vesicle, less
L. Citrome Journal of the Neurological Sciences 389 (2018) 61–66

Table 1
Deutetrabenazine and valbenazine for tardive dyskinesia, information from product labeling [8,9] and summary data from systematic reviews [10,11].

Deutetrabenazine Valbenazine

Brand name Austedo Ingrezza

Date approved by US FDA for TD
Dose/formulation
Other indications
Design rationale
Metabolites
Half-life
Boxed bolded warnings relevant to TD
Contraindications relevant to TD
Warnings and precaution contained in
Highlights of Prescribing Information
Dosing recommendations
CYP2D6 poor metabolizers
Drug-drug interactions
Hepatic impairment
Renal impairment
QT prolongation recommendations
Most common AEsa and rate vs. placebo
Responderb rates, pooled
NNT (95% CI) vs. placebo
CGI responderc rates, pooled
NNT (95% CI) vs. placebo
Discontinuation rates due to an AE,
pooled
NNH (95% CI) vs. placebo
August 2017 April 2017
Tablets: 6 mg, 9 mg, 12 mg Capsules: 40 mg, 80 mg
Chorea associated with Huntington's disease None
Deuteration results in slower drug metabolism Parent drug of [+]-α-HTBZ; no β-HTBZ
Active deuterated dihydro metabolites (HTBZ): α-HTBZ and β- Active metabolite: ([+]-α-HTBZ)
HTBZ
Total (α + β)-HTBZ from deutetrabenazine: 9–10 h Valbenazine and [+]-α-HTBZ: 15–22 h
None None
Hepatic impairment; taking reserpine, MAOIs, tetrabenazine, None
or valbenazine
QT prolongation; neuroleptic malignant syndrome; akathisia, Somnolence; QT prolongation
agitation, restlessness, and parkinsonism (latter not applicable
to TD); sedation/somnolence
Initial dose 12 mg/d, target dose 12–48 mg/d, administer BID Initial dose 40 mg/d, target dose 80 mg, administer once daily with
with food; titrate at weekly intervals by 6 mg/d based on or without food; titrate to 80 mg/d after one week on 40 mg/d
reduction of TD and tolerability
Maximum recommended dosage in poor CYP2D6 metabolizers Consider dose reduction based on tolerability
is 36 mg/d
Strong CYP2D6 inhibitors: maximum recommended dose is MAOIs: avoid; strong CYP3A4 inducers: not recommended; strong
36 mg/d; alcohol or other sedating drugs may have additive CYP3A4 inhibitors: reduce dose to 40 mg; strong CYP2D6
sedation and somnolence inhibitors: consider dose reduction based on tolerability
Contraindicated Recommended dose for patients with moderate or severe hepatic
impairment is 40 mg/d
No clinical studies have been conducted to assess the effect of No dosage adjustment is necessary for patients with mild to
renal impairment moderate renal impairment; use is not recommended in patients
with severe renal impairment
For patients at risk for QT prolongation, assess the QT interval For patients at increased risk of a prolonged QT interval, assess the
before and after increasing the total dosage above 24 mg/d QT interval before increasing the dosage
Nasopharyngitis (4% vs. 2%), insomnia (4% vs. 1%) Somnolence (10.9% vs. 4.2%)
30.0% vs. 14.8% 36.5% vs. 12.4%
7 (4–18) 5 (3–7)
46.9% vs. 33.0% 40.4% vs. 18.6%
8 (4–45) 5 (4–9)
3.6% vs. 3.1% 2.6% vs. 1.6%
189 (not significant) 76 (not significant)

AE – adverse event; CI – confidence interval; MAOI – monoamine oxidase inhibitor; NNH – number needed to harm; NNT – number needed to treat; TD – tardive dyskinesia.
a
As identified in product labeling in Highlights of Prescribing Information; for deutetrabenazine definition is 4% and greater than placebo; for valbenazine definition is ≥ 5% and
twice the rate of placebo.
b
Response defined as a ≥ 50% decrease in the AIMS dyskinesia score (sum of items 1–7) from baseline to endpoint (12 weeks in the deutetrabenazine studies and 6 weeks in the
valbenazine studies); for deutetrabenazine the fixed dose study arm of 12 mg/d was excluded was excluded from the calculations.
c
CGI response defined as a score of 1 (very much improved) or 2 (much improved) at endpoint (12 weeks in the deutetrabenazine studies and 6 weeks in the valbenazine studies); for
deutetrabenazine the fixed-dose study arm of 12 mg/d was excluded from the calculations.

dopamine is released into the synapse, and there is thus a lower like-
lihood of triggering an abnormal movement in a system made “hy-
persensitive” due to chronic post-synaptic dopamine receptor blockade
and subsequent upregulation of both number and function of post-sy-
naptic dopamine receptors. There are a limited number of studies of
tetrabenazine for TD, the first one published in 1972 [33], but overall
these studies have demonstrated antidyskinetic effects [23,33–35]. The
short half-life of tetrabenazine, sensitivity to CYP 2D6 metabolism
(genotyping is required for tetrabenazine doses above 50 mg/d), and its
relatively poor tolerability profile have led to drug development efforts
to improve upon tetrabenazine. Two different approaches have been
used and have resulted in two new VMAT2 inhibitors that are FDA-
approved for the treatment of TD in adults: deutetrabenazine [9,11]
and valbenazine [8,10]. Table 1 provides an overview of both products.
Both deutetrabenazine and valbenazine were approved on the basis of
placebo-controlled randomized clinical trials whose primary outcome
was change on the AIMS dyskinesia score (total score on items 1–7)
from baseline to end point (Week 6 for valbenazine studies and Week
12 for deutetrabenazine studies), as assessed by blinded video raters
who were movement disorder specialists otherwise not involved in the
study.
The first medication to be approved was valbenazine in April 2017
[8]. Valbenazine is itself a selective VMAT2 inhibitor (with relatively
low affinity) and is metabolized slowly to [+]-α-dihydrotetrabenazine
([+]-α-HTBZ), a potent and selective VMAT2 antagonist. [+]-α-HTBZ
is one of the metabolites of tetrabenazine and thus valbenazine and
tetrabenazine have this metabolite in common. The half-life of valbe-
nazine and [+]-α-HTBZ approaches 1 day and thus daily dosing is re-
commended (with or without food). Genotyping is not required but
dosing adjustments may be needed in the presence of CYP 2D6 or CYP
3A4 modulators. Efficacy is optimal at the 80 mg/d dose and is reached
after one week on 40 mg/d. In a placebo-controlled, double-blind,
flexible dose (25–75 mg/d) study of 6 weeks in duration, the least-
squares mean change in AIMS dyskinesia scores from baseline to end-
point was significantly greater with valbenazine (− 2.6) than placebo
(− 0.2; P = 0.0005) and no patients discontinued because of an ad-
verse event [36]. After titration, 76% of those taking valbenazine
reached the 75 mg/d dose. In a 6-week, double-blind, fixed-dose study
of valbenazine (40 and 80 mg/d) vs. placebo, the least-squares mean
change from baseline to endpoint was significantly greater for valbe-
nazine 80 mg/d (− 3.2) vs. placebo (− 0.1; P ≤ 0.001), as was also
observed for the 40 mg/d dose (− 1.9; P = 0.0021) [37]. Dis-
continuation because of an adverse event occurred in 3.8% of those
taking valbenazine (both doses pooled) versus 2.6% of those taking
placebo.
The second medication to be approved for TD was deutetrabenazine

64
L. Citrome
in August 2017 [9]. Deutetrabenazine is related to tetrabenazine in that
deuterium is substituted for hydrogen at the key sites of primary me-
tabolism, altering the pharmacokinetics of drug metabolism by slowing
it down. Thus, the active metabolites of deutetrabenazine, deuterated
α-HTBZ and β-HTBZ, have longer half-lives compared to the analogous
non-deuterated active metabolites of tetrabenazine. Deuterium is a
stable, nonradioactive, nontoxic, naturally occurring isotope of hy-
drogen and has the same size and shape as a hydrogen atom, differing
only in forming stronger chemical bonds. Genotyping is not required
but dosing adjustments may be needed in the presence of CYP 2D6
modulators. Twice daily dosing with food is recommended. In contrast
to the fixed target dose for valbenazine (80 mg/d), deutetrabenazine
dose is determined individually for each patient based on reduction of
TD and tolerability. The recommended starting dose of deute-
trabenazine for TD is 6 mg twice daily, and can be increased at weekly
intervals in increments of 6 mg/d to a maximum recommended daily
dosage of 48 mg. In a placebo-controlled, double-blind, flexible dose
(12–48 mg/d) study of 12 weeks in duration, the least-squares mean
change in AIMS dyskinesia scores from baseline to endpoint was sig-
nificantly greater with deutetrabenazine (−3.0) than placebo (− 1.6;
P = 0.019) and more patients discontinued because of an adverse event
with placebo (3.4%) than with deutetrabenazine (1.7%) [38]. After
titration, the mean total daily dose was 38.8 mg/d. In a 12-week,
double-blind, fixed-dose study of deutetrabenazine (12, 24 and 36 mg/
d) vs. placebo, the least-squares mean change from baseline to endpoint
was significantly greater for deutetrabenazine 36 mg/d (− 3.3) vs.
placebo (−1.4; P = 0.001), as was also observed for the 24 mg/d dose
(− 3.2; P = 0.003), but not for the 12 mg/d dose (− 2.1; P = 0.217)
[39]. The percentages of patients discontinuing because of an adverse
event were 4% for deutetrabenazine (all doses pooled) versus 3% for
placebo.
Clinically relevant reductions in dyskinetic movements can be ex-
pected in persons receiving either deutetrabenazine or valbenazine.
Note that participants who entered the studies on antipsychotic medi-
cations continued to receive them at the same dose. Using the outcome
of a 50% or greater reduction in baseline AIMS dyskinesia score as a
desired response, number needed to treat (NNT) can be calculated to
compare the active drug vs. placebo. NNT describes the number of
patients that needed to be randomized to active drug rather than pla-
cebo before expecting to encounter one additional responder [40]. NNT
values < 10 generally denote interventions that are found to be
clinically useful [41]. Fig. 1 illustrates the NNT for AIMS response from
the two pivotal fixed-dosed studies of deutetrabenazine [39] and val-
benazine [37]. The NNT values for deutetrabenazine 24 and 36 mg/d
(pooled) and valbenazine 40 and 80 mg/d (pooled) are the same: 5. The
NNT for AIMS response vs. placebo when pooling both of the available
deutetrabenazine studies (excluding the dose of 12 mg/d) was 7, and
Fig. 1. ≥ 50% reduction in AIMS dyskinesia score from baseline to endpoint, NNT vs.
placebo and 95% CIs, for the Phase III fixed-dose studies of deutetrabenazine (12 weeks)
and valbenazine (6 weeks); reproduced with permission from [11].
65
Journal of the Neurological Sciences 389 (2018) 61–66
was similar to that found for the NNT vs. placebo for Clinical Global
Impression (CGI) response (very much improved or much improved): 8
[11]. The analogous results from the pooled valbenazine trials were
similar: NNT for AIMS response was 5 and NNT for CGI response was
also 5 [10].
Although each agent can exhibit adverse effects, the rates for spe-
cific adverse effects in the clinical trials were generally low, with
number needed to harm (NNH) values > 10, denoting an intervention
that would be generally expected to be tolerable on that domain [41].
The adverse event with the largest drug-placebo difference for deute-
trabenazine was insomnia (4% vs. 1%), with a resultant NNH of 34
[11]. That for valbenazine was somnolence (11% vs. 4%), with a re-
sultant NNH of 15 [10]. Overall tolerability can be quantified by
comparing rates of discontinuation because of an adverse event. A total
of 3.6% of patients receiving deutetrabenazine (all doses pooled) dis-
continued because of an adverse event, compared with 3.1% for pla-
cebo, resulting in a NNH of 189 [11]. A total of 2.9% of patients re-
ceiving valbenazine (all doses pooled) discontinued because of an
adverse event, compared with 1.6% for placebo, resulting in a NNH of
76 [10]. Thus both deutetrabenazine and valbenazine are far more
likely to result in a response (≥50% reduction in the AIMS dyskinesia
score) than a discontinuation because of an adverse event. For deute-
trabenazine or valbenazine for TD there does not appear to be an ex-
acerbation of underlying psychopathology, nor are there any clinically
relevant signals regarding depression or suicidality. Warnings regarding
QT prolongation do not preclude the use of concomitant agents that
may also prolong QT (a common occurrence as many antipsychotics or
antidepressants are associated with prolongation of the QT interval);
however, ECG monitoring may be required in at-risk situations when
undergoing dose titration (see Table 1).
How long should VMAT2 inhibitors be used? At this point in time,
indefinite duration of therapy is recommended. As evidenced in the
long-term data available from valbenazine, dyskinetic movements re-
turned within 4 weeks after discontinuation [8].
Both deutetrabenazine and valbenazine require prescribing through
specialty pharmacies to assist in navigating the approval process for
these relatively expensive medications and to also minimize potential
patient co-pays.
6. Step 5: follow-up - check your work
When treatment is initiated, a baseline assessment should be ob-
tained and the AIMS examination is recommended for this purpose.
Advantages to using the AIMS include its ubiquity both in the clinic and
in drug development. Follow-up assessments should be done on a reg-
ular basis, and the AIMS in this case can facilitate communication
among providers. Lack of documented improvement can trigger addi-
tional enquiries regarding adherence or unidentified drug-drug inter-
actions, as well as inform decisions on dose titration or changing the
intervention. Measurement-based care has the potential to improve
outcomes, as has been demonstrated in the treatment of major de-
pressive disorder [42].
Measuring the AIMS dyskinesia items alone is inadequate. Follow-
up questions regarding functional impairments attributable to TD need
to be asked (e.g., interference with activities such as eating, drinking,
speaking, breathing, dressing oneself, writing, working, leisure activ-
ities, being with others). Fortunately, the AIMS contains items that
remind the clinician to evaluate impact on function and to determine
the degree of associated distress.
7. Summary
In the past, suggested treatments for TD have been largely in-
effective or limited. There are now 2 FDA-approved treatments for TD:
deutetrabenazine and valbenazine. Both are efficacious and tolerable.
They differ in terms of labeled instructions for frequency of
L. Citrome
administration (twice daily for deutetrabenazine vs. once daily for
valbenazine), titration requirements (dose to efficacy/tolerability for
deutetrabenazine vs. titrate to target dose of 80 mg/d for valbenazine),
need for food (administer deutetrabenazine with food), drug-drug in-
teractions (consider CYP 2D6 modulators for deutetrabenazine vs. both
CYP 2D6 and CYP 3A4 for valbenazine), contraindications (hepatic
impairment for deutetrabenazine), and minor differences in adverse
event profile. However, the clinical usefulness of these treatments is
rendered moot if TD goes unrecognized. Clinicians are encouraged to
assume that TD exists in their practice, that there is a need for harm
reduction, and to use assessments such as the AIMS for the evaluation of
TD and ongoing monitoring of TD treatment.
Disclosures
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors. In the past
48 months Leslie Citrome has engaged in collaborative research with,
or received consulting or speaking fees, from: Acadia, Alexza, Alkermes,
Allergan, AstraZeneca, Avanir, Boehringer Ingelheim, Bristol-Myers
Squibb, Eli Lilly, Forum, Genentech, Intra-Cellular Therapeutics,
Janssen, Jazz, Lundbeck, Merck, Medivation, Mylan, Neurocrine,
Novartis, Noven, Otsuka, Pfizer, Reckitt Benckiser, Reviva, Shire,
Sunovion, Takeda, Teva, Valeant, Vanda.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.jns.2017.11.019.
References
[1] L. Citrome, H.A. Nasrallah, On-label on the table: what the package insert informs
us about the tolerability profile of oral atypical antipsychotics, and what it does not,
Expert. Opin. Pharmacother. 13 (11) (2012) 1599–1613.
[2] A.N.I. Pharmaceuticals, REGLAN (Metoclopramide) Tablets, for Oral Use.
Prescribing Information, Available at https://www.accessdata.fda.gov/drugsatfda_
docs/label/2017/017854s062lbl.pdf, (August 2017) (Accessed 18 October 2017).
[3] H.C. Margolese, G. Chouinard, T.T. Kolivakis, L. Beauclair, R. Miller, Tardive dys-
kinesia in the era of typical and atypical antipsychotics. Part 1: pathophysiology
and mechanisms of induction, Can. J. Psychiatr. 50 (9) (2005) 541–547.
[4] M. Carbon, C.H. Hsieh, J.M. Kane, C.U. Correll, Tardive dyskinesia prevalence in
the period of second-generation antipsychotic use: a meta-analysis, J. Clin.
Psychiatry 78 (2017) e264–e278.
[5] S.W. Woods, H. Morgenstern, J.R. Saksa, et al., Incidence of tardive dyskinesia with
atypical versus conventional antipsychotic medications: a prospective cohort study,
J. Clin. Psychiatry 71 (4) (2010) 463–474.
[6] W. Guy, ECDEU Assessment Manual for Psychopharmacology, US Department of
Health, Education, and Welfare, Washington, DC, 1976.
[7] M.R. Munetz, S. Benjamin, How to examine patients using the Abnormal
Involuntary Movement Scale, Hosp. Commun. Psychiatry 39 (11) (1988)
1172–1177.
[8] Neurocrine Biosciences, INGREZZA (valbenazine) Capsules, for Oral Use.
Prescribing Information, Available at https://ingrezza.com/HCP/PI, (October
2017) (Accessed 18 October 2017).
[9] Teva Pharmaceuticals, AUSTEDO (Deutetrabenazine) Tablets, for Oral Use.
Prescribing information, Available at https://www.accessdata.fda.gov/drugsatfda_
docs/label/2017/209885lbl.pdf, (August 2017) (Accessed 18 October 2017).
[10] L. Citrome, Valbenazine for tardive dyskinesia: a systematic review of the efficacy
and safety profile for this newly approved novel medication - what is the number
needed to treat, number needed to harm and likelihood to be helped or harmed? Int.
J. Clin. Pract. 71 (7) (2017) e12964.
[11] L. Citrome, Deutetrabenazine for tardive dyskinesia: a systematic review of the
efficacy and safety profile for this newly approved novel medication - what is the
number needed to treat, number needed to harm and likelihood to be helped or
harmed? Int. J. Clin. Pract. 71 (11) (2017) e13030.
[12] N.R. Schooler, J.M. Kane, Research diagnoses for tardive dyskinesia, Arch. Gen.
Psychiatry 39 (4) (1982) 486–487.
[13] H. Morgenstern, W.M. Glazer, Identifying risk factors for tardive dyskinesia among
long-term outpatients maintained with neuroleptic medications. Results of the Yale
Tardive Dyskinesia Study, Arch. Gen. Psychiatry 50 (9) (1993) 723–733.
[14] American Psychiatric Association, Diagnostic and Statistical Manual of Mental
Disorders, Fifth edition, American Psychiatric Publishing, May 2013.
[15] A.F. Lehman, J.A. Lieberman, L.B. Dixon, et al., Practice guideline for the treatment
66
Journal of the Neurological Sciences 389 (2018) 61–66
of patients with schizophrenia, second edition, Am. J. Psychiatry 161 (2 Suppl)
(2004) 1–56.
[16] D.C. Mamo, R.A. Sweet, M.S. Keshavan, Managing antipsychotic-induced parkin-
sonism, Drug Saf. 20 (3) (1999) 269–275.
[17] Merck, COGENTIN (Benztropine Mesylate Injection). Prescribing Information,
Available at https://www.accessdata.fda.gov/drugsatfda_docs/label/2013/
012015s026lbl.pdf, (April 2013) (Accessed 18 October 2017).
[18] L. Citrome, R. Dufresne, J.M. Dyrud, Tardive dyskinesia: minimizing risk and im-
proving outcomes in schizophrenia and other disorders, Am. J. Manag. Care 13
(Suppl) (2007) 1–12.
[19] C.U. Correll, J.M. Kane, L.L. Citrome, Epidemiology, prevention, and assessment of
tardive dyskinesia and advances in treatment, J. Clin. Psychiatry. 78 (8) (2017)
1136–1147.
[20] S. Vinogradov, M. Fisher, H. Warm, C. Holland, M.A. Kirshner, B.G. Pollock, The
cognitive cost of anticholinergic burden: decreased response to cognitive training in
schizophrenia, Am. J. Psychiatry 166 (9) (2009) 1055–1062.
[21] H.C. Margolese, G. Chouinard, T.T. Kolivakis, L. Beauclair, R. Miller, L. Annable,
Tardive dyskinesia in the era of typical and atypical antipsychotics. Part 2: in-
cidence and management strategies in patients with schizophrenia, Can. J.
Psychiatr. 50 (11) (2005) 703–714.
[22] M.A. Richardson, M.L. Bevans, L.L. Read, et al., Efficacy of the branched-chain
amino acids in the treatment of tardive dyskinesia in men, Am. J. Psychiatry 160 (6)
(2003) 1117–1124.
[23] R. Bhidayasiri, S. Fahn, W.J. Weiner, et al., Evidence-based guideline: treatment of
tardive syndromes: report of the Guideline Development Subcommittee of the
American Academy of Neurology, Neurology 81 (5) (2013) 463–469.
[24] P.P. Lerner, C. Miodownik, V. Lerner, Tardive dyskinesia (syndrome): current
concept and modern approaches to its management, Psychiatry Clin. Neurosci. 69
(6) (2015) 321–334.
[25] G.K. Thaker, J.A. Nguyen, M.E. Strauss, R. Jacobson, B.A. Kaup, C.A. Tamminga,
Clonazepam treatment of tardive dyskinesia: a practical GABAmimetic strategy,
Am. J. Psychiatry 147 (4) (1990) 445–451.
[26] W. Zheng, Y.-Q. Xiang, C.H. Ng, G.S. Ungvari, H.F. Chiu, Y.T. Xiang, Extract of
ginkgo biloba for tardive dyskinesia: meta-analysis of randomized controlled trials,
Pharmacopsychiatry 49 (3) (2016) 107–111.
[27] S. Pappa, S. Tsouli, G. Apostolou, V. Mavreas, S. Konitsiotis, Effects of amantadine
on tardive dyskinesia: a randomized, double-blind, placebo-controlled study, Clin.
Neuropharmacol. 33 (6) (2010) 271–275.
[28] S. Angus, J. Sugars, R. Boltezar, S. Koskewich, N.M. Schneider, A controlled trial of
amantadine hydrochloride and neuroleptics in the treatment of tardive dyskinesia,
J. Clin. Psychopharmacol. 17 (2) (1997) 88–91.
[29] Adamas Pharma, GOCOVRI (Amantadine) Extended Release Capsules, for Oral Use.
Prescribing Information, Available at https://www.accessdata.fda.gov/drugsatfda_
docs/label/2017/208944lbl.pdf, (August 2017) (Accessed 16 October 2017).
[30] Lundbeck, XENAZINE (Tetrabenazine) Tablets, for Oral Use. Prescribing
Information, Available at https://www.accessdata.fda.gov/drugsatfda_docs/label/
2015/021894s010lbl.pdf, (June 2015) (Accessed 19 October 2017).
[31] D.R. Guay, Tetrabenazine, a monoamine-depleting drug used in the treatment of
hyperkinetic movement disorders, Am. J. Geriatr. Pharmacother. 8 (4) (2010)
331–373.
[32] J.J. Chen, W.G. Ondo, K. Dashtipour, D.M. Swope, Tetrabenazine for the treatment
of hyperkinetic movement disorders: a review of the literature, Clin. Ther. 34 (7)
(2012) 1487–1504.
[33] H. Kazamatsuri, C. Chien, J.O. Cole, Treatment of tardive dyskinesia, I: clinical
efficacy of a dopamine-depleting agent, tetrabenazine, Arch. Gen. Psychiatry 27 (1)
(1972) 95–99.
[34] H. Kazamatsuri, C.P. Chien, J.O. Cole, Long-term treatment of tardive dyskinesia
with haloperidol and tetrabenazine, Am. J. Psychiatry 130 (4) (1973) 479–483.
[35] W.G. Ondo, P.A. Hanna, J. Jankovic, Tetrabenazine treatment for tardive dyski-
nesia: assessment by randomized videotape protocol, Am. J. Psychiatry 156 (8)
(1999) 1279–1281.
[36] C.F. O'Brien, R. Jimenez, R.A. Hauser, et al., NBI-98854, a selective monoamine
transport inhibitor for the treatment of tardive dyskinesia: a randomized, double-
blind, placebo-controlled study, Mov. Disord. 30 (12) (2015) 1681–1687.
[37] R.A. Hauser, S.A. Factor, S.R. Marder, et al., KINECT 3: a phase 3 randomized,
double-blind, placebo-controlled trial of valbenazine for tardive dyskinesia, Am. J.
Psychiatry 174 (5) (2017) 476–484.
[38] H.H. Fernandez, S.A. Factor, R.A. Hauser, et al., Randomized controlled trial of
deutetrabenazine for tardive dyskinesia: the ARM-TD study, Neurology 88 (21)
(2017) 2003–2010.
[39] K.E. Anderson, D. Stamler, M.D. Davis, et al., Deutetrabenazine for treatment of
involuntary movements in patients with tardive dyskinesia (AIM-TD): a double-
blind, randomised, placebo-controlled, phase 3 trial, Lancet Psychiatry 4 (8) (2017)
595–604.
[40] L. Citrome, Compelling or irrelevant? Using number needed to treat can help de-
cide, Acta Psychiatr. Scand. 117 (6) (2008) 412–419.
[41] L. Citrome, T.A. Ketter, When does a difference make a difference? Interpretation of
number needed to treat, number needed to harm, and likelihood to be helped or
harmed, Int. J. Clin. Pract. 67 (5) (2013) 407–411.
[42] A.S. Yeung, Y. Jing, S.K. Brenneman, et al., Clinical Outcomes in Measurement-
based Treatment (Comet): a trial of depression monitoring and feedback to primary
care physicians, Depress Anxiety 29 (10) (2012) 865–873.