Neurol Med Chir (Tokyo) 43, 255¿258, 2003

Brainstem Congestion Caused by Direct
Carotid-Cavernous Fistula
—Case Report—
Hidetoshi MURATA, Takeshi KUBOTA, Masao MURAI, Hiroshi KANNO,
Satoshi FUJII, and Isao YAMAMOTO
Department of Neurosurgery, Yokohama City University School of Medicine,
Yokohama, Kanagawa

Abstract
A 41-year-old woman presented with tinnitus in the left ear and headache, followed by diplopia and
pain in the left cheek. Angiography showed a left high-flow direct carotid-cavernous fistula (CCF), causing steal of the blood flow from the internal carotid artery into the cavernous sinus. A few days later,
she rapidly developed right hemiparesis, dysarthria, and ocular conjugate deviation to the right, and
became somnolent. Angiography at that time revealed occlusion of the superior petrosal sinus, causing
engorgement of the veins in the surrounding brainstem. The CCF was completely embolized with interlocking detachable coils. Her consciousness disturbance and ophthalmoparesis dramatically improved
within a few days, and the right hemiparesis and dysarthria gradually resolved. Magnetic resonance
(MR) imaging after the treatment showed small pontine hemorrhage and perifocal edema but no ischemic lesions in the cerebral hemisphere. Re-evaluation of the MR imaging with gadolinium taken on
admission demonstrated engorged veins in the brainstem parenchyma, which corresponded to the
hemorrhagic lesion in the brainstem. Brainstem congestion caused by direct CCF is very rare, but it can
be life-threatening. Good outcome can be expected if the CCF is completely occluded before congestive
hemorrhage occurs.
Key words:

carotid-cavernous fistula,

venous congestion,

Introduction

no history of head trauma.
Neurological examination revealed a bruit in the
left orbit, left abducens nerve paralysis, and
hypesthesia in the first and second divisions of the
left trigeminal nerve. There was no chemosis or
proptosis. MR angiography showed a disordered signal void surrounding an extension of the cavernous
sinus. Angiography detected a left high-flow direct
CCF, which caused steal of the blood flow from the
internal carotid artery (ICA) into the superior
petrosal sinus (SPS) and inferior petrosal sinus (IPS),
resulting in cut off of the flow of the left middle
cerebral artery (MCA) (Fig. 1A). The MCA was
supplied by cross flow from the contralateral
side through the anterior communicating artery
(Fig. 1B). Single photon emission computed
tomography
with
technetium-99m-hexamethylpropyleneamine oxime showed preservation of the
cerebral blood flow of the left cerebral hemisphere.
Four days after admission, she rapidly developed
right hemiparesis, dysarthria, and ocular conjugate

Life-threatening complications associated with
carotid-cavernous fistulas (CCFs) include intracerebral hemorrhage, subarachnoid hemorrhage,
epistaxis, and cerebral ischemia due to steal of blood
flow by the CCF. We describe a case of brainstem
congestion due to a direct-type CCF, which is an
extremely rare complication.

Case Report
A 41-year-old woman complained of tinnitus in the
left ear and headache. Otolaryngological examination and magnetic resonance (MR) imaging in a
general hospital revealed no abnormalities. Five
days later, she developed diplopia and pain in the
left cheek, and was admitted to our hospital. She had
Received
2003

August 20, 2002;

Accepted

brainstem

January 15,

255

256

Fig. 1

H. Murata et al.

A: Left carotid angiogram on admission
showing a high-flow direct carotid-cavernous fistula (CCF) causing steal of the flow
from the internal carotid artery (ICA) into
the cavernous sinus, and no opacification of
the middle cerebral artery (MCA). B: Right
carotid angiogram on admission revealing
opacification of the left MCA via the anterior communicating artery. C: Left
carotid angiogram just before treatment
demonstrating occlusion of the superior
petrosal sinus, increased flow in the inferior
petrosal sinus, and engorgement of the veins
in the prepontine and interpeduncular
cisterns (arrowheads). D: Left carotid angiogram after embolization showing obliteration of the CCF and complete opacification
of the ICA and MCA.

deviation to the right, and became somnolent. We at
first suspected that ischemia of the left cerebral
hemisphere had developed due to the flow steal from
the left ICA into CCF. However, computed
tomography (CT) revealed no ischemic lesions in the
left cerebral hemisphere. We could not accurately
evaluate the brainstem lesions with CT because of
bone artifacts. Next day, the CCF was embolized
using 30 interlocking detachable coils. Left carotid
angiography just before the procedure revealed that
the flow dynamics of the CCF were little changed.
However, spontaneous occlusion of the SPS had
occurred, resulting in increased flow in the IPS and
engorgement of the veins in the prepontine and
interpeduncular cisterns (Fig. 1C).
Immediately after embolization, angiography
showed that the CCF was completely obliterated and

Fig. 2

A, B: T1-weighted (A) and T2-weighted (B)
magnetic resonance (MR) images one week
after the treatment showing small pontine
hemorrhage and perifocal edema. C: Reevaluation of the T1-weighted MR image
with gadolinium-diethylenetriaminepentaacetic acid taken on admission demonstrating engorgement of the perforating vein
within the pontine parenchyma coinciding
with the subsequent hemorrhage and edema
(arrow). D: T2-weighted MR image 4 weeks
after the treatment revealing marked reduction of the perifocal edema.

both the ICA and MCA were completely opacified
(Fig. 1D). The left orbital bruit disappeared. Her
consciousness disturbance and ophthalmoparesis
dramatically improved within a few days, and the
right hemiparesis and dysarthria gradually resolved.
MR imaging one week after the treatment showed
small pontine hemorrhage and perifocal edema but
no ischemic lesions in the cerebral hemisphere
(Fig. 2A, B).
Re-evaluation of the MR imaging with
gadolinium-diethylenetriaminepenta-acetic
acid
(Gd-DTPA) taken on admission identified engorged
veins in the brainstem parenchyma, which corresponded to the subsequent hemorrhagic and
edematous lesion in the brainstem (Fig. 2C). Four
weeks after the treatment, the right hemiparesis and
dysarthria were almost completely resolved. MR
imaging showed reduction of the brainstem edema
(Fig. 2D). Three months later, she returned to her
former office work.

Neurol Med Chir (Tokyo) 43, May, 2003

Brainstem Congestion Caused by Direct CCF
Table 1

257

Brainstem congestion caused by direct carotid-cavernous fistula
Angiography

Case
No.

Author (Year)

Age,
Sex

Location

Symptoms

Occlusion
of SPS

Engorged
veins in the
surrounding
brainstem

1

Turner et al.
(1983)7)
Teng et al.
(1991)6)

81, F

midbrain

NM

+

NT

—

36, M

midbrain,
thalamus

+

midbrain

+

dilation of
proximal
SPS
+

NT

26, M

transarterial
balloon
embolization
transarterial
balloon and
coil
embolization

no neurological
signs
no neurological
signs

49, F

pons

hemiparesis,
comatose
unconsciousness,
hemiparesis
numbness of
the right
side of
the body
and face
hemiparesis,
dysarthria,
somnolence

+

+

transarterial
coil
embolization

no neurological
signs

2
3

4

Present case

MR imaging
with Gd-DTPA

NT

engorged
veins in
pontine
parenchyma

Treatment

Outcome

death

Gd-DTPA: gadolinium-diethylenetriaminepenta-acetic acid, MR: magnetic resonance, NM: not mentioned, NT: not
tested, SPS: superior petrosal sinus.

Discussion
This patient presented with a high-flow direct CCF
associated with pontine hemorrhage and edema. We
first suspected that the flow steal from the left ICA
into the cavernous sinus3) had caused ischemia of
the left cerebral hemisphere when the patient developed hemiparesis and dysarthria, and subsequently
became somnolent. However, she showed ocular
conjugate deviation to the right, which can be
caused by lesions in the right side of the pons, but
not by lesions in the left cerebral hemisphere.
Moreover, neither CT before the treatment nor MR
imaging after the treatment had detected any
ischemic lesions in the left hemisphere. MR imaging
one week after embolization of the CCF detected
small hemorrhage and perifocal edema in the brainstem (Fig. 2A, B). MR imaging with Gd-DTPA had
previously showed engorged veins in the pontine
parenchyma, which corresponded to the subsequent
hemorrhage and edema (Fig. 2C). Angiography just
before embolization had demonstrated spontaneous
occlusion of the SPS, which is an essential part of
the venous drainage system from the brainstem via
the petrosal veins. Occlusion of the SPS had resulted
in backward flow into and congestion of the veins in
the prepontine and interpeduncular cisterns
(Fig. 1C). Therefore, the small hemorrhage probably
resulted from brainstem venous congestion due to
the CCF.
Brainstem hemorrhage may occur after intravascular embolization or incomplete embolization of
direct CCFs.2) However, the treatment did not cause
hemorrhage in the present case because the critical

Neurol Med Chir (Tokyo) 43, May, 2003

symptoms presented before and improved soon after
the treatment.
Brainstem congestion caused by direct CCF has
been reported in three cases, which all involved congestion in the midbrain (Table 1).6,7) In contrast, the
present case had pontine hemorrhage associated
with direct CCF. The important factors in the
development of brainstem congestion are changes in
the drainage pattern of the SPS and engorged veins
in the surrounding brainstem. Occlusion of the SPS
was found in three cases including ours, and dilated
veins related to brainstem congestion were found in
all four cases. MR imaging with Gd-DTPA clearly
showed the engorged perforating veins causing the
hemorrhage in the present case, so is very useful for
detecting venous congestion.
CCF associated with a high risk of intracerebral
venous hemorrhage may appear as parenchymal
vermicular enhancement of brain vessels on CT, and
as a pattern of dilated and tortuous cerebral veins on
angiography.1) Our case demonstrated all these findings in the brainstem. Five cases of pontine congestion have been caused by indirect dural CCF.4,5,8)
Nonopacification of the SPS was demonstrated in
four of these cases as well as in some cases of direct
CCFs. Abnormal angiographical staining was seen
at the pons in two cases8) and unusual pontine
enhancement on MR imaging, demarcated by the
pontine raphe, in three cases,4,5) but these characteristics were not seen in cases of direct CCFs. In
contrast, no engorged veins were found in the surrounding brainstem in any case of pontine congestion due to dural CCF. Such abnormal staining may
result from abnormal communication of the multi-

H. Murata et al.

258

ple dural branches in the nearby brainstem, which
reflects a different pathogenesis to that caused by
direct CCF.
Changes occurring in the venous drainage during
and after embolization require special attention, and
incomplete occlusion must be avoided because of
the possibility of malignant venous congestion such
as brainstem hemorrhage rather than resolution of
the CCF. Good outcome can be expected if the CCF
is completely occluded before brainstem hemorrhage occurs.

4)

5)

6)

7)

References
1)

d'Angelo VA, Monte V, Scialfa G, Fiumara E, Scotti G:
Intracerebral venous hemorrhage in ``high-risk''
carotid-cavernous fistula. Surg Neurol 30: 387–390,
1988
2) Goto K, Hieshima GB, Higashida RT, Halbach VV,
Bentson JR, Mehringer CM, Pribram HF: Treatment of
direct carotid cavernous fistulae. Various therapeutic
approaches and results in 148 cases. Acta Radiol Suppl
369: 576–579, 1986
3) Iida K, Uozumi T, Arita K, Nakahara T, Ohba S, Satoh
H: Steal phenomenon in a traumatic carotid-cavernous fistula. J Trauma 39: 1015–1017, 1995

8)

Shintani S, Tsuruoka S, Shiigai T: Carotid-cavernous
fistula with brainstem congestion mimicking tumor on
MR imaging. Neurology 55: 1929–1931, 2000
Takahashi S, Tomura N, Watarai J, Mizoi K, Manabe
H: Dural arteriovenous fistula of the cavernous sinus
with venous congestion of the brain stem: report of
two cases. AJNR Am J Neuroradiol 20: 886–888, 1999
Teng MM, Chang T, Pan DH, Chang CN, Huang CI,
Guo WY, Chen CC, Pang RG, Lee LS: Brainstem
edema: an unusual complication of carotid cavernous
fistula. AJNR Am J Neuroradiol 12: 139–142, 1991
Turner DM, Vangilder JC, Mojtahedi S, Pierson EW:
Spontaneous intracerebral hematoma in carotidcavernous fistula. Report of three cases. J Neurosurg
59: 680–686, 1983
Uchino A, Kato A, Kuroda Y, Shimokawa S, Kudo S:
Pontine venous congestion caused by dural carotidcavernous fistula: report of two cases. Eur Radiol 7:
405–408, 1997

Address reprint requests to: H. Murata, M.D., Department
of Neurosurgery, Yokohama City University School
of Medicine, 3–9 Fukuura, Kanazawa–ku, Yokohama
236–0004, Japan.
e-mail: hmurata＠h2.dion.ne.jp.

Neurol Med Chir (Tokyo) 43, May, 2003