Original Articles

Alternating Hemiplegia Syndrome:
Electroencephalogram, Brain
Mapping, and Brain Perfusion
SPECT Scan Study
in a Chinese Girl
Virginia Wong, MBBS, MRCP(UK), DCH(London), DCH(Glasgow); Gerrit C.L. Ho,
MBBS, MSc; Henry W.D. Yeung, MBBS, FRCR(UK), ABIM; K.M. Ma, MBBS, MRCP
Abstract
A 3-year-old Chinese girl with alternating hemiplegia syndrome failed to respond to anticonvulsants, antimigrainous
drugs, and calcium channel blockers. She made a complete remission with a 4-week course of steroid, and relapsed after
steroid withdrawal. Electroencephalogram and brain mapping during the hemiplegic attack showed unilateral highvoltage sharp slow-wave discharges in the temporo-occipital region contralateral to the hemiplegic side and diffuse
high-voltage slowing during attacks of quadriplegia or other clinical manifestation such as dullness, lethargy, or
yawning. Brain perfusion single photon emission computed tomographic (SPECT) scan study during the attack showed
decreased uptake in the temporoparietal region contralateral to the hemiplegic side and in the ipsilateral basal ganglia,
whereas the perfusion was normal between attacks. Electroencephalogram background activity was improved while the
child was in clinical remission with steroid treatment. Computed tomographic and magnetic resonance imaging scans of
the brain were normal. Carotid angiogram failed to show any structural or dynamic changes of the carotid arteries. The
possible mechanism underlying alternating hemiplegia syndrome might be transient and reversible cerebral ischemia
with high-voltage slow-wave discharges shown in the electroencephalogram and decreased perfusion in SPECT scan. (J

Child Neurol 1993;8:221-226).

syndrome
Alternating hemiplegia hemiplegia
occurring

consists of

tal or neurologic abnormalities.’ The exact mechanism causing the hemiplegic attacks is not known. Pos-

on alternating sides associated with oculomotor abnormalities and other paroxysmal phenomena such
as dystonia, choreoathetosis, or autonomic disturbances. These children usually have associated men-

sible relationship to migrainous or epileptic phenomhas been postulated. 1-6 The overall prognosis
is guarded, because trials with anticonvulsants and
antimigrainous drugs have failed to abort the at-

recurrent attacks of

Received Feb 25, 1992. Received revised May 20, 1992. Accepted for publication June 11, 1992.
From the Department of Paediatrics (Dr Wong), University
of Hong Kong, Queen Mary Hospital and the Department of Nuclear Medicine (Drs Ho, Yeung, and Ma), Queen Mary Hospital,

Hong Kong.

Address correspondence to Dr Virginia Wong, Department
of Paediatrics, University of Hong Kong, Queen Mary Hospital,
Pokfulam Road, Hong Kong.

ena

tacks.1,2,5
We report changes in the electroencephalogram
(EEG), brain mapping and brain perfusion single

photon emission computed tomographic (SPECT)
scan in a 3-year-old Chinese girl with intractable
episodes of alternating hemiplegia or quadriplegia.
She failed to respond to anticonvulsants, antimigrainous drugs, and calcium channel blockers but
221

Downloaded from jcn.sagepub.com at FLORIDA INTERNATIONAL UNIV on July 7, 2015

made a complete remission with a short course of
steroid treatment and then relapsed on its withdrawal.
z

Case Report
L.W.M. was a 3-year-old Chinese girl born by cesarean section for breech presentation with a birth weight of 3.1 kg.
Antenatally, the mother noted decreased fetal movement

all through her pregnancy. There was no perinatal asphyxia. She was noted to be hypotonic from birth.
At the age of 2 months, she was noted to have deviation of eyes and head to her left, with loss of consciousness
for about 30 minutes. There was no clonic twitching of the
limbs. She remained drowsy after the attack. At 5 months
of age, she had staring followed by deviation of eyes and
head to her left, with clonic twitching of all four limbs and
associated loss of consciousness for 5 minutes. She was admitted into Queen Mary Hospital, the University of Hong
Kong teaching hospital. Thereafter, she was noted to have
repeated attacks of up-rolling of eyes and deviation of eyes
or head turning to her left or right side. EEG did not show
any focal abnormalities. She was started on carbamaze-

pine.
At the age of 71/2months, she was noted to have deviation of eyes to the left with sudden loss of movement of
her left upper and lower limbs and a blank expression lasting for 36 hours. With the return of movement on her left
side, she was noted to have decreased movement of her
right upper and lower limbs lasting for 24 to 36 hours. She
was fully conscious during the attack.
Thereafter, she had recurrent attacks of alternating
hemiparesis/hemiplegia and quadriparesis/quadriplegia.
The frequency varied from one attack per week to 30 attacks per day. Each attack lasted from 10 minutes to 48
hours. Preceding the attack, she was usually dull and tired
looking. Her face sometimes turned pale, and circumoral
cyanosis sometimes appeared. Sometimes, dilated veins
were seen on one side of the forehead unrelated to the
hemiparetic side. She was very miserable and irritable during the attack and cried a lot. The hemiparesis alternated
on both sides repeatedly or remained on the same side
during the attack. She was fully conscious during the attack, and she could still shake her head to indicate that she
could not move her limbs. Her speech was slurred with a
low pitch quality. Occasionally, she was aphasic, and tears
sometimes ran out continuously from one eye, usually on
the same side as the hemiparesis. Her muscle tone was extremely hypotonic during the attacks, but fluctuation to
slight rigidity occasionally occurred. Her big toes were
sometimes dorsiflexed with a clustering together of the
other toes. There was sometimes associated coarse tremor
of the hemiparetic or normal hand. Her pupils were sometimes relatively dilated. Tachycardia and, sometimes, decreased respiratory movement occurred during the attack.
After the attack, she was usually tired and slept for a long
time. She usually woke up feeling refreshed, with disappearance of the hemiparesis. Her attacks may have been
precipitated by stress but were unrelated to sleeping pattern.

Her developmental milestones were delayed. She
achieved head control at 4 months of age, sat with support

at age 10 months, and walked with support at 21/2 years of
age. She started to say single words at age 2 years. Her
mental age was in the mild mental retardation range.
The mother had a history of recurrent frontal headache since her youth, but she had no symptoms suggestive

of migraine. There was no family history of periodic paralysis. Physical examination of the child showed generalized
hypotonia with no muscle weakness between attacks. During attacks, she had generalized hypotonia and weakness
(unilateral or generalized) and brisk deep tendon reflexes,
but negative Babinski reflexes.
The following investigations were found to be normal:
complete blood count; erythrocyte sedimentation rate; renal, liver, and thyroid function tests; and serum levels of potassium, magnesium, glucose, ammonia, creatine kinase,
lactate, and pyruvate. Tests for antinuclear factor, rheumatoid factor, and anti-DNA antibodies were negative, as were
anticardiolipin antibody, lupus anticoagulant, and antineutrophil cytoplasmic antibody examinations. Urine analysis for amino acids, organic acids, and metabolic screening
was negative. Nerve conduction study and electromyogram
results were normal. Muscle biopsy showed no evidence of
mitochondrial myopathy. Computed tomographic and magnetic resonance imaging scans of the brain and brain-stem
auditory evoked potential were normal. Carotid angiogram
failed to show any stenosis of the carotid arteries. Echocardiogram and Holter monitoring were normal.
Prolonged video-EEG monitoring and brain mapping
with a 16-channel Concerto machine (Dantec) showed that
the clinical attacks were not associated with any spike
wave discharges. Diffuse high-voltage slow delta activities
occurred during attacks of dullness, lethargy, loss of
speech, yawning, and tearing without clinical evidence of
hemiplegia (Figure 1). Bursts of generalized high-voltage
slow delta waves occurred for a prolonged period of time
during the sleeping stage after an attack. During a hemiplegic attack, slowing of the background activity with highvoltage slow delta or theta waves occurred in both
temporo-occipital regions or sometimes contralateral to the
hemiplegic side (Figure 2). Photic stimulation induced
asymmetrical photic recruitment response, with absent response ipsilateral or contralateral to the hemiplegic side
during the attack.
Three sets of 99mTc-hexamethylpropylenamine oxide
(HMPAO) SPECT scans were performed between attacks,
in the left hemiplegic state, and in the right hemiplegic
state. Commercially supplied HMPAO (Amersham Co)
brain agent was freshly labeled with 25 to 30 mCi 99mTc in 5
mL eluate. In the baseline state between attacks (upper
row, Figure 3), the uptake of 99&dquo;’Tc-HMPAO was normal
and symmetrical in the cerebral hemispheres, basal ganglia, and cerebellum. However, during the hemiplegic attack, there was decreased uptake detected in the cerebral
hemisphere contralateral to the side of hemiplegia. In the
case of left hemiplegic attack (middle row, Figure 3), hypoperfusion was detected in the right frontal, parietal, and
occipital regions and the left basal ganglia. During an attack of right hemiparesis, the child had dystonic posturing
of the right upper limb with mild tremor and the paralysis
was only mild (lower row, Figure 3). Hypoperfusion during this attack was more subtle because it was detected
only in a small segment of the left anterior parietal lobe
and in the right basal ganglia. The uptake in the cerebel-

222

Downloaded from jcn.sagepub.com at FLORIDA INTERNATIONAL UNIV on July 7, 2015

FIGURE 1
EEG and brain mapping during an attack of dullness, irritability, prolonged crying, and
slurring of speech, with no hemiplegia. The red to blue areas are indicative of regions of
high to low spectral energy, respectively, in different frequency bands (delta, theta,
alpha, and beta). Diffuse high-voltage slow delta waves (upper left) and bilateral
temporo-occipital theta waves (upper right) are seen in the brain mapping study (vertex

view) .

FIGURE 2
EEG and brain mapping during an attack of left hemiplegia showing diffuse highvoltage slow delta wave discharges (upper left) and increased theta activities over right
temporocentral and left temporoparietal regions.

223

Downloaded from jcn.sagepub.com at FLORIDA INTERNATIONAL UNIV on July 7, 2015

FIGURE 3

99mTc-HMP AO SPECT scan showing four coronal cuts of the three sets of imaging: (1)

between attacks (top row), (2) left hemiplegia (middle row), and (3) right hemiplegia
(bottom row). The yellow to blue areas are indicative of high to low 99mTc-HMP AO up-

take, respectively.
lum was symmetrical during the three clinical states and
was therefore used as the maximal reference
intensity for

qualitative interpretation.
Our patient was administered sodium valproate, carbamazepine, phenytoin, and propranolol, with no improvement. A short trial of flunarizine for 2 months failed
to abort the attacks. Because of the possibility that arteritis

causing vasospasm, betamethasone, 4 mg twice daily
(the equivalent of 2 mg/kg prednisone per day), was tried.
The attacks were completely aborted 2 days after administration, associated with marked improvement in EEG backwas

ground activity. She could walk around steadily, and her
mood was improved. Due to the appearance of marked
cushingoid features, betamethasone was stopped at the request of the parents after a 4-week course. This was associated with immediate relapse of the attacks. Because the
parents refused to retry steroid due to the occurrence of
marked cushingoid features, we could not substantiate its
continued therapeutic usefulness.
,

Discussion
Our case, with typical features of alternating hemiplegia syndrome, showed poor response to anticonvulsants, antimigrainous drugs, and calcium channel
blockers but made a complete remission with a 4-week
course of steroid and relapsed after its withdrawal.
Alternating hemiplegia syndrome was first described in eight children with intermittent alternating

hemiplegia associated with vasomotor prodrome.2
Periodic cerebral vascular ischemia related to migraine was postulated as the basis of the symptoms.
Symptoms of migraine occurred in some of these children on follow-up. Subsequent reports’ -6 of alternating hemiplegia syndrome showed controversy in
differentiating the underlying mechanism for the attacks, whether related to seizure or migraine. &dquo;Ictal&dquo;
EEG in some cases showed slow-wave abnormalities
or seizure discharges. The response to anticonvulsants was notably poor.

In 1984, Caesar and AzOU7 reported dramatic improvement in alternating hemiplegia syndrome with
flunarizine, a calcium channel blocker. A subsequent
international trial with 12 children’ showed the use-

fulness of flunarizine in aborting the hemiplegic attacks. The mechanism of action might be related to
the antihypoxic, antivasoconstrictive, and anticonvulsive activity of flunarizine.9 Other studies also
substantiated the usefulness of flunarizine in aborting the attacks.lO,l1 The response of alternating
hemiplegia syndrome to flunarizine was more rapid
than the usual gradual improvement of migrainous
symptoms in children and adults.9-11 However, our
patient failed to respond to a trial of 2 months’ treatment with 5 mg flunarizine daily.

224

Downloaded from jcn.sagepub.com at FLORIDA INTERNATIONAL UNIV on July 7, 2015

EEG

findings of high-voltage slow-wave dis-

charges in the contralateral and/or bilateral temporo-

occipital regions during the prodrome or hemiplegic
attacks in our case suggest decreased blood flow as
the possible mechanism of hemiplegia and the associated autonomic manifestation. We failed to demonstrate any anatomic or dynamic occlusion of the
carotid arteries with the angiogram. A short, 4-week
of betamethasone was tried for possible arteritis in causing recurrent cerebrovascular ischemia.
The dramatic response with total absence of hemicourse

plegic attacks and improvement of behavior, together with immediate deterioration with relapse of
hemiplegia, quadriplegia, and tremor at a lower
dose of betamethasone proved the clinical efficacy of
the steroid. Serial EEG recordings documented the
return to normal background activity, with loss of
high-voltage slow-wave activity. Four days after
stopping steroid treatment, the clinical attacks recurred and the EEG became abnormal again.
The usefulness of steroid in alternating hemiplegia syndrome has not been previously reported. The
clinical and EEG remission induced by steroid therapy, with relapse during the withdrawal period in
our case is interesting. Unfortunately, we were unable to document the continued usefulness of steroid in this disease, due to parental refusal of drug
continuation. Alternating hemiplegia syndrome is a
rare disease and further case-control study with steroid treatment is worthwhile. It is still too premature
to document the usefulness of steroid in this rare
condition because the therapeutic trial was short.
Recent reports 12-15 on cerebral perfusion during
and between attacks with the use of brain perfusion
SPECT scan showed conflicting results. Zupanc et
al 14 showed hypoperfusion in the left cerebral hemisphere during an episode of right hemiplegia in a
17-month-old girl, which resolved with clinical improvement. The authors postulated vasospasm the
probable cause of the transient reversible hypoperfusion of the contralateral hemisphere. EEG also
showed focal slowing over the hypoperfused hemisphere during the attack. This case was similar to
our child with focal hypoperfusion in SPECT scan
and focal slowing in EEG except that our case also
showed hypoperfusion in the ipsilateral basal ganglia as well.
However, Kanazawa et a115 showed ictal hyperperfusion in the hemisphere contralateral to the
hemiplegic side and symmetric uptake during the
normal period. Thus, the authors postulated alternating hemiplegia syndrome as an atypical manifestation of epilepsy. Their case was a 12-month-old

boy with possible cerebral palsy and mental retardation, and focal spike discharges were shown in the
interictal EEG. Thus, the SPECT scan study during
the attack with hyperfusion might be actually detecting an underlying seizure phenomenon. Whether
the hyperfusion was related to a genuine epileptic
attack instead of a vasospasm-induced hemiplegia is
unknown because children with alternating hemiplegia syndrome usually had epilepsy as well.
The tracers used were different in the two stud123I-Iodoamphetamine (IMP) was used in the
study by Zupanc et a114 with hypoperfusion,
whereas 99mTc-HMPAO was used in the study by
Kanazawa et al with hyperperfusion.I5 Because both
the Kanazawa group and our study used the same
perfusion agent but yielded contradictory results, it
is unlikely that the radiopharmaceutical biochemistry or kinetics resulted in the differences. This can
be further supported by the fact that despite the different uptake kinetics and intracerebral binding stability of 99’T’Tc-HMPAO and 123I_IMP, we have
obtained similar hypoperfusion results to those of
Zupanc’s group. Furthermore, the Kanazawa group
reported only slight, relative hyperfusion without
EEG paroxysmal discharges during an attack, which
is again different from our findings.
The results of our SPECT scan study using
99mTc-HMPAO are more in line with the finding by
Zupanc et al of decreased cerebral perfusion during
an attack. Strangely enough, in our case, the clinical
attack of left hemiplegia was associated with a markedly decreased perfusion in the right temporoparietal region, whereas the mild right hemiparesis with
dystonia of right upper limb and tremor of right
hand was associated with only slightly decreased
perfusion in the left anterior parietal area. There
seemed to be a good correlation between the degree
of clinical attack and the degree and extent of hypoperfusion in our case. It is interesting to note that
there was mildly decreased uptake shown in the
basal ganglia ipsilateral to the hemiplegic side during the attacks. Whether this might be due to the results of the remote interruptive effects of transhemispheric diaschisisI6 is still to be investigated. This
is similar to positron emission tomographic scan finding in alternating hemiplegia syndrome as reported
by Tada et al 17 in an 11-year-old boy who showed a
slight decrease in cerebral blood flow at the insula,
putamen, and claustrium of the left side during an attack of right hemiplegia.
Thus, we postulate that transient vasospasm is
the underlying mechanism for the recurrent hemiplegic attacks due to EEG, brain mapping, and brain
ies.

225

Downloaded from jcn.sagepub.com at FLORIDA INTERNATIONAL UNIV on July 7, 2015

perfusion SPECT scan changes seen during and between the attacks. Alternating hemiplegia syndrome
might be an intermittent paroxysmal movement disorder rather than being epileptic in nature. The
mechanism underlying the dramatic response to steroid and relapse after its withdrawal is unknown.
The decreased blood supply to the middle cerebral
artery with resultant decreased blood flow to the
frontal, parietal, and temporal lobes could account
for the hemiplegia, quadriplegia, and aphasia. Decreased blood supply to the posterior cerebral artery
with resultant ischemia to the basal ganglia could account for the extrapyramidal manifestation of dystonia, rigidity, choreoathetosis, and tremor of the
limbs. Further study of this disease is warranted, especially using brain perfusion SPECT scan studies in
relation to steroid treatment, in order to elucidate
the basic mechanism underlying this rare disorder.

Acknowledgment
The authors would like to thank Mrs T. Wong for her secretarial assistance and Mr J. Ng for technical support.

References
1.

dren. Pediatrics 1975;56:722-726.
5. Hosking GP, Cavanagh NPC, Wilson J: Alternating hemiplegia : Complicated migraine of infancy. Arch Dis Child

1978;53:656-659.
6. Shirasaka Y, Ito M, Okuno T, et al: Epileptic seizures difficult
to differentiate from alternating hemiplegia in infants: A case
report. Brain Dev 1990;12:521-524.
7. Caesar P, Azou M: Flunarizine in alternating hemiplegia in
childhood. Lancet 1984;2:579.
8. Casaer P, Aicardi J, Curatolo P, et al: Flunarizine in alternating hemiplegia in childhood. An international study in 12
children. Neuropediatrics 1987;18:191-195.
9. Curatolo P, Cusmai R: Letter to the editor. Lancet 1984;2:980.
10. Caers LI, De Beukelaar FD, Amery WK: Flunarizine, a calciumentry blocker, in childhood migraine, epilepsy and alternating
1987;10:162-168.
hemiplegia. Clin Neuropharmacol
11. Campistol PJ, Sans Fito A, Pineda Marfa M, et al: Alternating
hemiplegia in infancy; Clinical features, clinical course and
treatment based on three cases. An Esp Pediatr 1990;32:
336-338.
12. Siemes H, Caesar P: Alternating hemiplegia in childhood.
Clinical report and single photon emission computed tomography study. Monatsschr Kinderhielkd 1988;136:467-470.
13. Hattori H, Hashizuka S, Matsuoka O, et al: Alternating hemiplegia in infants: A case report with abnormal findings in
ABR (auditory brainstem responses) and SPECT (single photon emission CT). Jpn J Pediatr 1989;42:77-82.
14. Zupanc M, Dobkin JA, Perlman SB: 123
I-Iodoamphetamine
SPECT brain imaging in alternating hemiplegia. Pediatr Neurol

1991;7:35-38.

Krageloh I, Aicardi J: Alternating hemiplegia in infants: Re-

1980;21:784-791.
port of five cases. Dev Med Child Neurol
2. Clarke JM: On recurrent motor paralysis in migraine (with report of a family in which recurrent hemiplegia accompanied
the attacks). Br Med J 1910;1534-1538.
3. Dynes JB: Alternating hemiparetic migraine syndrome. Br

MedJ
1939;2:446-447.

4. Golden GS, French JH: Basilar artery migraine in young chil-

15. Kanazawa O, Yoshiyuki S, Hattori H, et al: Ictal TC99m
HMPAO SPECT in alternating hemiplegia. Pediatr Neurol

1991;7:121-124.
16. Andrew RJ: Transhemispheric diaschisis, a review and comment. Stroke 1991;22:943-949.

17. Tada H, Miyake S, Yamada M, et al: A patient with alternating hemiplegia in childhood. No To Hattatsu 1989;21:283-288.

226

Downloaded from jcn.sagepub.com at FLORIDA INTERNATIONAL UNIV on July 7, 2015