Keywords:
Chorea - Dyskinesia - Stroke - Emergencies - Movement Disorders
Palavras-chave:
Coreia - Discinesias - Acidente Vascular Cerebral - Emergências - Transtornos dos
Movimentos
INTRODUCTION
Chorea is derived from the Greek word choros, meaning dance. It is a hyperkinetic movement disorder characterized by a continuous
flow of random, brief and migratory involuntary muscle contractions. In the diagnostic
workup of chorea, at least ask four questions should be asked: 1) What was the patient’s
age at onset? 2) What was the time course of chorea? 3) Was the patient exposed to
drugs capable of inducing chorea? 4) Is there a positive family history for chorea?[1]
According to the temporal pattern, chorea is defined as acute when present within
hours to days. Unlike chronic chorea, in which heredodegenerative causes such as Huntington's
disease stand out, three main nosologies have major roles in acute chorea: vascular,
metabolic and inflammatory[2],[3].
In the context of acute or subacute chorea, hemiballism-hemichorea is usually the
most common presentation. Ballism is a term derived from the Greek word ballismus, which means “jumping about or dancing”. This is generally considered to be a very
severe variant of chorea characterized by violent, large-amplitude flinging movements
involving proximal extremities on one side of the body. Due to its often-related acute
onset, this presentation is the one most frequently seen in the emergency room[4],[5]. Chorea or ballism are usually displayed in the same patient during the course of
the disease; hemiballism is more prominent earlier, and lower-amplitude hemichorea
emerges as the disorder evolves. Commonly, the terms are used interchangeably.
Abnormal movements develop as complications in 1-4% of all patients after stroke[6]. Although hemichorea is considered to be a rare complication reported in less than
1% of cases, with or without hemiballism, it is the most frequent hyperkinetic post-stroke
movement disorder, followed by dystonia. In addition, hemichorea-hemiballism may be
part of or even the only sign of acute presentation of chorea[6]. On the other hand, a previous case series on sporadic chorea in adults showed that
stroke is the main cause of acute chorea (approximately 50%)[7].
Metabolic nosology often appears as the second main etiology, in up to 30% of cases[1],[3]. Drugs, hyperglycemia, hyponatremia and hypoxemia have all been listed as potential
causes. The possibly remaining etiologies are usually inflammatory, whether autoimmune
or infectious (e.g. Sydenham or HIV-AIDS). Regarding case series of children with
acute chorea, up to 95% present Sydenham chorea, such that inflammatory etiology is
the most frequent cause in this specific population[1]. Other immunological etiologies resulting in chorea include systemic lupus erythematosus,
primary antiphospholipid antibody syndrome, vasculitis and paraneoplastic syndromes.
Considering that acute chorea or hemiballism-hemichorea syndromes are uncommon entities
that are found in the context of neurological emergencies, we conducted a retrospective
study to investigate acute chorea cases evaluated over the past four years.
Regarding vascular causes, we described the vascular regions more commonly affected
in cases of acute chorea. Concerning metabolic causes, we investigated which were
the major underlying etiologies. From this, we put forward suggestions regarding the
findings from clinical features that potentially would be capable of distinguishing
between vascular and nonvascular (metabolic or inflammatory) etiologies, in order
to guide approaches towards acute chorea syndromes and their management.
Hence, the main purpose of this study was to investigate which were the main etiologies
and clinical features of acute chorea in the emergency department of a tertiary-level
center. Additionally, we suggested an approach for optimizing the diagnostic workup
and for the initial management of this uncommon condition in the emergency room.
METHODS
This was a retrospective observational study. We reviewed the electronic medical records
of patients seen at the neurological emergency room of Hospital das Clínicas, University
of São Paulo School of Medicine, from 2015 to 2019. We searched for the term "chorea"
in summarized patient data. When records were found, we accessed the full medical
record of each patient from the time of the emergency room admission. A total of 10
patients were included in the definition of acute chorea or hemiballism-hemichorea,
based on clinical examination and according to the medical records. The data were
then grouped into three main causes: vascular, toxic-metabolic and inflammatory.
In each case, we characterized the age at onset, gender, body distribution (hemichorea
versus generalized chorea), duration of chorea at the time of hospital admission,
neuroimaging characteristics and final etiology registered in the medical records.
For acute chorea related to stroke, we also characterized the vascular regions.
All the analyses were performed using the Statistical Package for the Social Sciences
software, version 21.0 (SPSS, IBM Statistics, Chicago, IL, USA). Categorical variables
were expressed as absolute and relative frequencies and compared using Pearson's chi-square
in univariate analyses. All tests were two-tailed. Statistical significance for final
values was taken to be p<0.05.
RESULTS
In our series, data from 10 patients presenting with acute chorea in the emergency
room were retrospectively evaluated. The investigations performed and described in
the medical records were generally based on neuroimaging and laboratory tests. The
patients’ median age was 59 (± 18.38) years. The majority were female (70%) and the
median duration of symptoms was six (±5.63) days. Regarding the etiology of the 10
patients with acute chorea, five cases (50%) presented vascular etiology, four cases
(40%) had a metabolic cause and one case (10%) had an inflammatory cause ([Table 1]).
Table 1
Demography, body distribution, etiology, neuroimaging and duration of symptoms of
10 cases with acute chorea.
|
Case
|
Age
|
Gender
|
Distribution
|
Etiology
|
Neuroimaging
|
Duration of symptoms
|
|
1
|
20
|
F
|
Hemichorea
|
Vascular
|
Caudate ischemia
|
17 hours
|
|
2
|
59
|
M
|
Hemichorea
|
Vascular
|
Subthalamic nuclei ischemia
|
7 days
|
|
3
|
65
|
M
|
Hemichorea
|
Vascular
|
Parietal ischemia
|
1 day
|
|
4
|
86
|
F
|
Hemichorea
|
Vascular
|
Parietal ischemia
|
2 hours and 40 minutes
|
|
5
|
59
|
F
|
Hemichorea
|
Vascular
|
Insular ischemia
|
1 hour
|
|
6
|
58
|
F
|
Hemichorea
|
Metabolic (diabetes)
|
Basal ganglia hyperdensity
|
14 days
|
|
7
|
61
|
F
|
Generalized
|
Metabolic (lithium)
|
Normal
|
5 days
|
|
8
|
49
|
M
|
Generalized
|
Metabolic (SSRI)
|
Normal
|
7 days
|
|
9
|
69
|
F
|
Hemichorea
|
Metabolic (diabetes)
|
Basal ganglia hyperdensity
|
14 days
|
|
10
|
27
|
F
|
Generalized
|
Inflammatory (Sydenham)
|
Normal
|
14 days
|
We compared age at onset, body distribution and median duration of symptoms in two
groups: vascular and nonvascular causes. There were no statistical differences regarding
age at onset, in a direct comparison between vascular and nonvascular causes (median
age 59±21.34 vs 58±14.4 years). All vascular causes presented as hemichorea, whereas
the most common body distribution in nonvascular etiologies was generalized (60%).
To compare the etiologies concerning presentation timing, we divided the acute chorea
sample into two groups: from the beginning until 24 hours and more than 24 hours.
Interestingly, the group with presentation of less than 24 hours was statistically
related to the vascular etiology (p=0.04). It was noteworthy that the median duration
of symptoms for vascular causes was 17 (±63.12) hours, in contrast to 14 (±3.97) days
for nonvascular causes.
Considering the vascular-related cases, all of these five patients had asymmetrical
presentation and ischemic strokes. Only one patient in this group (case 2 in [Table 1]) showed ischemia in the contralateral subthalamic nuclei (which is classically associated
with hemichorea-hemiballism). Other regions of occurrence that were found included
the caudate, insula and parietal lobe ([Figure 1]). Patient 5, a 59-year-old woman, developed hyperacute left hemichorea-hemiballism
and was admitted within 60 minutes from the clinical onset, in our emergency department.
Her head CT scan and blood glucose were normal. She underwent thrombolysis with intravenous
alteplase, from which she evolved with partial improvement after 2 hours and complete
resolution after 24 hours, without any other treatments. Brain MRI demonstrated acute
right insular stroke. This remarkable example of a dramatic response emphasizes the
importance of considering chorea in stroke protocols when there is an ictal onset,
and of considering it to be a potentially treatable entity[8].
Figure 1 Vascular chorea: although classically described in the subthalamic nuclei, our cases
showed ischemic lesions in other parts of the basal ganglia network: caudate nuclei,
globus pallidus and parietal and insular cortex.
Regarding toxic-metabolic etiology, there were two cases of hemichorea-hemiballism
secondary to decompensated diabetes with a nonketotic hyperglycemic state. In addition
to high serum glucose levels, characteristic imaging signs of striatal hyperdensity
on CT scan or hyperintensity on T1-weighted MRI in the basal ganglia aided the diagnosis.
Both cases presented asymmetrical hemichorea syndromes and had benign evolution, with
remission in which it took from days to weeks for the symptoms to improve after glucose
control had been achieved.
Another two cases were defined as drug-induced. A 61-year-old patient with generalized
chorea and encephalopathy, with laboratory results showing a lithium level of 3.17
mmol/L (therapeutic range 0.5-0.8 mmol/L), underwent hemodialysis and subsequently
became more alert, and the chorea subsided. Patient 8, a 49-year-old male, developed
generalized chorea after introduction of a selective serotonin reuptake inhibitor
(SSRI), fluoxetine, at a standard dose. Further laboratory and imaging investigations
were unremarkable and no other potential factor was identified. The medication was
discontinued and chorea improved after SSRI withdrawal.
One case in the sample had an inflammatory cause. This comprised probable reactivation
of Sydenham chorea in a 27-year-old female patient. She presented involuntary movements
days after a dental procedure, with phenomenology similar to what had been seen when
she was previously diagnosed with Sydenham chorea in childhood. After ruling out other
causes of chorea through her history, neuroimaging and laboratory testing, we considered
that this case probably represented reactivation of Sydenham chorea.
Neuroimaging was useful in all vascular cases and in metabolic chorea associated with
nonketotic hyperglycemia. There were no typical relevant abnormalities presented in
neuroimaging, concerning drug-related and inflammatory causes (Sydenham reactivation),
in contrast to the vascular etiologies and hyperglycemic states (brain MRI was performed
on all our cases). The most useful laboratory tests in our sample were seen to be
the serum glucose and drug serum levels (e.g. lithium).
DISCUSSION
In our case series, acute chorea was more commonly found in middle-aged adults, probably
due to the main vascular etiology. Although our sample was of limited size, this finding
is in line with a previous Italian case series in which 51 patients with acute chorea
were described[7]. On the other hand, a previous Brazilian case series demonstrated a younger mean
age[9].
Vascular etiology was the main cause of acute chorea in our patients. This result
is consistent with two other previous case series[7],[9]. Ischemic stroke occurred in all the vascular cases, as was seen in these previous
studies, in which all the cases were due to ischemic stroke except for one in which
acute chorea was developed after a left caudate hematoma[7].
Although occurrences of chorea are classically associated with subthalamic nuclei
infarction, previous studies have shown that neuroimaging findings in other region
are the rule and not the exception ([Figure 1])[7],[9]. Chorea should be recognized as caused by dysfunctional neuronal networks connecting
the basal ganglia and motor cortical areas, and not due to a single lesion related
to a specific brain topography (e.g. subthalamic nuclei). One possible network with
a considerable role is the posterolateral putamen pathway, which was shown to be damaged
in 90% of the patients in a hemichorea-hemiballismus case series with 29 patients[10]. The sites involved included the caudate nuclei, insular cortex and parietal lobe,
regions that were also identified in our case series ([Figure 1]).
Toxic-metabolic causes were the second most frequent etiology. Nonketotic hyperglycemia
was the main etiology in this group. Similar findings were reported in a previous
case series[7],[9]. It has been acknowledged that, although metabolic etiologies are often associated
with diffuse brain lesions and generalized chorea, the distribution in nonketotic
hyperglycemia is usually unilateral[1],[11]. This distribution possibly makes nonketotic hyperglycemia the main differential
diagnosis of vascular chorea. One reasonable previously mentioned explanation is that
hyperglycemia leads to multifocal asymmetrical petechial hemorrhages of the basal
ganglia, which is highlighted by the typical asymmetric hyperdense basal ganglia lesions
in brain CT, which correlate with SWI-positive findings in brain MRI[1]. This imaging pattern was found in both our cases, thus aiding in the differential
diagnosis, and is in contrast with a previous case series in which it was only seen
in three out of six cases[12].
We did not find any cases of the following causes of metabolic chorea that have previously
been described: acute intermittent porphyria, hypo/hypernatremia, hypocalcemia, hyperthyroidism,
hypoparathyroidism or hepatic/renal failure.
Additionally, drug-related chorea cases presented as generalized chorea syndrome.
Several centrally-acting drugs, such as as neuroleptics, metoclopramide, lithium,
SSRI and cocaine, have previously been identified as possible causes of medication-induced
chorea[7],[13],[14]. One previous case report describing lithium toxicity leading to sporadic chorea
mentioned risk factors such as advanced age, use of concomitant medications and impaired
kidney function[14]. Other less common substances that have been reported as causes of drug-induced
chorea include carbon monoxide, manganese, mercury and organophosphate poisoning[1].
In our case series, only one case had an inflammatory cause, consisting of probable
Sydenham chorea reactivation. We assume that the reason for this was that the most
common inflammatory chorea, Sydenham chorea, is typically a pediatric disease and
our service is an adult-only emergency department.
Furthermore, differently from the previous case series, we did not see any HIV-related
chorea, which has been mentioned as the most common infectious cause[7]. One reasonable explanation for this could be the lower incidence of uncontrolled
HIV-AIDS nowadays. It is worth mentioning that we also did not find any other causes
of inflammatory chorea that have been described, such as chorea gravidarum, antiphospholipid
antibody syndrome, postinfectious or postvaccinal encephalitis. paraneoplastic chorea,
neurosyphilis or viral encephalitis (mumps, measles and varicella), mainly due to
our small sample. A longer period for an observational study would probably give rise
to some of these etiologies.
Taking into account our findings, we developed the DANCE investigational acronym to
suggest some first steps in the approach towards patients presenting with acute chorea
in the emergency room. This is a mnemonic that represents the following: Diagnosis
of chorea, Acute stroke protocol, Normal glucose levels, Check neuroimaging and Exposure
to drugs (DANCE). Its aim is to enable identification of common etiologies that require
urgent management, as shown in [Table 2]. We recommend that acute stroke protocol evaluation should be performed even in
patients with diabetes and elevated serum glucose at admission, because these patients
also present increased risk of stroke, and thrombolysis is a time-limited form of
therapy. Brain MRI imaging is often able to help differentiate stroke and hyperglycemia
as the cause of acute chorea.
Table 2
The DANCE approach. The five critical points in patients with acute chorea are the
following: identification of a chorea syndrome; use of an acute stroke protocol; presence
of normal serum glucose; checking of neuroimaging; and checking of exposure to drugs.
|
Diagnosis of chorea
|
Involuntary migratory, random, chaotic muscle contractions
|
|
Acute stroke protocol
|
Thrombolysis can be as adequate for acute hemichorea as it is for acute hemiparesis
|
|
Normal glucose protocol
|
Nonketotic hyperglycemia is a major cause of acute chorea
|
|
Check neuroimaging
|
Order brain imaging (consider MRI if CT is normal)
|
|
Exposure to drugs
|
Stop use of centrally-acting drugs as neuroleptics, lithium or SSRI
|
As a retrospective case series, there were some limitations regarding collection of
data from non-standardized medical reports. Two patients were excluded prior to the
analyses due to incomplete descriptions of the etiology investigation. Furthermore,
the low volume of data prevented us from obtained more robust statistical relevance.
Nonetheless, we consider that one strength of our study was that it illustrates a
case series of a rare movement disorder presentation in the emergency department.
In addition, we proposed a diagnostic workup based on our clinical observations, within
the acute context of generalized chorea or hemiballism-hemichorea syndromes. Therefore,
considering the emerging developments regarding stroke protocols and neuroimaging,
we are sure that a straightforward approach might help in making prompt diagnoses
and in managing this potentially treatable entity.
Acute chorea is an uncommon problem in the emergency room that is associated with
potentially treatable causes. The main etiology is vascular and, thus, it should be
considered in stroke protocols aiming towards thrombolysis. The initial investigation
also should focus on serum glucose levels, given that nonketotic hyperglycemic states
are other frequent cause. Ascertaining the patient’s drug history is paramount, in
order to rule out centrally-acting drugs as an etiology relating to chorea (e.g. lithium).
Neuroimaging can help in the differential diagnosis between stroke and nonketotic
hyperglycemia. We proposed the DANCE approach ([Table 2]) to guide the diagnostic workup and to emphasize the first steps in evaluating patients
with acute chorea, in order to focus on causes that require prompt management, such
as thrombolysis, hemodialysis or intensive glucose control.
