Movement disorders associated with antiseizure medications: a systematic review

Abstract

New onset movement disorders have been frequently reported in association with the use of antiseizure medications (ASMs). The frequency of specific motor manifestations and the spectrum of their semiology for various ASMs have not been well characterized. We carried out a systematic review of literature and conducted a search on CINAHL, Cochrane Library, EMBASE, MEDLINE, PsycINFO, and Scopus from inception to April 2021. We compiled the data for all currently available ASMs using the conventional terminology of movement disorders. Among 5123 manuscripts identified by the search, 437 met the inclusion criteria. The largest number of reports of abnormal movements were in association with phenobarbital, valproic acid, lacosamide, and perampanel, and predominantly included tremor and ataxia. The majority of attempted interventions for all agents were discontinuation of the offending drug or dose reduction which led to the resolution of symptoms in most patients. Familiarity with the movement disorder phenomenology previously encountered in relation with specific ASMs facilitates early recognition of adverse effects and timely institution of targeted interventions.

1. Introduction

Antiseizure medications (ASMs) are commonly prescribed for seizure prophylaxis or management, and for other indications, such as bipolar disorder, anxiety, and chronic pain [1]. Both the therapeutic effectiveness and the adverse effects of the ASMs are attributed to their ability to cross the blood brain barrier and act on the targets within the central nervous system (CNS). CNS adverse effects can emerge as cognitive slowing, psychiatric disturbances, or impairment of motor controls [2]. New onset movement disorders are frequently reported in association with the use of ASMs and can manifest with symptoms of vestibular, ocular, and motor dysfunction or global impairment of motor control [3]. Recognition of these symptoms would allow for the implementation of timely interventions aimed at reducing medication toxicities, improving patient compliance, and preventing permanent neurological disability.

The development of adverse effects of ASMs may be defined by drug-specific factors (e.g., mechanism of action and pharmacokinetic profile) or clinical factors (e.g., demographic characteristics and seizure syndrome) [4,5]. Most CNS-related adverse effects, including abnormal movements, are thought to be dose-dependent and triggered by an overall decrease in cerebral neuronal activity rather than by the specific mechanisms of action of these drugs [6]. For example, diplopia and nystagmus frequently encountered during the use of sodium channel blockers for seizure prophylaxis were thought to be due to inhibition of high-frequency action potential firing in vestibular and oculomotor circuits [7]. Concomitant administration of ASMs with other centrally acting drugs may also exacerbate their neurological adverse effects. For example, myoclonus was observed in patients receiving a combination of escitalopram and lamotrigine [8]. The myoclonus was thought to be caused by additive effects of the drugs on serotonin 5-HT1A receptors or by synergetic inhibition of voltage-gated calcium channels [9]. Co-administration of the ASMs with medications that inhibit their metabolism may lead to increased serum ASM levels resulting in more severe manifestations of the CNS toxicity [10].

The objective of this review is to provide an up-to-date, comprehensive summary of the existing literature on the development of new-onset abnormal movements triggered by the administration of ASMs. In the present review, we categorized patients’ symptoms according to the conventional movement disorder terminology and recorded relevant clinical and demographic characteristics of patients presenting with specific movement disorders.

2. Materials and Methods

The systematic review protocol was developed using guidance from the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [11] and registered in the International Prospective Register of Systematic Reviews (PROSPERO; Center for Reviews and Dissemination number CRD42021242850) [12].

2.1. Search strategy

Literature searches were carried out between April 9 and 11, 2021. EMBASE (embase.com, version including 1974-present), MEDLINE (EBSCOhost), CINAHL (EBSCOhost), PsycINFO (EBSCOHost), Scopus, and The Cochrane Library (including the Cochrane Database of Systematic Reviews and The Cochrane Central Register of Controlled Trials, wiley.com) were searched from inception to the search date.

The list of drugs included in the search and movement disorder keywords were developed by a librarian (C.S.) in consultation with neurologists with subspecialty training in Epilepsy (O.T.) and Movement Disorders (M.S.K.). The ASMs included in the search were first and second generation agents (phenobarbital, phenytoin, primidone, valproate, carbamazepine, and ethosuximide) as well as new generation medications (brivaracetam, cannabidiol, cenobamate, clobazam, clonazepam, diazepam, eslicarbazepine, felbamate, fenfluramine, gabapentin, lacosamide, lamotrigine, levetiracetam, lorazepam, midazolam, oxcarbazepine, perampanel, pregabalin, rufinamide, stiripentol, tiagabine, topiramate, vigabatrin, and zonisamide) [13,14].

Movement disorder phenomenology were described according with the Movement Disorders Society classification [15]. Specifically, tremor (excluding parkinsonism), myoclonus (positive and negative), dystonia (including blepharospasm and spasmodic dysphonia), tics, akathisia, athetosis, ballism, chorea, choreoathetosis, restless leg syndrome (RLS), periodic limb movement disorder (PLMD), eyelid myokymia, and nonspecific dyskinesias were characterized as hyperkinetic movement disorders [15]. Parkinsonism was defined as a hypokinetic movement disorder [15]. Ataxia was a distinct category defined as incoordination of voluntary muscle movement [15]. Nystagmus, oculogyric crisis, opsoclonus, opsoclonus-myoclonus, and ophthalmoplegia were defined as eye movement disorders [16].

Full search strategies are available through UNMC’s digital repository at https://digitalcommons.unmc.edu/search/9. Several types of searches were performed: 1) All databases were searched for articles with a term for the ASM medication class, an ASM family, or one of the included drug’s English generic names in close proximity to one of a list of words for the “induced” concept that was itself in close proximity to a word for the “movement disorder” concept. 2) Databases with heading/subheading systems (EMBASE, MEDLINE, CINAHL, and The Cochrane Library) were searched for articles with a) an ASM heading in combination with an adverse effect-related subheading and b) a movement disorder heading in combination with an etiology-related subheadings. Because of the extensive indexing of EMBASE records, EMBASE searches required that the ASM-related and movement disorder-related headings be major headings. 3) Triple-indexed records in EMBASE were searched for articles indexed with an anti-epileptic drug heading linked to an “adverse drug reactions” subheading, which in turn was linked to a movement disorder-related term. Major heading designation was not required in these searches.

Because no funds were available for translation, search results were limited to English-language articles. Conference abstracts were removed when database filters allowed. Filters in CINAHL and PsycINFO were used to limit results to academic journal articles. Records indexed as concerning animals that were not also indexed as concerning humans were removed from the EMBASE search. Because the lengthy searches taxed the abilities of the other databases’ search engines, no attempt to remove animal studies was made in these databases.

The database searches retrieved 7469 total records (284 from CINAHL, 246 from the Cochrane Library, 4460 from EMBASE, 1190 from MEDLINE, 326 from PsycINFO, and 963 from Scopus). All search results were imported into RefWorks. Of these records, 2346 duplicates were removed using RefWorks’ duplicate detection tool, and 1804 were categorized by RefWorks as review articles, animal studies, and conference abstracts, to be excluded. Consequently, 3319 records for unique publications remained for title/abstract review. Two independent reviewers (D.Z. and S.P.) reviewed the article titles and abstracts and then the selected articles (D.Z. and I.S.). Disagreements between reviewers and inquiries by reviewers were resolved by another reviewer.

2.2. Eligibility criteria

The title, abstract, or full text articles that contained a mention of a new onset movement disorder and ASM were reviewed (Fig. 1). Reports that were not in English, not peer-reviewed (other than letters to the editor), or those that contained no original data were excluded. Reports of exacerbation of a previous movement disorder during the use of the ASM and articles on movements consistent with seizures or those in the settings of severe metabolic derangements known to cause abnormal movements were excluded. Manifestations of abnormal movements reported to be caused by a combination of ASMs were recorded separately. Reports of movement disorders arising in the settings of co-administration of ASM with other agents in which the precipitating factor for abnormal movements was not clear were excluded.

Figure 1.

Flow diagram of study selection process.

2.3. Data extraction

We extracted the numbers of patients presenting with various movement disorders for all ASMs. If available, we recorded patients’ age, sex, clinical indication of ASM, maximum daily dose and route of administration of ASM, duration of drug use prior to the onset of movement disorder, and serum ASM levels. We also extracted the data on interventions aimed to alleviate the abnormal movements as well as outcomes of these interventions. We categorized duration of ASM prior to the onset of adverse effects as acute (<3 days), subacute (3 days to 3 weeks), and chronic (>3 weeks).

2.4. Quality of evidence

Case reports and case series were evaluated based on completeness and quality of reporting and were excluded if they did not provide the pertinent information [17]. In the large studies, the risk of bias was assessed based on Cochrane Risk of Bias tools for randomized and non-randomized studies [18,19].

2.5. Data synthesis

The information from all reports was consolidated and total counts of movement disorders was recorded for each ASM. Cumulative incidences of movement disorders were calculated, when possible. Mean age for each group was determined through weighted means, when possible. The doses, routes of administration, interventions, and outcomes were consolidated. Comparisons of results between ASMs were descriptive only and not statistically assessed.

3. Results

Our search yielded 5123 manuscripts (duplicates removed), of which 437 met eligibility criteria with publication dates from 1973 to 2021 yielding 4481 patients with new-onset abnormal movements associated with ASMs (Fig. 1). The overall distribution of movement disorders with various phenomenology is summarized in Table 1 while the details of demographic characteristics or treatment regimens and clinical outcomes are reported separately (Table 2).

Table 1.

Antiseizure medications and movement disorder phenomenology.

ASM	Number of patients	Movement disorders (number of patients)
Brivaracetam	26	Tremors (9), ataxia (17)
Carbamazepine	147	Akathisia (3), choreoathetosis (1), dystonia (17), myoclonus (3/+, 6/−), RLS (1), tics (8), tremors (55), other dyskinesia (2), parkinsonism (2), ataxia (48), nystagmus (28), oculogyric crisis (5), ophthalmoplegia (2)
Cenobamate	13	Ataxia (13)
Clobazam	18	Dystonia (1), tics (1), tremors (3), parkinsonism (1), ataxia (12)
Clonazepam	32	Tics (1), ataxia (31)
Diazepam	12	Tremors (3), other dyskinesia (1), parkinsonism (4), ataxia (3), ophthalmoplegia (1)
Eslicarbazepine	220	Myoclonus (1/+), tremors (17), ataxia (202)
Ethosuximide	5	Akathisia (2), chorea (1), other dyskinesia (2), ataxia (2)
Felbamate	4	Akathisia (1), dystonia (2), tics (1), ataxia (1), nystagmus (1)
Fosphenytoin	36	Tremors (5), other dyskinesia (2), ataxia (8), nystagmus (30)
Gabapentin	98	Ballism (1), chorea (4), choreoathetosis (5), dystonia (3), myoclonus (32/+, 20/−), tics (1), tremors (5), other dyskinesia (7), parkinsonism (1), ataxia (32), nystagmus (1), oculogyric crisis (1)
Lacosamide	452	Myoclonus (5/+), tics (1), tremors (303), other dyskinesia (3), ataxia (140), nystagmus (86)
Lamotrigine	130	Ballism (1), chorea (2), choreoathetosis (1), dystonia (7), myoclonus (8/+), tics (14), tremors (66), other dyskinesia (1), parkinsonism (3), ataxia (27), nystagmus (5), oculogyric crisis (4), opsoclonus (1)
Levetiracetam	105	Choreoathetosis (1), tics (2), tremors (71), other dyskinesia (6), parkinsonism (2), ataxia (23)
Lorazepam	3	Myoclonus (2/+), other dyskinesia (1)
Midazolam	7	Akathisia (1), athetosis (1), dystonia (1), myoclonus (1/+), RLS (1), tremors (1), other dyskinesia (2), parkinsonism (1)
Oxcarbazepine	27	Myoclonus (6/+), tremors (8), other dyskinesia (1), parkinsonism (1), ataxia (7), nystagmus (2), oculogyric crisis (1)
Perampanel	264	Myoclonus (1/−), tremors (10), ataxia (253), oculogyric crisis (1)
Phenobarbital	1623	Dystonia (3), myoclonus (2/−), tics (6), other dyskinesia (3), ataxia (1614), nystagmus (2), oculogyric crisis (1)
Phenytoin	209	Athetosis (4), ballism (2), chorea (16), choreoathetosis (25), dystonia (18), myoclonus (2/+, 31/−), tics (1), tremors (16), other dyskinesia (30), parkinsonism (2), ataxia (103), nystagmus (55), ophthalmoplegia (3), opsoclonus-myoclonus (1)
Pregabalin	183	Akathisia (8), dystonia (3), myoclonus (10/+, 14/−), tremors (18), other dyskinesia (5), parkinsonism (7), ataxia (122), nystagmus (2), ophthalmoplegia (1)
Primidone	2	Ataxia (2)
Rufinamide	6	Tremors (1), ataxia (5)
Stiripentol	44	Tremors (15), ataxia (33)
Tiagabine	4	Athetosis (1), dystonia (3)
Topiramate	37	Dystonia (1), eyelid myokymia (9), myoclonus (6/+), RLS (4), PLMD (1), tremors (6), ataxia (11)
Valproate	623	Akathisia (13), chorea (7), choreoathetosis (1), dystonia (12), myoclonus (2/+, 7/−), tremors (513), other dyskinesia (1), parkinsonism (60), ataxia (52), nystagmus (11)
Vigabatrin	34	Athetosis (1), chorea (2), choreoathetosis (3), dystonia (2), myoclonus (2/+), tremors (2), other dyskinesia (22), parkinsonism (1), ataxia (4)
Zonisamide	37	RLS (3), tremors (12), parkinsonism (1), ataxia (33)

/+, positive myoclonus. /−, negative myoclonus.

Table 2.

Patient demographics, clinical course, and management outcomes.

				Mean daily dose		Route of delivery		Serum level Monitoring	Latency to symptom onset			Action taken			Outcome			Time to improved outcome (days)
Antiseizure medication	Cases (n)	Mean age (years)	% Female	mg	mg/kg	Oral (n)	IV (n)	Serum levels (n, % normal)	<3d	3d-21d	>21d	D/C	↓ dose	No Δ	++	+	−	<3d	3d-21d	>21d
Carbamazepine	41	32.0	46%	855	11.2	41		30 (63%)	12	8	21	25	12	4	38	2	1	21	7	6
Clonazepam	1	37.0	0%	8.0		1					1		1			1				1
Diazepam	2	70.0	0%	6.0		2				1	1	2			2			2
Ethosuximide	3	8.7	33%	483		3			3			3			2	1		2	1
Felbamate	3	19.1	0%	1800	41.0	3		2 (50%)	2			2	1		3			1	2
Fosphenytoin	3	17.6	67%	536			3	1 (100%)	2	1		3			3			3
Gabapentin	32	56.8	53%	1120		32		4 (75%)	12	7	9	29	3		32			19	7	4
Lacosamide	3	9.7	67%		9.0	3			1	1	1	1	2		3			2		1
Lamotrigine	41	26.8	56%	254	12.9	41		11 (9%)	9	5	24	23	17	1	35	5	1	11	3	13
Levetiracetam	3	40.7	33%	917		2	1		2		1	2	1		1	2		1	1
Lorazepam	3	20.0	100%	9.0	0.25	1	2		2	1		3			2	1		2		1
Midazolam	7	42.3	43%	5.8	0.15	2*	5		5		1	7			7			6	1
Oxcarbazepine	5	27.2	60%	1050	30.0	5			2	2	1	4		1	5			2	2	1
Phenobarbital	10	10.6	20%	97.4	12.5	9	1	4 (75%)	2	5	2	10			5	4	1	1	4	2
Phenytoin	89	30.2	40%	395	14.5	70	19	80 (36%)	20	19	48	63	17	9	71	16	2	32	31	15
Pregabalin	17	61.3	65%	180		17†			9	2	5	17			15	2		5	5	5
Tiagabine	4	21.3	50%	27.5		4			4				1	3	4				2	1
Topiramate	13	35.3	77%	140		13			5	4	4	10	2	1	12	1		7	3	2
Valproate	35	47.4	49%	1223	17.9	34	1	21 (90%)	2	7	25	26	4	5	27	7	1	4	10	15
Vigabatrin	4	4.4	25%	2000	150	2	2		2	1	1	3	1		4			1		2
Zonisamide	2	38.5	100%	300		2		1 (100%)	2			1	1		1		1	1

^*Subcutaneous route.

^†Intranasal route.

IV, intravenous. <3d, acute (<3 days). 3d-21d, subacute (3 to 21 days). >21d, chronic (>21 days). D/C, discontinued. ↓ dose, decreased dose. No Δ, no change. ++, resolved. +, improved. −, no change.

3.1. Phenobarbital and primidone

Phenobarbital was encountered in the largest number of reports of new onset abnormal movements, with a total of 1623 patients described in 10 papers (Table 1) [20–29]. The most frequent manifestation was ataxia (1614 cases, cumulative prevalence of 19.6%) [27–29]. Other adverse effects included tics (6 cases), dystonia (3), negative myoclonus (2), oculogyric crisis (1), and other dyskinesia (3).

Among all studies, there were 6 reports of 10 patients for whom further details were available (Table 2) [20–25]. The average cumulative daily dose of phenobarbital was 97.4 mg (or 12.5 mg/kg). Serum levels of phenobarbital were obtained in 4 patients and were above therapeutic range in one patient. Phenobarbital was discontinued in all patients which resulted in resolution of symptoms in half of the patients and improvement of movements in 40% of patients. We also identified 2 case reports of ataxia induced by primidone [29].

3.2. Phenytoin and fosphenytoin

Phenytoin was described in association with various acute-onset abnormal movements in 80 reports of 209 patients (Table 1) [26,29–107]. Ataxia was the most frequently reported symptom and was described in 103 patients (13.5%). Various hyperkinetic disorders also included myoclonus (33 cases), athetosis (4), chorea (16), choreoathetosis (25), ballism (2), dystonia (18), tremors (16), and other unspecified dyskinesias (30). There were 55 cases of nystagmus, of which four and five presentations were characterized as downbeat [33,36] and horizontal nystagmus [37,45,57,68,75], respectively. Other less commonly encountered complications included tics (1), parkinsonism (2), ophthalmoplegia (3), and opsoclonus-myoclonus (1).

Among all studies, additional information was available in 67 reports of 89 patients (Table 2) [30–94,100,107]. These patients with new onset abnormal movements received an average 395 mg (or 14.5 mg/kg) of phenytoin daily prior to the development of adverse effects. Of 80 patients who had serum phenytoin levels recorded, 51 patients (63.8%) were noted to have their levels exceeding the therapeutic range. The majority of patients had a sustained exposure to phenytoin prior to the occurrence of adverse effects. Specifically, among 87 patients for whom the duration of treatment was recorded, 48 patients (55.2%) received phenytoin for over 3 weeks. On the other hand, only 20 patients (23.0%) developed symptoms within 3 days from the initial exposure to phenytoin. The most frequent management approach undertaken to combat these adverse effects was discontinuation of phenytoin (63, 70.8%). This was followed by dose reduction in 17 patients (21.3%). In one case of acute focal dystonia associated with phenytoin, the patient was treated with benztropine [44]. The dystonia resolved, and the patient was continued on the same dose of phenytoin.

Fosphenytoin was reported to induce abnormal movements in 36 patients (Table 1) [108–111]. In a clinical trial of 60 patients who received intramuscular fosphenytoin, 28 patients developed nystagmus, while 8 and 5 manifested ataxia and tremor, respectively [111]. Three case reports described the presence of dystonia and two reports described nystagmus, one of which was characterized as peduncular; the symptoms resolved following discontinuation of the drug [108–110].

3.3. Valproate

We identified 78 studies of 623 patients who received valproate (in various preparations such as valproic acid, sodium valproate, and divalproex sodium) and developed new onset abnormal movements (Table 1) [27,95,97–99,101–103,112–181]. The most frequent movement disorders were tremor (513, 11.6%), parkinsonism (60, 6.1%), and ataxia (52, 9.9%). Other manifestations of hyperkinetic disorders included akathisia, chorea, choreoathetosis, dystonia, myoclonus, and other dyskinesias in 43 patients. In a large prospective cohort study, the odds of developing parkinsonism in patients taking valproate were 5 times higher than in those taking all other ASMs, including carbamazepine, phenytoin, topiramate, clobazam, and lamotrigine [103].

We extracted further details from a subset of these studies (27 case reports of 35 patients) (Table 2) [112–138]. The mean cumulative daily dose of valproate in these patients was 1223 mg, with 34 cases of it administered orally and one administered intravenously (IV). Serum levels were available in 21 patients (60%) among whom only 2 patients (9.5%) had levels above the therapeutic range. More than two thirds of patients (73.5%) developed adverse effects following prolonged exposure to valproate for 3 or more weeks while 2 patients (5.9%) manifested the symptoms acutely, within 3 days from initiation of the drug. The interventions included discontinuation and dose reduction of valproate in 74.2% and 11.4% of patients, respectively. Of the patients who had an improvement with these interventions, the latency to improvement was at least 3 weeks in more than half of the patients (51.7%). Additionally, 7 patients with valproate-induced parkinsonism and one with chorea were treated with levodopa [123,126,130] and benzodiazepines and haloperidol [124], respectively.

3.4. Carbamazepine

We identified 43 reports of 147 patients with carbamazepine-induced abnormal movements, which included tremor (55, 12.2%), ataxia (48, 3.7%), nystagmus (28), and dystonia (17, 5.8%) (Table 1) [26,95,103,141,157,178,181–217]. In one pediatric study, 19 children (24.7%) developed nystagmus in the settings of accidental or intentional ingestion of carbamazepine [182].

Further details of clinical characteristics were available in 29 reports of 41 of the patients (Table 2) [189–217]. Specifically, these patients developed akathisia (3 cases), choreoathetosis (1), blepharospasm (1), other dystonia (7), myoclonus (8), restless leg syndrome (1), orofacial dyskinesia (2), parkinsonism (2), nystagmus (9), oculogyric crisis (5), and ophthalmoplegia (2). The nystagmus was characterized as downbeat, gaze-evoked, and horizontal with direction-changing or torsional components in 3 patients in each category [190,196,197,206,212,213,216]. Patients treated with carbamazepine received a mean cumulative daily dose of 855 mg (or 11.2 mg/kg) prior to onset of the abnormal movements. Of the 30 patients for whom the serum levels were reported, 11 patients (36.7%) had levels above therapeutic range. The symptoms manifested after chronic use (i.e., >21 days) in over half of patients (51.2%). The complications developed acutely in 29.3% of these patients and subacutely in 19.5% of patients. To remediate the adverse effects, carbamazepine was discontinued in 61% of patients and its dose was reduced in 29.3% of patients. This resulted in resolution of symptoms in most patients (92.7%). Carbamazepine-induced dystonia described in 4 cases was additionally treated with promethazine, diazepam, trihexyphenidyl, or gastric lavage [192,195,198,213].

3.5. Benzodiazepines

Clonazepam was reported to cause ataxia (31) and Tourette syndrome (1) (Table 1) [165,218,219]. Diazepam was reported to be associated with abnormal movements in 12 patients, including parkinsonism (4, 14.8%), ataxia (3, 2.3%), tremor (3, 2.3%), divergence paralysis (1), and oral-buccal dyskinesia (1) (Table 1) [220–223]. There were 7 case reports of midazolam-induced akathisia, athetosis, dystonia, myoclonus, RLS, tremor, parkinsonism, and orofacial or lingual dyskinesias, all of which resolved following midazolam discontinuation (Table 2) [224–230]. Lorazepam was reported to be associated with myoclonus (2) and orofacial dyskinesias (1) (Table 1) [231,232].

3.6. Clobazam and ethosuximide

There were five reports of 18 patients with clobazam-induced ataxia (12), tremors (3), dystonia (1), tics (1), and parkinsonism (1) (Table 1) [95,103,155,233,234]. Ethosuximide was reported in association with ataxia (2), akathisia along with orofacial dyskinesias (2), or generalized chorea (1) (Table 1) [29,235,236].

3.7. Lacosamide

We identified 26 reports of 452 patients with new-onset abnormal movements associated with the use of lacosamide (Table 1) [186,237–261]. The most common presentations were tremor (303, 7.1%) and ataxia (140, 6.3%). There were 86 reported cases of nystagmus, 52 (60.5%) of which were also included in cases of ataxia. Other hyperkinetic disorders included tics (1), positive myoclonus (5), and unspecified dyskinesias (3). Further, lacosamide triggered orofacial and limb dyskinesia in 3 patients [237].

3.8. Perampanel

We identified 19 studies of 264 patients with abnormal movements triggered by perampanel (Table 1) [262–280]. The most frequently reported manifestation was ataxia (253, 4.8%). Other movement disorders included tremor (10), negative myoclonus (1), and oculogyric crisis (1).

3.9. Gabapentin and pregabalin

We identified 36 studies of 98 patients with gabapentin-associated abnormal movements, including myoclonus (52, 12.5%) and ataxia (32, 6.7%) (Table 1) [98,103,281–314]. Among these studies, further details were available for 32 patients (Table 2) [289–313]. Specifically, 12 patients developed positive myoclonus, of which 7 had elevated serum creatinine and blood urea nitrogen (BUN) levels. Three other patients had negative myoclonus. Other manifestations included ballism (1 case), chorea (4), choreoathetosis (5), dystonia (3), tics (1), tremors (5), oculogyric crisis (1), parkinsonism, and other dyskinesias (7). The mean cumulative daily dose of gabapentin was 1120 mg. Nearly 40% of patients for whom the onset of symptoms was documented developed adverse effects within 3 days from the initiation of treatment while the remaining patients had subacute (25%) or delayed (32.1%) onset of these complications. The gabapentin dose was reduced in 3 patients while drug was discontinued in all other cases. In addition, some patients with myoclonus underwent dialysis [295,315,316] or received haloperidol, diphenhydramine, lorazepam, or procyclidine [289,290,304,305]. In all instances, the interventions resulted in resolution of symptoms. The recovery was achieved in nearly two-thirds of the patients (63.3%) within 3 days after the intervention and was delayed in the remaining patients.

There were 32 reports of 183 patients with abnormal movements associated with pregabalin, mostly including ataxia (122, 7.6%), myoclonus (24, 2.1%), and tremor (18, 10.6%) (Table 1) [282,283,317–346]. These included case reports of 17 patients with mean daily dose of pregabalin of 180 mg (Table 2) [330–346]. In these studies, myoclonus developed in 8 patients, of whom 4 had elevated serum creatinine and BUN while downbeat nystagmus emerged in 2 patients [331,334]. Other manifestations included ataxia (5), akathisia (1), blepharospasm (1), and parkinsonism (4). Motor symptoms developed acutely in more than a half of the patients (56.3%). All patients discontinued pregabalin which led to either improvement or complete resolution of symptoms. Abnormal movements improved within 3 days in 33% of patients.

3.10. Oxcarbazepine and eslicarbazepine

We identified 10 studies of 27 patients with abnormal movements triggered by oxcarbazepine, most commonly tremors (8, 0.8%), ataxia (7, 2.1%), and positive myoclonus (6, 0.7%) (Table 1) [95,347–355]. They also included two incidences of jerky see-saw and multidirectional nystagmus which resolved after carbamazepine was discontinued [351,355]. Other rare manifestations included tardive dyskinesia (1), parkinsonism (1) and oculogyric crisis (1) [352–354]. We also identified 11 reports of 220 patients with ataxia (202, 3.2%), tremor (17, 2.1%), and positive myoclonus (1, 5.6%) following administration of eslicarbazepine (Table 1) [184,356–365].

3.11. Lamotrigine

Lamotrigine-induced abnormal movements were encountered in 38 studies reporting 130 patients whose manifestations included tremor (66, 12.1%), ataxia (27, 11.2%), tics (14), and myoclonus (8) (Table 1) [95,101,103,105,155,173,178,181,366–395]. Additional data were available in 41 of these patients (Table 2) [366–390]. Specifically, 14 patients developed tics, 10 patients had abnormal eye movements, and 7 suffered from various forms dystonia. Five of these patients developed nystagmus with downbeat (3) or horizontal gaze-evoked components (1) [367,374,379,380]. The mean cumulative daily dose of lamotrigine in this subset of patients was 254 mg (or 12.9 mg/kg). Serum levels were obtained in 11 patients and were above therapeutic range in 10 patients (90.9%). The movement disorders developed acutely in 9 patients (23.7%), subacutely in 5 patients (13.2%), and after chronic use in 23 patients (60.5%). To treat these complications, lamotrigine was discontinued in over half of the patients (56.1%), and its dose was reduced in 41.4% of patients. These measures lead to resolution or significant improvement of symptoms in the majority of instances (97.6%). One patient with simultaneous manifestations of blepharospasm, complex tics, and obsessive-compulsive disorder was also treated with risperidone [368]. There was variability in times to resolution of symptoms; 40.7% of patients have recovered within 3 days after the intervention and 48.1% of patients still experienced symptoms after 3 weeks.

3.12. Levetiracetam and brivaracetam

There were 13 reports of 105 patients with new onset movements triggered by administration of levetiracetam (Table 1) [95,155,157,173,178,183,187,392,396–400]. These manifestations included tremor (71, 8.2%) and ataxia (23, 20.2%). Other symptom included parkinsonism (2), which improved after adjustment of the dose or discontinuation of the drug [398,400]. Brivaracetam was found to be associated with tremor (9, 2.7%) or ataxia (17, 2.0%) in large observational studies (Table 1) [401–405].

3.13. Topiramate, zonisamide, and stiripentol

We identified 16 studies of 37 patients with new-onset abnormal movements induced by topiramate (Table 1) [103,173,406–419]. These manifestations included ataxia (11, 8.5%), eyelid myokymia (9, 5.7%), tremor (6, 3.0%), myoclonus (6), and RLS (4). Among these, in 13 patients (Table 2) [406–414], topiramate was administered at a mean dose of 140 mg daily for the prophylaxis of seizures or migraine headache as well as treatment of eating disorders. Symptoms developed acutely, subacutely, or were delayed in similar proportions of patients. To treat these adverse effects, topiramate was discontinued in 76.9% of patients, and its dose was reduced in 15.4% of patients. One patient with multifocal myoclonus was additionally treated with cyproheptadine [406], and another patient with new onset restless leg syndrome was administered cabergoline [411]. The abnormal movements resolved in 12 patients (92.3%) and significantly improved in the other patient (7.7%). Over half of the patients achieved recovery within 3 days from interventions.

Zonisamide was associated with 37 cases of abnormal movements, which primarily included ataxia (33, 4.4%) and tremor (12, 3.0%), RLS (3, 2.5%), or parkinsonism (1) (Table 1) [103,420–427]. In two case reports, RLS resolved in one of the patients following drug cessation and remained unchanged in the other patient despite dose decrease [426,427]. Among 44 patients treated with stiripentol, 33 (29.8%) developed ataxia and 12 (12.1%) developed tremor (Table 1) [428–432].

3.14. Vigabatrin and tiagabine

There were 9 reports of 34 patients with vigabatrin-induced abnormal movements, most of which were hyperkinetic and manifested as athetosis (1), chorea (2), choreoathetosis (3), dystonia (2), myoclonus (1), tremors (2), or other unspecified dyskinesias (22) (Table 1) [103,433–440]. Some patients improved after dose reduction or discontinuation of vigabatrin (Table 2).

Tiagabine was found to be associated with athetosis in 1 patient and dystonia in 3 patients (Table 1) [441,442]. In all cases, the movement disorders resolved following discontinuation or dose reduction of the tiagabine (Table 2).

3.14.1. Other antiseizure medications

In a randomized control trial, cenobamate was found to be associated with ataxia in 13 patients (4.0%) (Table 1) [443]. Rufinamide was associated with ataxia in 5 patients (5.6%) [444,445] and tremor in 1 patient (0.6%) (Table 1) [446]. Felbamate was associated with tics in 1 patient (2.6%) [447] while ataxia with downbeat nystagmus was reported in another patient with felbamate intoxication [448]; dystonia was reported in 2 patients (Table 1) [449].

3.15. Movement disorders associated with combinations of ASMs

We identified 16 reports of 80 patients with new onset abnormal movements that developed during the administration of ASM combinations [98,168,185,212,379,401,450–459]. The common combination was valproate and lamotrigine, representing 46 patients with manifestations of abnormal movements which included tremor (42, 20.5%), ataxia (5), opsoclonus (2), nystagmus (2), and tics (2) [168,185,379,454,457,459].

Four other case reports described 8 patients with uncommon manifestations of abnormal movements that developed in patients treated with various ASM combinations. In one patient, the combination of IV phenytoin and diazepam administration for seizures triggered an acute reversible opsoclonus, ataxia, and dysarthria [450]. Primidone administered with phenobarbital over 8 years resulted in periodic alternating nystagmus in one patient [451]. Intentional ingestion of toxic doses of carbamazepine and phenytoin triggered choreoathetosis, dystonia, and nystagmus in another patient, and the symptoms were successfully treated with induced vomiting and administration of magnesium sulfate-charcoal [212]. Combined lamotrigine and phenytoin induced chorea in 3 patients [453].

Other larger studies included various combinations of phenytoin, valproate, carbamazepine, gabapentin, lacosamide, levetiracetam, brivaracetam, and tiagabine causing ataxia, tremor, or asterixis [98,401,455,458]. Furthermore, while not matched to specific ASMs and therefore not included in our counts of abnormal movements, in a large single-center study of 201 patients, Zadikoff et al. described 89 patients (44.3%) who developed postural or action tremor [103].

4. Discussion

In the present systematic review of literature, we compiled the available data on association of currently used ASMs with new onset movement disorders of various phenomenology. The spectrum of manifestations of abnormal movements was broad, and the latency to onset of symptoms has varied from a few hours to several days. The majority of adverse effects were reversible and resolved following the discontinuation or dose reduction of ASMs.

A previous review of the abnormal motor manifestations associated with 13 different ASMs was compiled more than 15 years ago [3]. The categorization of the abnormal movements used in the previous systematic review was similar to the system employed in our study. However, we used a more granular approach and applied additional and more specific movement disorder subcategories, such as restless legs syndrome and eyelid myokymia, to highlight the entire spectrum of motor manifestations of the ASM toxicity. We also explored ASM dosing, identified toxic levels, and timing of onset and resolution based on duration of ASM use and timing after ASM dose changes. In the previous review, the authors found that phenytoin and carbamazepine were frequently reported in association with ataxia and dyskinesias while valproate was associated with tremor and parkinsonism [3]. Our present findings are largely in agreement with those reported in the previous review. In the present review, we also included 18 additional ASMs and reviewed 311 new relevant reports that have appeared in the literature since the previously published review. To our knowledge there were no other systematic reviews of the literature on this topic.

Understanding of pathophysiological mechanisms underlying the development of new abnormal movements can help to anticipate their emergence. However, the knowledge of these mechanisms is currently lacking, and the available data are largely restricted to the animal studies. Given the higher propensity of specific movement disorders to appear following administration of certain ASMs, such as parkinsonism with valproate or tics with lamotrigine, it is likely that the development of adverse effects is related to the specific mechanisms of action for each ASM. Likewise, while the same movement disorder may appear during treatment with various ASMs, such as myoclonus with gabapentin or phenytoin, the pathophysiology of these manifestations is likely distinct for different drugs. Since the majority of symptoms have resolved after adjustment of the medication doses, these presentations appeared to be most consistent with a type A category of adverse effects, which are dose-dependent and related to the pharmacologic properties of the drug [460]. However, in some cases where the movement disorders persistent despite the adjustment of the ASM dose, a type B (idiosyncratic) adverse effects could be present. In many patients the relevant data were missing; therefore, an accurate categorization of adverse effects was not feasible.

Phenobarbital and phenytoin were among the oldest ASMs that were notorious for causing ataxia. It is proposed that ataxia in phenobarbital and phenytoin use is triggered by folate depletion [165]. Chronic phenytoin use is also known to be associated with reduced cerebellar volume, that can further contribute to worsening of ataxia [96,461]. Despite the high number of reports of phenytoin-induced dyskinesia, the pathophysiology of this manifestation remains poorly understood. In preclinical studies, a two-week long exposure to phenytoin in rats induced a higher number of stereotypies following a challenge with apomorphine, a dopamine receptor agonist [462].

Valproate was frequently reported in association with tremor and parkinsonism. The putative pathophysiologic mechanisms of valproate-induced parkinsonism included an enhancement of the GABAergic signaling in the basal ganglia and altered gene expression in dopamine signaling by upregulation of the expression of dopamine transporter, thus leading to increased dopamine reuptake and decreased availability in the synaptic cleft [463]. The valproate-induced unmasking or potentiation of the overt pre-existing degeneration of dopaminergic neurons in the substantia nigra was also proposed to contribute to the manifestations of parkinsonism [463]. Tremor and various other hyperkinesias, particularly choreoathetoid movements, were thought to result from the valproate-induced increase in inhibitory tone of the basal ganglia neurons and resultant disinhibition of the efferent pathways to the motor cortex [116,117,122].

Carbamazepine-induced hyperkinetic disorders are thought to be related to its effects on dopaminergic neurotransmission in the basal ganglia pathways. More specifically, carbamazepine induced a biphasic effect on extracellular dopamine release in the striatum of rats such that dopamine levels were increased following the administration of lower doses of carbamazepine and decreased in response to the toxic doses of drug [464]. Other animal studies similarly reported that administration of carbamazepine appears to enhance dopamine release in the striatum [465,466]. Carbamazepine was also noted to increase serotonin levels, measured in the hippocampus of rats [467], which could lead to manifestations of myoclonus. Ataxia and nystagmus which were often reported in patients treated with carbamazepine were thought to be caused by a direct transient effect of the drug on brainstem nuclei and cerebellum [206].

Gabapentin and pregabalin were found to be frequently associated with new onset myoclonus. The mechanism of this manifestation remains unclear [282]. One theory suggests that myoclonus emerges due to the action of these drugs on voltage-gated calcium channels in the thalamocortical pathways [283]. Another thought was that gabapentin and pregabalin could cause an increase in serotonin levels in the brain and peripheral nervous system, thereby promoting the development of myoclonus [468–470]. In animal studies, the use of serotonin receptor agonists precipitated the myoclonus [471,472]. Myoclonus was frequently encountered in patients with high serum levels of gabapentin or pregabalin and poor renal function and symptoms have subsided following the dose reduction or institution of renal dialysis.

Lamotrigine was associated with tremor, ataxia, tics, dystonia, and abnormal eye movements. The pathophysiologic mechanism underlying the emergence of tics was proposed to be related to blockade of presynaptic glutamate release by lamotrigine leading to enhancement of dopaminergic transmission in the substantia nigra and striatum [383]. Similar indirect effect of lamotrigine on the dopaminergic pathways may also be associated with manifestations of tremors, oculogyric crisis, and dystonia [371,378,385]. At toxic levels, lamotrigine can cause a transient dysfunction of the vestibulo-cerebellar pathways, which may explain some presentations of ataxia and nystagmus [367,380].

Topiramate and zonisamide were found to be associated with restless leg syndrome and periodic limb movements of sleep. It was proposed that topiramate and zonisamide can modulate the neurotransmission in dopaminergic pathways indirectly by enhancing the GABAergic neurotransmission and inhibiting the glutamatergic neurotransmission in the brain [411,426,427,473]. Benzodiazepines and barbiturates that are GABA_A receptor agonists are thought to trigger dyskinesia via potentiation of the inhibitory tone in the basal ganglia pathways [25,225,227]. Levetiracetam-induced parkinsonism was attributed to the indirect inhibition of dopamine release in patients with Huntington disease [398,400]. Perhaps in a similar way, dysregulation of dopaminergic signaling may also result in hyperkinetic disorders during treatment with levetiracetam [399].

We identified the reports of various forms of nystagmus in patients treated with ASMs, including downbeat nystagmus and gaze evoked nystagmus. Nystagmus can be generated by disruption of the fixation, vestibulo-ocular reflex or gaze-holding mechanisms. The directionality of nystagmus may provide clues to localization of dysfunction within the CNS. For example, gaze-evoked nystagmus involves disruption of neurotransmission within the horizontal and vertical gaze centers in the caudal pons and midbrain, respectively and was reported during exposure to exogenous GABA [474]. This form of nystagmus was found in patients exposed to phenytoin, carbamazepine, lamotrigine, and oxcarbazepine [31,42,53,57,60,74,81,196,206,212,216,355,374]. On the other hand, downbeat nystagmus is thought to occur in dysfunction of vestibulo-cerebellar pathways and caudal medulla. We found several reports describing downbeat nystagmus in association with phenytoin, carbamazepine, lamotrigine, pregabalin, and felbamate [33,36,196,212,213,331,334,367,380,448,452,475]. Periodic alternating nystagmus was found to be associated with phenobarbital and primidone and was thought to be caused by the medication effects on the cerebellar nodulus [451,475]. One patient taking oxcarbazepine was found to have jerky see-saw nystagmus, which in other disorders is associated with dysfunction of the medial longitudinal fasciculus and lesions to the midbrain, pons, or cerebellum [351]. Pendular nystagmus that developed acutely following the administration of IV fosphenytoin was thought to be resulting from blockade of sodium conductance in the brainstem neurons [109]

Our study has several limitations. Because of the limitations of the literature database search engines and the number of ASMs under consideration only drug class names, English generic names and subject headings were used to identify the drugs in the searches performed. Unindexed records utilizing proprietary names, scientific names, some of the foreign generic names, or investigational names for the ASMs would have been missed. Despite the rigorous approach taken in categorization of movement disorders, it is possible that we did not capture all manifestations of abnormal movements. While we made all attempt to recognize the description of neurological comorbidities that may have led to manifestations of abnormal movements, these conditions were not always explicitly reported. Therefore, it is possible that some patients have had additional factors independently contributing to the development of abnormal movements. Finally, in the larger studies, patients were likely taking multiple ASMs; therefore, it is possible that their adverse effects were complex and more than one ASM contributed to the toxicity.

5. Conclusions

This systematic review provides an up-to-date summary of the movement disorders reported in association of ASMs. The most frequently encountered manifestations of abnormal movements were tremor, ataxia, nystagmus, and various hyperkinetic disorders, as well as parkinsonism which were largely reversed following the removal of the offending drugs. The knowledge of the described associations and their clinical phenotypes may facilitate early diagnosis and treatment of these disorders.

Highlights.

• Administration of antiseizure medications (ASMs) is often associated with the development of new onset movement disorders

• Phenobarbital and valproic acid, are most frequently associated with various movement disorders, including tremor and ataxia

• Nearly all patients with ASM-induced movement disorders improved following the cessation of drugs or their dose adjustment.

• Familiarity with the phenomenology of motor adverse events in respect to specific ASMs facilitates their early recognition and triggers targeted interventions.