Archives ofDisease in Childhood 1990; 65: 1245-1248

1245

Central nervous system vasculitis after
chickenpox-cause or coincidence?
Avinoam Shuper, Eileen P G Vining, John M Freeman

Abstract
A 7-5 year old boy, known to have a seizure
disorder, presented with an infarct in the left
middle cerebral artery territory, 10 weeks
after severe chickenpox. Immunofluorescent
antibody titre to the variceila zoster virus in
the cerebrospinal fluid was 1:32. Cerebral
angiography showed evidence of focal vasculitis. He presented again seven months later
with an acute exacerbation of seizures.
Magnetic resonance imaging of the brain
showed an old posterior extension of the
infarct, but a repeated angiography demonstrated an improvement in the vasculitic process. Cerebrospinal fluid antibody titre was
again 1:32. Although this may have been an
unfortunate coincidence, a possible association between chickenpox and vasculitis, similar to that reported with herpes zoster, and
with potentially significant clinical implications, should be considered. As a definite
proof can be obtained only by a brain biopsy,
however, which is generally not indicated in
such cases, only additional clinical reports
can lead to delineation of this association as a
definite entity.
Nervous system complications of varicella are
not rare in general paediatric practice. A diver-

Department of
Neurology,
Division of
Pediatric Neurology,
The Johns Hopkins
Medical Institution,
600 N Wolfe Street,
Baltimore,

MD 21205,
USA
Avinoam Shuper
Eileen P G Vining
John M Freeman
Correspondence to:
Dr Shuper.
Accepted 25 June 1990

sity of neurological conditions, including
meningoencephalitis, acute cerebellar ataxia,
and the Guillain-Barre and Reye's syndromes
have been reported.' The varicella zoster virus
may cause cerebral angiitis after herpes zoster
ophthalmicus,2 and angiitis in adults with and
without immune deficiencies after other herpes
zoster infections not involving the trigeminal
nerve.5 The association of vascular brain
lesions after chickenpox, although well described,'9 is much less well recognised. We
describe a child with cerebral vasculitis resulting in acute aphasia and hemiplegia 10 weeks
after the onset of chickenpox. The dilemmas
resulting from this possible association are discussed.
Case report
A 7'5 year old boy followed up for a longstanding seizure disorder presented with acutely progressive right sided weakness and aphasia. He
had weighed 2682 g at birth after an unremarkable pregnancy and delivery. His development
thereafter was delayed, with a stable developmental quotient of around 70. At age 3 years he
developed a mixed seizure disorder with generalised tonic-clonic seizures, partial right sided

motor seizures, and drooling spells. His neurodevelopmental problems were thoroughly
evaluated in another hospital, with negative
results. Of note, plasma and urine amino acid
pattern and magnetic resonance imaging of the
brain, which were done at the age of 5 5 years,
were normal. Physical and neurological examinations were unremarkable. Seizures were well
controlled by the use of valproic acid, 1125 mg
daily.
Chickenpox developed 10 weeks before his
hemiparesis and was described as severe and
widespread, with no signs of central nervous
system involvement. The lesions crusted and
the illness resolved after about two weeks.
Twenty four hours before his admission he
was noted to have weakness in his right hand,
which progressed over hours to involve the arm,
after which right sided facial weakness with
drooling developed, followed by inability to
speak. The next day he had a right hemiplegia
and expressive and receptive aphasia.
Physical examination was remarkable for
numerous generalised healing chickenpox scars.
He had a profound expressive aphasia, weakness of the right lower half of his face, and
motor weakness (3 in a scale of 0 to 5) on the
right. The deep tendon reflexes were hyperactive and his toe was 'up going' on the same side.
Sensory examination was normal. His right arm
and leg were mildly tremulous.
On admission his packed cell volume fraction
was 0 35, white cell count 8 9x 109/1 with
differential count fractions of 0-08 band forms,
0 3 polymorphonuclears, 0 5 lymphocytes, 011
monocytes, and 0-01 atypical lymphocyte. The
sedimentation rate was 6 mm in the first hour.
Concentrations of serum electrolytes, results of
liver function tests and determinations of
anion gap, prothrombin time, and activated
partial thromboplastin time gave normal
results. Blood lactate concentration was 0-6
mmol/l (normal up to 3 0 mmol/l) and lactate
in the cerebrospinal fluid was 1 6 mmol/l.
Rheumatoid factor and anti-DNA antibodies
were negative, antinuclear antibodies were positive at a titre of 1:40 (not significant below
1:160). A lumbar puncture showed normal cerebrospinal fluid pressure, normal concentrations
of glucose and protein, and 1 mononuclear
white blood cell/ml and 2 red blood cells/ml.
The antivaricella immunofluorescent antibody
titre was 1:1024 in the peripheral blood and
1:32 in the cerebrospinal fluid.
Electrocardiography and echocardiography
were normal. Electroencephalography showed a
consistent focal slowing over the left central
parietal region, and recurrent bursts of bilateral

Shuper, Vining, Freeman

1246

spikes, which occasionally were more apparent
over the right hemisphere.
Magnetic resonance imaging of the head (fig
1) was compatible with acute left temporal and
parietal lobe infarct with an abnormal signal
intensity within the left basal ganglia, compatible with lacunar infarcts. Bilateral cerebral
arteriography showed long discontinuous segments of smooth, concentric narrowing, and
beading involving the left anterior and predominantly the middle cerebral arteries, and
some involvement of the left posterior cerebral
artery (fig 2). The beading angioarchitecture of

Figure I Magnetic resonance imaging of the brain showing
the acute infarct in the left temporal and parietal lobes, with
the involvement of the left basal ganglia.

the arteries was considered to be typical of
central nervous system vasculitis.
In view of the vasculitic appearance of the
blood vessels, steroid treatment was initiated,
first with intravenous methylprednisolone for
three days and then oral prednisone for a
period of two months. Over several days the
child showed slow but progressive clinical
improvement with some return of speech and
improvement of his right hemiparesis. Seizure
control was also noted to be excellent, leading to
a 40% decrease in the valproic acid dose three
months later. The blood concentration remained within the high therapeutic range.
Seven months after the acute stroke, the child
presented again with acute onset of a flurry of
myoclonic seizures and aphasia, and was
admitted to the hospital. At this time his packed
cell volume fraction was 0-35, white cell count
7-2x109/l, platelets 204x109/l, and the sedimentation rate was 6 mm/hour. Concentrations
of serum electrolytes, activities of liver
enzymes, and anion gap and coagulation studies
were normal. Blood lactate concentration was
1-6 mmol/l. Amino acid patterns in serum and
urine were normal. Rheumatoid factor, antiDNA antibodies and antinuclear antibodies
were negative, and C3 and C4 concentrations
were normal. Examination of the cerebrospinal
fluid showed normal concentrations of glucose
and protein, 2 mononuclear white blood cells
/ml and 121 red blood cells/ml. The antivaricella
immuofluorescent antibody titre was 1:1024 in
the peripheral blood and 1:32 in the cerebrospinal fluid.
Electroencephalography showed bilateral
multifocal polyspikes, and generalised slowing
of the background, which was more pronounced on the left, where the basic rhythm
was decreased at times to 1-3/sec. The record
suggested a diffuse encephalopathy. Magnetic

Figure 2 Cerebral angiography showing narrowing and beading of the left middle, anterior
and, to a lesser degree, the posterior cerebral arteries.

Figure 3 Magnetic resonance imaging of the brain showing
siknt posterior extmion of the infarct.

1247

Central nervous system vasculitis after chickenpox-cause or coincidence?

resonance imaging of the head showed an old
posterior extension of the previous infarct
(fig 3). Angiography was repeated, and showed
noticeable resolution of the vasculitic process,
except for a residual irregularity of the proximal
portion of the left posterior cerebral artery.
The seizures were controlled by increasing
the valproic acid to the previous dose and
adding phenytoin. When seizures were controlled, neurologic examination showed that the
child was in general back to his preadmission
state, with recovery of the aphasia. Drooling, as
well as some degree of increased mental dullness, persisted.
Discussion
This child presented with acute onset of right
hemiparesis, with changes on magnetic
resonance imaging compatible with acute cerebral infarction and angiographic evidence of
unilateral central nervous system vasculitis. The
lack of evidence for other causes of cerebral
infarcts, such as heart disease, mitochondrial
angiopathy, amino acid abnormalities, hypercoagulable state, or a collagen vascular disease,
together with absence of systemic disease
suggest the diagnosis of a primary central
nervous system vasculitis. The angiographic
pattern, with segmental narrowing and beading,
is compatible with granulomatous angiitis of the
nervous system, a rare disorder in the paediatric
age group.2
The association between another clinical
manifestation of varicella zoster virus-the
herpes zoster ophthalmicus-and delayed hemiplegia is well documented.2 Within several
weeks after the primary infection, patients with
herpes zoster ophthalmicus can develop a contralateral hemiplegia, as a result of focal cerebral
angiitis ipsilateral to the viral infection.
Although the roentgenographic appearance of
the affected vessels is similar to the cases of
granulomatous angiitis of the nervous system,
the two conditions differ clinically. In comparison with granulomatous angiitis of the nervous
system, the herpes zoster ophthalmicus associated hemiplegia is generally less severe, is not
associated with 'ence,phalopathic' symptoms,
and is more limited in extent, and more likely to
resolve.2
The mechanism for the vasculitis in herpes
zoster ophthalmicus associated hemiplegia is
unclear. Virus like particles have been reported
in the cytoplasm and nuclei of smooth muscle
cells from the wall of the affected vessel.2 3 10 It
has been suggested that the virus is activated in
the trigeminal ganglion and spreads to the
adjacent carotid artery. Some patients with
herpes zoster associated cerebral angiitis have
had truncal rather than trigeminal zoster eruptions. In such patients, dissemination by either
a haematogenous or cerebrospinal fluid pathway
must be considered. " Hilt et al demonstrated
necrotising angiitis in a patient with herpes zoster ophthalmicus, but did not detect viral antigen in vessels. They suggested an immune
mediated angiitis triggered by the herpes zoster
as an alternate mechanism. Another patient
with central nervous system vasculitis and

underlying Hodgkin's disease was found to have
antibodies to varicella zoster virus infection in
the central nervous system without any previous
vesicular eruption.5
The association of chickenpox and acute or
delayed stroke is also documented. Griffith -et al
described a 31 year old woman who in the acute
phase of chickenpox developed embolic occlusion of several cerebral arteries, due to sterile
endocarditis. These authors have found more
examples of chickenpox associated acute stroke
in the literature.6 An additional case of a 26 year
old patient who developed cerebral infarction 11
days into disseminated varicella was also
reported.7 This patient showed evidence of mild
systemic alteration in immune function, which
was interpreted as suggestive of vasculitis.
Kamholz and Tremblay described a 6 year old
girl who developed a left hemiparesis six weeks
after the onset of chickenpox.9 A computed
tomogram showed hypodensity in the region of
the right basal ganglia and small areas of enhancement in the right centrum semiovale.
Arteriography showed stenosis of the proximal
middle and anterior cerebral arteries with distal
occlusions in both territories. The authors
suggested that corneal inoculation could have
resulted in contiguous spread of the virus to the
internal carotid artery branches. Eda et al described four children (age range 1 to 5 years)
with acute hemiplegia and lacunar infarcts that
occurred one to three months after chickenpox.8
They failed, however, to demonstrate vasculitis
in their patients.
The presumed association of varicella and
vasculitis rests mainly on this temporal occurrence. The findings in the cerebrospinal fluid
are not pathognomonic and may range from
totally normal to a mild monocytic pleocytosis
and a raised protein concentration. The presence of detectable varicella zoster antibodies in
the cerebrospinal fluid (in a titre of 1:2 or
greater) closely correlates with central nervous
system infection with the virus, and is not found
in infected patients without brain involvement. 12 Nevertheless, this finding does not
directly relate the vascular findings to the infection. Definite proof of this linkage can only be
obtained by cerebral biopsy, which is not indicated in our patient.
Our patient's symptoms are similar to those
of the five previously reported children, with
the additional radiographic demonstration of
unilateral angiitis resembling that reported in
both granulomatous angiitis of the nervous
system and after herpes zoster ophthalmicus.
Antibody production within the central nervous
system was suggested by a raised cerebrospinal
fluid: serum antibody ratio (1:30). We speculate
that during a process of generalised dissemination in the central nervous system, the virus was
able to invade the wall of blood vessels, causing

anguitis.

The subsequent course of our patient was not
benign. Although the arteriography showed
that the vasculitic process had improved, there
was a silent extension of the infarct. Moreover,
it may not be excluded that the presence of antibodies in the cerebrospinal fluid in an unchanged titre, represents a reactivation of the

Shuper, Vining, Freeman

1248

virus. Furthermore this reactivation could be
the cause of the child's new onset of seizures
and the evidence on electroencephalography of
a diffuse encephalopathy.
The possible association between the delayed
hemiplegia and chickenpox is clinically important. A large number of management related
questions remain unanswered. If this disease
entity carries a relatively good prognosis,
steroids may not be needed. Indeed, because of
a possible reactivation of the virus, steroids
might be contraindicated and acyclovir would
be indicated instead. The natural history of the
vasculitis, and the possibility of recurrence of a
thrombotic event, necessitating immunosuppressive or antiviral or antithrombotic treatment, are unknown.
It is possible that the acute vasculitis 10
weeks after chickenpox was an unfortunate
coincidence. If so, it is an interesting coincidence. We suggest that, in the future,
physicians treating acute strokes in children
should carefully ask about a history of antecedent chickenpox, and proceed with an appro-

priate evaluation.
The authors thank Dr Richard T Johnson, director of the department of neurology, The Johns Hopkins University and neurologist in chief, The Johns Hopkins Hospital, and Dr R Nick

Bryan, director of neuroradiology division, The Johns Hopkins
Hospital, for reviewing the manuscript and for their helpful
comments.
AS is the recipient of a fellowship from the American
Physicians Fellowship.

1 Kennedy PGE. Neurological complications of varicella-zoster
virus. In: Kennedy PGE, Johnson RT, eds. Infections of the
nervous system. London: Butterworths, 1987:177-208.
2 Sigal HL. The neurologic presentation of vasculitis and
rheumatologic syndromes. Medicine 1987;66:157-80.
3 Morgello S, Block GA, Price RW, Petito CK. Variceila-zoster
virus leukoencephalitis and cerebral vasculopathy. Arch
Pathol Lab Med 1988;112:173-7.
4 Eidelberg D, Sotrel A, Horoupian S, Neuman PE, PumarolaSune T, Price RW. Thrombotic cerebral vasculopathy associated with herpes zoster. Ann Neurol 1986;19:7-14.
5 Vilchez-Padilla JJ, Redon J, Ruiz A, Lopez-Aldeguer J. CNS
varicella-zoster vasculitis. Arch Neurol 1982;39:785.
6 Griffith JF, Salam MV, Adams RD. The nervous system
diseases associated with varicella. Acta Neurol Scand 1970;
46:279-300.
7 Gibbs MA, Fisher M. Cerebral infarction in an adult with
disseminated varicella. Bull Clin Neurosci 1986;51:65-7.
8 Eda I, Takashima S, Takeshia K. Acute hemiplegia with
lacunar infarct after varicella infection in childhood. Brain
Dev 1983;5:494-9.
9 Kamholz J, Tremblay G. Chickenpox with delayed contralateral hemiparesis caused by cerebral angiitis. Ann Neurol
1985;18:358-60.
10 Tanaka K, Hokamura K, Kinjo M. Fukumoto S. Subarachnoid hemorrhage and granulomatous angiitis of the
basilar artery: demonstration of the varicella zoster virus in
the basilar artery lesions. Stroke 1986;17:1024-8.
11 Hilt DC, Buchholz D, Krumholz A, Weiss H, Wolinsky JS.
Herpes zoster ophthalmicus and delayed contralateral
hemiparesis caused by cerebral angiitis: diagnosis and
management approaches. Ann Neurol 1983;14:543-53.
12 Gershon A, Steinberg S, Greenberg S, Taber L. Varicellazoster associated encephalitis: detection of specific antibody
in cerebrospinal fluid. Clin Microbiol 1980;12:764-7.