Journal of Neurology, Neurosurgery, and Psychiatry 1989;52:1402-1409

Regional cerebral blood flow by SPECT imaging in
Sturge-Weber disease: an aid for diagnosis
C CHIRON,* C RAYNAUD,t N TZOURIO,t C DIEBLER,4 0 DULAC,*
M ZILBOVICIUS,t A SYROTAt
From the Clinique de Pediatrie et de Puericulture, Hopital Saint Vincent de Paul, Paris,* Service Hospitalier
Frederic Joliot, Orsay,t Centre Medico-Chirurgical Foch, Suresnes, Francet
SUMMARY Regional cerebral blood flow (rCBF) was studied using SPECT (single photon emission
computed tomography) with 133-Xenon in 13 patients with confirmed Sturge-Weber disease, aged 9
months to 18 years. CT scan, performed at the same time, showed evident cerebral angioma in 10 but
not in three. A marked hypoperfused area was found in all patients, ranging from - 32% to - 72%
and of the same location as the CT signs. The hypoperfusion seems to result from post ictal
phenomenon as well as from chronic ischaemia. SPECT imaging is therefore a sensitive method for
visualising intracranial angioma in Sturge-Weber disease and it provides an aid for diagnosis when a
CT scan is not reliable.

Sturge-Weber (SW) angiomatosis is a non-inherited
neurocutaneous syndrome characterised by a congenital trigeminal port-wine stain and a piamater
angioma.' This intracranial angioma, venous and
purely meningeal,23 is embryologically derived from
the neural crest. Along with the angioma, cerebral
lesions, which are ischaemic in nature, are seen in the
cortex underlying the angioma.4
The diagnosis of SW disease is based on the
association of three characteristics - cutaneous,
neurological and radiological. Among the
neurological features are epilepsy, hemiplegia and
mental retardation. CT is currently the most reliable
technique used to confirm the existence of intracranial
angioma by visualising calcifications, focal cortical
atrophy and opacification after contrast enhancement.5 Diagnosis is easy in severe cases of this
disease since all clinical and radiological characteristics are present, but it is difficult in milder cases where
one or more of these features is absent, such as an
absence of cutaneous angioma,6 a delayed onset of
neurological signs7 or an atypical appearance of the
angioma on the CT, on which calcifications, focal
atrophy or angioma contrast may be absent, particularly in young children.8 Progressive aggravation
of the neurological condition, for which seizures seem
Correspondence to: Dr Chiron, Clinique de Pediatrie et de Puericulture, Inserm U29, Hopital Saint Vincent de Paul, 74 Avenue Denfert
Rochereau, 75674 Paris, Cedex 14, France.
Received 3 January 1989 and in revised form 10 May 1989.
Accepted 14 June 1989

to be responsible, occurs in SW disease.4 Therefore, the
diagnosis of this disease, that is, the presence of an
intracranial angioma, needs to be determined.
Hypometabolism but also hypermetabolism have
recently been reported using positron emission
tomography (PET).9 We studied regional cerebral
blood flow (rCBF) in children with confirmed SW
disease to determine whether a characteristic rCBF
pattern exists and to test the diagnostic value of this
technique with that of CT.

Patients

We studied thirteen patients, eight boys and five girls, aged
between 9 months and 18 years (mean 6 years). The
examinations were performed with the informed consent of
the children's parents.
All 13 were diagnosed as having SW disease. Diagnostic
criteria are summarised in table 1. Facial angioma was found
in all patients except one (number 5). In this case, the
neurological and radiological characteristics were those of
Sturge-Weber disease; arteriography was normal, excluding
other types of vascular malformation. Two patients did not
have seizures before SPECT imaging (numbers 8, 9); one of
these showed neither neurological disorders nor mental
retardation (number 8) but there was no doubt about
cutaneous and cerebral angioma. All patients showed asymmetrical electroencephalographic features, including depressed background activity on the side of the angioma (fig lb),
interictal focal epileptiform activity (fig 4b), or focal epileptic
discharges (fig lc). In ten patients, the CT revealed a typical
pattern of intracranial angioma; for the three others, the CT
disclosed only focal atrophy (number 9), only contrast
enhancement (number 7), or no abnormality (number 12).

1402

Regional cerebral bloodflow by SPECT imaging in Sturge- Weber disease: an aidfor diagnosis

1403

able 1 Patient data
Age at Cutaneous
'umber of
Sex SPECT angioma
atients

Seizures
(First-Last)

EEG data

2

F

14 y

Bilateral

Right hemispheric depression
Bilateral discharges
21 m-13 y 6 m Right hemispheric depression

3

F

7y

Bilateral

8 mr- y 6 m

4

M

6y

Right

6 m- y 10 m Right hemispheric slow waves

5

M

18 y

No angioma 11 m-17 y

6

F

7 y 6 m Right

7

M

I y 5 m Bilateral

8

M

Iy

Right

Q

M

8y

Right

I

F

10 y

Bilateral

6 m- 10 y

Right hemispheric depression

Neurological
examination

CTsigns
Mental
development (Age)

Left hemiplegia

Tetraplegia

Severely
delayed
Severely
delayed
Severely

Normal

Mildly

Left hemiparesis

Left posterior slow waves

Right hemiparesis

31 m-7y6m

Right hemispheric slow waves

Normal

13 m-2 y

Normal

No seizure

Left posterior spikes
Right posterior slow waves
Right anterior slow waves

No seizure

Right slow waves

Normal

0

M

10 m

Left

6 m-9 m

Left posterior slow waves

I

M

11 m

Bilateral

4 m-8 m

Left hemispheric depression

2

F

9m

Left

9 m-9 m

Left hemispheric depression

3

M

24

Left

2 m-24 m

Left hemispheric depression
Left temporal discharges

Normal

*delayed

delayed
Mildly

delayed
Mildly

delayed
Mildly
delayed
Normal

Mildly
delayed
Right hemiparesis Mildly
delayed
Right hemiplegia Mildly
delayed
Right hemiparesis Normal
Right hemiplegia

Mildly
delayed

C + A, CE,
(10 y)
C, A, CE,
(14 y)
C, A, CE,
(7 y)

C, A,CE,

(6 y)

C, A,
(18 y)

A, CE,
(5 y 4 m)
CE,
(2 y)
A, CE,

(1 Y)

A
(8 y)

A, CE,
(10 m)
C, CE,
(11 m)
None
(9 m)
A, C, CE,
(24 m)

ACD

PB

CBZ, CLB
CBZ, CZP
CBZ, CLB
PB, CBZ

CBZ, CZP,
PG
PB, PHT
PB
PB

PB, CBZ
CBZ, CLB
CBZ, CLB

PB, CBZ

1, male; F, female; m, month; y, year; A, atrophy; CE, contrast enhancement; C, calcifications; ACD, anticonvulsant drug; PB, phenobarbital;
BZ, carbamazepine; PG, progabide; CZP, clonazepam; CLB, clobazam; PHT, phenytoin.

ROI rCBF was classified as "decreased" when the difference
with the controlateral value was greater than 20%.
A CT, with contrast injection, was performed on each
Two CTs given when the patient was aged six months and patient, using the same head position as that required for the
SPECT study. Nine mm-thick slices were obtained at the
one year were considered normal. Treatment was therefore
stopped after the second CT. One month later, partial motor same levels OM + 20, + 40, + 60, + 80, and + 100, as in the
status occurred, confirming the diagnosis of SW disease. One SPECT study. Both studies were performed at the same time
in all except two patients for whom there was a time interval
year later, although epilepsy was severe, the patient showed
of 7 and 26 months between the CT and the SPECT (numbers
no permanent focal motor deficit and a CT revealed only
6 and 7 respectively); no neurological changes were observed
moderate abnormalities (table 1).
during this time interval. The ROI employed on SPECT
images were traced on the CT images after the size of the ROI
Methods
was homothetically reproduced on the CT image. Three types
The rCBF was measured by SPECT with a Tomomatic 564 of abnormalities were studied on the CT scan: calcifications,
using 133 Xenon, and injected intravenously in doses ranging atrophy and contrast enhancement after injection.
from 1 5 to 2-5 mCi/Kg. Data collection lasted 4 5 minutes.
Results were obtained on five 20 mm-thick slices, parallel to Results
the orbito-meatal (OM) line, at levels OM + 20, OM + 40,
OM + 60, OM + 80 mm and OM + 100 mm. The radiation
the CT abnordose to the target organ, the lung, was 250 to 350 mrad. To In six patients (numbers 1-4, 11, 13),
and
enhancement
contrast
since
were
large
malities
children
acquisition,
during
from
moving
head
prevent the
under 6 years old received 4 mg/Kg of rectal pentobarbital calcifications covered an entire hemisphere. In all six
and 0 5 mg/Kg of intramuscular droperidol. All patients were patients, an rCBF decrease was observed in a large
receiving anticonvulsant drugs (table 1) and none of the cortical and subcortical area of the same hemisphere.
patients experienced seizures on the day of the study.
In the cortex, the hypoperfused area was approxThe rCBF was measured on 20 cortical regions of interest imately of the same size as the CT abnormal area (fig
(ROI) per slice. The ROI were circular and ranged from 1-5 I a). In the subcortical regions, the size of the hypoperto 2 5 cm2 in size. For each patient, the mean rCBF or "global fused area was easily measured, but could not be
rCBF", was calculated as the mean half-slice rCBF for slices
the CT abnormal area which
OM + 40, + 60, and + 80, on the undamaged or the less compared with that of
intensity of the rCBF decrease
The
defined.
poorly
was
a
as
were
values
expressed
rCBF
damaged side. ROI
percentage of the individual mean rCBF and also as a was considerable, with maximal value ranging from
percentage of the contralateral value in the symmetrical ROI. 32% to 72% depending on the patients.
Patient number 7 had a special history. Anticonvulsant

treatment had been given since birth for suspected SW
disease because of facial angioma and to prevent seizures.

Chiron, Raynaud, Tzourio, et al

1404

(a)
Fig 1 Patient number 13. (a)
large abnormal CT area involving
the whole left hemisphere (top) with
a hypoperfused area (bottom) of
approximately the same size. (b)
Depressed interictal background
activity on the whole left
hemisphere. (c) Left temporal
epileptic discharge at the same age.

379

(b)

iwrY
--vw-vv~SWN
N8r4
25

(c)

M

a

vvlit' V l1
Al

'.' 1'

1

bet
;..t'.'t.rfi'ui'j
.'A..;&4
it..tlI.vvllrTIrFIt1N

*

A ,

AA

$4

$

N9N~~r4hA\ltWNI,$4ivA1kv~

V

F

1405

Regional cerebral bloodflow by SPECT imaging in Sturge- Weber disease: an aidfor diagnosis
Fable 2 Comparative data on SPECT and CT scan
Number ofpathological
cortical roi

Vumber of Location of
patients cT abnormalities
I
2

3
4
5
6
7
8
9
10
11
12
13

RRdiffuse
Bilateral atrophy
R diffuse
L posterior
R diffuse
Bilateral atrophy
R diffuse
L posterior
R posterior
Lposterior
R posterior
R anterior
R posterior
L posterior
L diffuse
R posterior
L diffuse

Location of SPECT

Minimal rCBF observed

ml/100 g/mn

Mean rCBF

Maximal rCBF
decrease*

21 (58)

52

-59%

31(43)
25 (37)
29 (85)
31(70)
35 (57)
39(71)
52 (63)
42 (54)
36(64)
44 (85)
36 (80)
27 (95)

46

-32%

47

-45%

73
62
53
73

-50%
-33%
-45%

(symmetrical rCBF)

hypoperfusion

On CT

On SPECT

ml/100 g/mn

Rdiffuse

30

30

L central
R diffuse
L posterior
R diffuse
L central
R diffuse
L posterior
R posterior

Lposterior

R posterior
R anterior
R posterior
L posterior
L diffuse
R posterior
L diffuse
L diffuse

41

41

30

30

30
8
10
6

28
10
11
11

5
2
6
30

3
5
10
30

0
30

20
30

-60%

- 32%
- 37%

76
67
63
72

-43%
-39%

67
80

-46%
-72%

'Expressed in percentage of mean rCBF.
'I, right; L, left; ROI, region of interest.

In six patients (numbers 5-10), the CT abnormalities covered a small area. In all six, an rCBF
decrease was found in an area corresponding to the CT
scan abnormal area. In the cortical region, the size of

Fig2

the hypoperfused area was similar to that of the CT
abnormal image in three patients (fig 2) and slightly
larger in the other three (numbers 7, 9, 10) (figs 3, 4). In
these six patients, the rCBF decrease was also marked,

----------area
-d
corresponding
atintnumer5.SmalbnomaCare(op)wih

Fig 2 Patient number S. Small abnormal CT area (top) with a corresponding hypoperfused area on SPECT (bottom).

1406

Chiron, Raynaud, Tzourio, et al

Fig 3 Patient number 9. CTscan performed after contrast injection (top) shows no enhancement but onlyfocal atrophy (the
posterior contrast area corresponds to the superior longitudinal sinus). The hypoperfused area is clearly seen on rCBF images
(bottom).

with individual values ranging from 32% to 45%.
In the remaining patient (number 12), SPECT
showed a localised rCBF decrease, with maximal
decrease of 50%, although the CT was normal (fig 5).
These results show that in all the patients in this
series, SPECT clearly defined a hypoperfused area,
which was also seen in the patients with or without the
typical appearance of angioma on CT.
Discussion
The diagnosis of Sturge-Weber disease is based on the
association of clinical and radiological characteristics.
Cutaneous angioma is usually present at birth.
Epilepsy usually appears in the first years of life and
hemiplegia follows the first prolonged seizures although some patients may remain seizure free.
Cerebral angioma is constant but radiological characteristics may be delayed, or reduced to non-specific
features, or may even be occasionally absent. SturgeWeber disease is easily diagnosed when all the features
are present, but diagnosis is difficult when the
neurological characteristics are absent, and especially
when a CT provided no additional information. An

additional diagnostic method would therefore be
useful.
The method used to measure rCBF has been
previously described.'0 It is easy to perform, does not
require any blood samples, and only takes 5 minutes.
The SPECT system used is especially adapted to the
brain and is highly sensitive. The validity of this
method has been studied for the cortex by comparing
rCBF values in the same patients with those obtained
by SPECT and PET using continuous inhalation of
C'502. The rCBF values obtained by both methods
were significantly correlated." Adapting the SPECT
technique for children required two modifications:
intravenous (IV) injection of the 133-Xenon instead of
inhalation and the use of premedication.'2 Using the
IV injection method, rCBF values increased by a mean
of 10%, probably due to the shape of the Xenon input
curve.'3 The premedication used (pentobarbital and
droperidol) did not cause any significant change in
rCBF in our research.
A study of patients with confirmed Sturge-Weber
disease was necessary to test the ability of SPECT
imaging to visualise the intracranial angioma. Our
results show that all patients had a marked hypo-

Regional cerebral bloodflow by SPECT imaging in Sturge- Weber disease: an aidfor diagnosis

1407

1

Fig 4 Patient number 7. (a) CTscan after contrast injection at one year is considered normal (top). SPECT imaging at one
yearfive months shows a well definedposterior hypoperfused area (middle). At two years, CT scan after injection only shows a
discrete enhancement in the posterior cortex and a hypertrophy of the left choroidplexus (bottom). (b) Left temporalfocus of
spikes and right temporalfocus of slow waves on the interictal EEG recorded at one year six months.

1408

Fig 5 Patient number 12. CT scan after injection is normal
(top) but SPECTshows at the same age a left diffuse
hypoperfusion (bottom).

Chiron, Raynaud, Tzourio, et al
meningeal. Ischaemia progressively develops in the
brain underlying the angioma and involves cortical
and subcortical regions, due to the obstruction by the
angioma of the normal venous return. This produces
an abnormal drainage into the deep plexus and
hypertrophy of the choroid plexus visible on CT scan.
Stasis and slowing in venous cortical circulation have
been demonstrated by dynamic angiography,202' and
hypoxia, preoperatively, has been shown.22 The
ischaemic mechanisms alone may explain the
hypoperfusion seen in our two non epileptic patients.
Hypoperfusion may also be a consequence of the
seizures in the post-ictal period. Evidence ofconsistent
presence of epileptic foci related to the angioma is well

known in SW disease23 and was also seen in our series.
This mechanism alone may be involved in our patient
with seizures without CT scan abnormalities, as
decribed in partial epilepsy.2425 Alexander suggests
that both ischaemic and epileptic mechanisms may be
involved together.4 The complex circulatory condition
in SW disease may explain why seizures are particularly severe and often followed by permanent
hemiplegia. In normal conditions, cerebral blood flow
locally increases during seizures.26 In SW disease, the
insufficient increase in blood supply during seizures
may exacerbate ischaemic brain lesions. But the
respective responsibility of ischaemic and epileptic
phenomena in lesional mechanisms remains unclear,
as well as the relation between hypoperfusion and the
reported abnormalities on metabolism.9
Our results show that the considerable hypoperfusion observed by SPECT imaging in patients with
Sturge-Weber disease is a reliable characteristic of the
disease. Since it is consistently observed, even in cases
without clear evidence of angioma on CT, SPECT
imaging may be considered a useful complement to
CTs for diagnosing SW disease in childhood. Further
studies are needed to test its diagnostic value in the first
year of life.

perfusion in the pathological cerebral regions, suggesting that such hypoperfusion is a reliable sign of SturgeWeber disease. In our series, the CT scan pattern was
incomplete in some patients and diagnosis was established after the neurological features appeared: in one
of them CT scan was normal, in another it showed
such mild abnormalities that they had been
overlooked on first examination. In three patients, the
clinical pattern was incomplete. Two of the three had
never had any seizures. However, in these patients, a
hypoperfused area was also clearly defined. This
indicates that rCBF decrease may be present even
when CT and neurological features are incomplete or
absent. SPECT imaging may therefore be considered a
complementary method for diagnosing Sturge-Weber
disease, which is especially useful in cases with an
atypical CT scan.
The hypoperfusion described in our patients should We thank Doctors J Motte, C Billard, P Pineau and C
not be confused with the hypoperfusion of immature Boulloche for their kind assistance.
areas, which is seen before the age of six months in the
temporo-parieto-occipital cortex and is always symmetrical.'4"5 Hypoperfusion should not be related to
the anticonvulsant drugs taken by our patients
because such drugs produce diffuse changes in blood References
flow and metabolism but not focal defects.'6"7 The
hypoperfusion observed here is therefore probably I Alexander GL, Norman RM. The Sturge- Weber syndrome. Bristol: Wright and Son, 1960.
due to the lesions of SW disease.
0, Riche MC, Merland J. Facial port-wine stains and
Focal hypoperfusion is a non-specific finding that 2 Enjolras
Sturge-Weber syndrome. Pediatrics 1985;76:48-51.
can reflect different mechanisms. Two of them may be
3 Couly GF, Le Douarin MM. Mapping of the early neural
taken into account in this disease, ischaemic and
primordium in quail-chick chimeras. II. The prosencephalic
neural plate and neural folds: implications for the genesis of
epileptic. The anatomical lesions described in Sturgecephalic
human congenital abnormalities. Dev Biol
Weber are mainly angioma and ischaemic lesions.489
1987;120:198-214.
The angioma affects the superficial areas because it is 4 Alexander GL. Sturge-Weber syndrome. In: Vinken PJ, Bruyn

Regional cerebral bloodflow by SPECT imaging in Sturge- Weber disease: an aidfor diagnosis
GW, eds. Handb Clin Neurol. Amsterdam: North Holland
Publishing, 1972;14:223-40.
5 Welch K, Naheedy MH, Abroms IF, Strand RD. Computed
Tomography of Sturge-Weber Syndrome in Infants. J Comput
Assist Tomogr 1980;4:33-6.
6 Gorman RJ, Snead OC. Sturge-Weber syndrome without portwine nevus. Pediatrics 1977;60:785.
7 Gilly R, Lapras C, Tommasi M, Revol M, Challamel MJ, Clavel
D, Manuelle JC. Maladie de Sturge-Weber-Krabbe. Pediatrie
1977; 1:45-64.
8 Diebler C, Dulac 0. Pediatric Neurology and Neuroradiology.
Berlin: Springer Verlag, 1987:99-101.
9 Chugani HT, Mazziotta JC, Phelps ME. Sturge-Weber syndrome:
a study of cerebral glucose utilization with positron emission
tomography. J Pediatr 1989;1 14:244-53.
10 Lassen NA. Cerebral blood flow tomography with Xenon-133.
Semin Nucl Med 1985;15:347-56.
11 Raynaud C, Mazoyer B, Soucy JP, et al. Regional cerebral blood
flow (rCBF) measured by SPECT with 133-Xenon: validation
using a PET/SPECT comparison. Eur J Nucl Med 1988;14:308.
12 Dulac 0, Raynaud C, Chiron C, Plouin P, Syrota A, Arthuis M.
Etude du debit sanguin cerebral dans le syndrome de West
idiopathique: correlation avec les donnees electroencephalographiques. Rev EEG Neurophysiol lin 1987;17:169-82.
13 Mazoyer BM, Raynaud C, Tzourio N, et al. Error analysis of
regional cerebral blood flow (rCBF) measured in children by
SPECT with 133-Xe. J Nucl Med 1988;29:918.
14 Tzourio N, Chiron C, Raynaud C, Dulac 0, Mazoter B, Bourguignon M, Syrota A. Cerebral maturation during the first 20
months of life studied with the regional cerebral blood flow
measurements using SPECT. J Nucl Med 1988;29:743.
15 Chugani HT, Phelps ME, Mazziotta JC. Positron emission
tomography study of human brain functional development.
Ann Neurol 1987;22:487-97.

1409
16 Theodore WH, Dichirio G, Margolin R, Porter RJ, Brooks RA.
Barbiturates depress human cerebral glucose metabolism.

Neurology 1986;36:60-4.
17 Theodore WH, Bairamina D, Newmark ME, et al. Effect of
phenytoin on human cerebral glucose metabolism. J Cereb
Blood Flowv Metab 1986;6:615-320.
18 Greenfields JG. Neuropathology. Edinburgh: Edward Arnold
Publisher, 1976:417-20.
19 Norman MG, Schoene WC. Ultrastructure of Sturge-Weber
disease. Acta Neuropathol 1977;37:199-205.
20 Bentson JR, Wilson GH, Newton TH. Cerebral venous drainage
pattern of the Sturge-Weber syndrome. Neuroradiolog-

1971;101:1 11-8.

21 Probst FP. Vascular morphology and angiographic flow patterns
in Sturge-Weber angiomatosis: facts, thoughts and suggestions. Neuroradiology 1980;20:73-8.
22 Alexander GL, Cooper R, Crow HJ. EEG, Ecg and oxygen
availability (AO2) in the cortex of a case of Sturge-Weber
syndrome. Electroencephalogr Clin Neurophysiol 1972;14:284.
23 Brenner RP, Sharbrough FW. Electroencephalographic evaluation in Sturge-Weber syndrome. Neurology 1976;26:629-32.
24 Bonte FJ, Devous MD, Stokely EM, Homan RW. Single-photon
tomographic determination of regional cerebral blood flow in
epilepsy. AJNR 1983;4:544-6.
25 Engel J, Brown WJ, Kuhl E, Phelps ME, Mazziota JC, Crandall
PH. Pathological findings underlying focal temporal lobe
hypometabolism in partial epilepsy. Ann Neurol 1982;12:
518-28.
26 Franck G, Sadzot B, Salmon E, et al. Regional cerebral blood flow
and metabolic rates in human focal epilepsy and status
epilepticus. In: Delgado-Escueta AV, Ward A, Woodbury
DM, Porter RJ, eds. Advances in Neurology. New York: Raven
Press, 1986.