J Neurosurg 121:63–66, 2014
©AANS, 2014

Extensive basal ganglia edema caused by a traumatic
carotid-cavernous fistula: a rare presentation related to a
basal vein of Rosenthal anatomical variation
Case report
Isabelle Ract, M.D.,1 Aurélie Drier, M.D., 2 Delphine Leclercq, M.D., 2
Nader Sourour, M.D., 2 Joseph Gabrieli, M.D., 3 Marion Yger, M.D., 4
Aurélien Nouet, M.D., 5 Didier Dormont, M.D., 2 Jacques Chiras, M.D., 2
and Frédéric Clarençon, M.D. 2
1
Department of Radiology, Pontchaillou Hospital, CHU Rennes; Departments of 2Neuroradiology, 4Vascular
Neurology, and 5Neurosugery, Pitié-Salpêtrière Hospital, Paris VI University, Paris, France; and
3
Department of Radiology, University of Padova, Italy

The authors report a very rare presentation of traumatic carotid-cavernous fistula (CCF) with extensive edema of
the basal ganglia and brainstem because of an anatomical variation of the basal vein of Rosenthal (BVR). A 45-yearold woman was admitted to the authors’ institution for left hemiparesis, dysarthria, and a comatose state caused by
right orbital trauma from a thin metal rod. Brain MRI showed a right CCF and vasogenic edema of the right side of
the brainstem, right temporal lobe, and basal ganglia. Digital subtraction angiography confirmed a high-flow direct
CCF and revealed a hypoplastic second segment of the BVR responsible for the hypertension in inferior striate veins
and venous congestion. Endovascular treatment was performed on an emergency basis. One month after treatment,
the patient’s symptoms and MRI signal abnormalities almost totally disappeared.
Basal ganglia and brainstem venous congestion may occur in traumatic CCF in cases of a hypoplastic or agenetic
second segment of the BVR and may provoke emergency treatment.
(http://thejns.org/doi/abs/10.3171/2014.1.JNS132016)

Key Words
•
traumatic carotid-cavernous fistula
•
MRI
basal vein of Rosenthal
•
edema
•
vascular disorders

A

direct carotid-cavernous fistula (CCF) is an abnormal communication between the internal carotid artery (ICA) and the cavernous sinus leading to a high-flow shunt.11 The mechanisms that may be
responsible for a direct CCF are traumatic conditions and
spontaneous rupture of an aneurysm located on the cavernous segment of the ICA. Clinically, patients with a direct CCF usually present with pulsating proptosis, corneal edema, and chemosis that are related to engorgement
of the superior ophthalmic vein draining into the cavernous sinus.12 Rarely, CCFs are responsible for brainstem
edema due to reflux into a superior petrosal sinus (SPS)
draining the lateral mesencephalic vein.7
We describe a rare presentation of traumatic CCF

Abbreviations used in this paper: BG = basal ganglia; BVR =
basal vein of Rosenthal; CCF = carotid-cavernous fistula; DSA =
digital subtraction angiography; ICA = internal carotid artery; SPS
= superior petrosal sinus; SWAN = susceptibility-weighted angiography; TOF = time of flight.

J Neurosurg / Volume 121 / July 2014

•

basal ganglia

•

with edema of the basal ganglia (BG) due to an anatomical variation of the basal vein of Rosenthal (BVR).

Case Report
History and Examination. A 45-year-old woman was
emergently admitted to our hospital for dysarthria, left
hemiparesis, and headache rapidly evolving to a comatose state. Right ocular trauma from a thin metal rod had
recently occurred. Brain MRI (3 T) was emergently performed (Fig. 1). Three-dimensional time of flight (TOF)
acquisition showed hyperintense signal within the right
cavernous sinus; these findings were suggestive of a direct CCF. Both FLAIR and T2-weighted images showed
a hyperintense signal in the right side of the pons, the
mesencephalon extending to the right middle cerebellar
peduncle, the right BG, and the medial aspect of the right
temporal lobe. Since apparent diffusion coefficient values
were not decreased in these areas, the T2-weighted hyperintense signals were suggestive of vasogenic edema.
63

I. Ract et al.

Fig. 1. A 45-year-old woman presented with left hemiparesis, dysarthria, and headache rapidly evolving to a comatose state
because of right ocular trauma from a thin metal rod. Three-dimensional TOF image (A) showing hyperintense signal within the
right cavernous sinus, which is bulging laterally. These findings are suggestive of a direct CCF. Coronal FLAIR (B) and axial
T2-weighted (C) images showing a hyperintense signal involving the brainstem, the medial aspect of the right temporal lobe,
and the right BG. Three-dimensional axial SWAN acquisition (D) and MinIP SWAN coronal reconstruction (E) showing a dilated
vein close to the BG (white arrows). Note the presence of an intraventricular hemorrhage in the occipital horn of the right lateral
ventricle (D), probably related to the direct trauma.

Susceptibility-weighted angiography (SWAN) images
showed a focal hematoma in the right side of the pons associated with a faint intraventricular hemorrhage probably
from the direct trauma caused by the thin metal rod. Additionally, a tube-shaped structure suggestive of a dilated
vein was seen close to the right BG (Fig. 1).

Treatment. Digital subtraction angiography (DSA)
was performed with the patient under general anesthesia
for diagnostic and therapeutic purposes (Fig. 2). These
imaging studies confirmed a high-flow direct CCF in the
right cavernous portion of the ICA with reflux toward the
contralateral cavernous sinus. Precise analysis of the venous network on the DSA studies revealed the presence
of a hypoplastic second segment of the BVR; the first
segment and its branches mainly drained into the cavernous sinus via the uncal vein. Thus, at the arterial phase,
both the right cavernous sinus and the first segment of the
BVR and its branches were opacified via the torn right
ICA. It is noteworthy that the right SPS did not appear
patent on DSA.
Because there was no filling of the supraclinoid
branches of the right ICA, the circle of Willis was functional, and the tear on the ICA was large, the decision was

64

made to close both the fistula and the right ICA endovascularly with coils via a double-lumen balloon (Ascent 6–9
mm, Micrus, Johnson & Johnson). The procedure was
performed under full heparinization (activated clotting
time between 2- and 3-fold greater than baseline) without
any technical difficulty.
Posttreatment Course. At the 1-month clinical follow-up, the patient had a satisfactory recovery, with only
slight residual left hemiparesis. On follow-up MRI (3 T),
the T2 and FLAIR signal abnormalities had almost totally disappeared (Fig. 3).
The angiography findings (first segment of the BVR
draining into the cavernous sinus) and the resolution of
MRI signal abnormalities on follow-up after shunt occlusion confirmed the initial diagnosis of engorgement of the
veins draining into the first segment of the BVR due to
retrograde filling related to the direct CCF.

Discussion

Carotid-cavernous fistulas can be classified into 4
groups according to Barrow et al.1 The first group, Type
A, the most severe type, is a direct connection between
J Neurosurg / Volume 121 / July 2014

Atypical traumatic carotid-cavernous fistula

Fig. 3. Axial FLAIR image obtained at the 1-month follow-up, showing total disappearance of the signal abnormalities.

Fig. 2. Right ICA digital subtraction angiograms, anteroposterior
(upper) and lateral (lower) projections. Opacification of both cavernous
sinuses is seen at the arterial phase, confirming the diagnosis of a right
direct CCF (arrows). On the lateral projection, a dilated first segment of
the BVR appears as well as the connection between the BVR and the
cavernous sinus via the uncal vein. The second segment of the BVR
appears hypoplastic. Note the absence of opacification of the SPS (arrowhead). Inf = inferior; I, II, and III = first, second, and third segments
of the BVR.

the ICA and the cavernous sinus. The other groups, Types
B, C, and D, are each an indirect dural communication
between dural arteries and the cavernous sinus. Our patient had a posttraumatic Type A direct CCF.
The clinical presentation of CCFs depends on the
degree of shunting and the venous drainage pathways.
The cavernous sinus receives drainage anteriorly from
the ophthalmic veins and superiorly from sphenoparietal
sinus and cortical veins. Then blood flow into the cavernous sinus drains posteriorly into the inferior petrosal
J Neurosurg / Volume 121 / July 2014

sinus and the SPS and inferiorly into the pterygoid plexus
via emissary veins.7,11
Generally, patients present with the classic clinical
triad of proptosis, chemosis, and pulsatile bruits caused
by orbital venous congestion. Less commonly, the venous
congestion can be posterior (cerebral and brainstem),
causing neurological symptoms rather than ocular ones.
Rarely, reflux into the cortical veins can be observed and
represents a risk factor for focal neurological deficit, seizure, venous infarction, or even intracranial hemorrhage.11
Mechanisms leading to venous infarction caused by
venous congestion have been studied for indirect CCFs,2,8
but remain poorly understood for traumatic direct CCFs.
In indirect dural CCFs, risk factors for venous infarction
or intracranial hemorrhage are as follows: exclusive cortical venous drainage, ectasia of the draining vein, agenesis
of the second and/or third segments of the BVR, thrombosis of the superior orbital vein or SPS, and distal cortical venous drainage.3–5 For direct CCFs, venous cerebral congestion has been reported in 6 cases,4,6,9,14,15 all of
which showed brainstem congestion, including one case
associated with congestion of the cervical spinal cord.6
In 3 cases, as seen in our patient, the SPS was not patent. However, to our knowledge, there is no case in the
literature of BG venous congestion related to flow stagnation of veins drained by the first segment of the BVR and
caused by hypoplasia of its second segment.
The BVR is an anastomotic venous channel generally
65

I. Ract et al.
connecting the cavernous sinus to the great cerebral vein
(the so-called vein of Galen). Its course turns around the
cerebral peduncle and presents a close relationship with
the posterior cerebral artery. Typically, the BVR is divided into 3 segments:8,13 The first segment, also called the
anterior or striate segment, mainly drains the inferior striate veins, deep cerebral vein, insular veins, and olfactory
vein. The second, or peduncular, segment mainly drains
the peduncular vein, inferior ventricular vein, inferior
choroidal vein, and hippocampal veins. The third, or posterior mesencephalic, segment mainly drains the lateral
mesencephalic vein and the posterior thalamic vein.
However, the BVR presents some anatomical variations. Indeed, Suzuki et al.,13 after performing 3D CT angiography in 250 patients, reported that the connection
between the first and second segments of the BVR was
not seen (agenesis or hypoplasia) in 36.9% of the cases.
A simple classification of the potential venous drainages of the BVR has been established according to the
5 drainage pathways formed during the early embryonic
stage.10 Drainage can be toward the vein of Galen, toward
the cavernous sinus or sphenoparietal sinus, toward the
SPS via the lateral mesencephalic vein, toward the SPS
via a peduncular vein, and toward the transverse sinus or
straight sinus via the tentorium sinuses.
If the uncal vein, which connects the cavernous sinus to the first segment of the BVR, is well developed,
the connection between the first and second segments is
usually hypoplastic or agenetic, as observed in our case.
Since the first segment of the BVR drains the inferior striate veins and some hippocampal veins, engorgement of
this segment by the CCF via the uncal vein explains the
signal abnormalities seen on emergency MRI. Finally, the
CCF occlusion in our case resolved engorgement of the
first segment of the BVR and its tributaries, which was
confirmed by the favorable clinical and imaging outcome.

Conclusions

Brainstem and BG venous congestion is a rare but
potentially serious complication of traumatic CCF. The
association of a large tear on the ICA, a developed uncal
vein with a hypoplastic second segment of the BVR, and
a thrombosed or hypoplastic SPS may favor this rare complication of a direct CCF. Neurosurgeons as well as interventional neuroradiologists should be aware of this unusual
presentation of CCF that provokes emergency treatment.
Disclosure
Dr. Sourour is a consultant for Covidien and Penumbra. The
authors report no conflict of interest concerning the materials or
methods used in this study or the findings specified in this paper.
Author contributions to the study and manuscript preparation include the following. Conception and design: Clarençon,
Ract, Drier, Leclercq, Sourour, Gabrieli, Dormont. Acquisition of
data: Clarençon, Ract, Drier, Leclercq, Sourour, Gabrieli, Yger,
Chiras. Analysis and interpretation of data: all authors. Drafting
the article: Clarençon, Ract, Drier, Leclercq, Sourour, Gabrieli,
Nouet, Dormont, Chiras. Critically revising the article: all authors.
Reviewed submitted version of manuscript: all authors. Study supervision: Clarençon.

66

References
1. Barrow DL, Spector RH, Braun IF, Landman JA, Tindall SC,
Tindall GT: Classification and treatment of spontaneous carotid-cavernous sinus fistulas. J Neurosurg 62:248–256, 1985
2. Blanc R, Maia Barros AD, Brugieres P, Méder JF, Gaston A:
[Cavernous sinus dural arteriovenous fistula complicated by
edematous cerebral lesions from venous etiology.] J Neuroradiol 31:220–224, 2004 (Fr)
3. Brown RD Jr, Wiebers DO, Nichols DA: Intracranial dural
arteriovenous fistulae: angiographic predictors of intracranial
hemorrhage and clinical outcome in nonsurgical patients. J
Neurosurg 81:531–538, 1994
4. Bussière M, Lownie SP, Pelz DM, Nicolle D: Direct carotidcavernous fistula causing brainstem venous congestion. J Neuroophthalmol 29:21–25, 2009
5. Cognard C, Gobin YP, Pierot L, Bailly AL, Houdart E, Casasco A, et al: Cerebral dural arteriovenous fistulas: clinical and
angiographic correlation with a revised classification of venous drainage. Radiology 194:671–680, 1995
6. Herrera DA, Vargas SA, Dublin AB: Traumatic carotidcavernous fistula with pontomesencephalic and cervical cord
venous drainage presenting as tetraparesis. J Neuroimaging
21:73–75, 2011
7. Kiyosue H, Hori Y, Okahara M, Tanoue S, Sagara Y, Matsumoto S, et al: Treatment of intracranial dural arteriovenous
fistulas: current strategies based on location and hemodynamics, and alternative techniques of transcatheter embolization.
Radiographics 24:1637–1653, 2004
8. Miyamoto N, Naito I, Takatama S, Shimizu T, Iwai T, Shimaguchi H: Clinical and angiographic characteristics of cavernous sinus dural arteriovenous fistulas manifesting as venous
infarction and/or intracranial hemorrhage. Neuroradiology
51:53–60, 2009
9. Murata H, Kubota T, Murai M, Kanno H, Fujii S, Yamamoto
I: Brainstem congestion caused by direct carotid-cavernous
fistula—case report. Neurol Med Chir (Tokyo) 43:255–258,
2003
10. Padget DH: The cranial venous system in man in reference to
development, adult configuration, and relation to the arteries.
Am J Anat 98:307–355, 1956
11. Ringer AJ, Salud L, Tomsick TA: Carotid cavernous fistulas:
anatomy, classification, and treatment. Neurosurg Clin N Am
16:279–295, viii, 2005
12. Shownkeen H, Bova D, Origitano TC, Petruzzelli GJ, Leonet­
ti JP: Carotid-cavernous fistulas: pathogenesis and routes of
approach to endovascular treatment. Skull Base 11:207–218,
2001
13. Suzuki Y, Ikeda H, Shimadu M, Ikeda Y, Matsumoto K: Variations of the basal vein: identification using three-dimensional CT angiography. AJNR Am J Neuroradiol 22:670–676,
2001
14. Teng MM, Chang T, Pan DH, Chang CN, Huang CI, Guo WY,
et al: Brainstem edema: an unusual complication of carotid
cavernous fistula. AJNR Am J Neuroradiol 12:139–142,
1991
15. Turner DM, Vangilder JC, Mojtahedi S, Pierson EW: Spontaneous intracerebral hematoma in carotid-cavernous fistula.
Report of three cases. J Neurosurg 59:680–686, 1983
Manuscript submitted September 23, 2013.
Accepted January 13, 2014.
Please include this information when citing this paper: published
online February 14, 2014; DOI: 10.3171/2014.1.JNS132016.
Address correspondence to: Frédéric Clarençon, M.D., Department of Neuroradiology, Pitié-Salpêtrière Hospital, 47 Blvd de
l’Hôpital, Paris 75013, France. email: fredclare5@msn.com.

J Neurosurg / Volume 121 / July 2014