Fistula after Foramen Ovale Telemetry

Internal Carotid-Inferior Petrosal Sinus
Fistula Complicating Foramen Ovale
Telemetry: Successful Treatment with
Detachable Coils: Case Report
and Review
Laurence A.G. Marshman, F.R.C.S., Steven Connor, F.R.C.R.,
Charles E. Polkey, M.D.
Departments of Neurosurgery (LAGM, CEP) and Neuroradiology (SC), King’s College Hospital,
Denmark Hill, London, England

OBJECTIVE AND IMPORTANCE: Iatrogenic internal carotid arterycavernous sinus fistula complicating percutaneous foramen ovale (FO)
instrumentation (e.g., retrogasserian rhizotomy for trigeminal neuralgia)
has been reported in only four patients to date. To our knowledge, no
case of fistula has previously been reported either to complicate FO
telemetry or to involve the inferior petrosal sinus (IPS); moreover, most
patients have presented within 48 hours. We present a case of internal
carotid artery-IPS fistula that complicated FO telemetry in which the
clinical syndrome was delayed by 4 weeks.
CLINICAL PRESENTATION: Four weeks after undergoing bilateral FO telemetry during Phase 2 investigations for surgery for epilepsy, a 37-year-old
man suddenly developed a painful bilateral pupil-sparing oculomotor
palsy, poor visual acuity, proptosis, conjunctival suffusion, and an audible
bruit over the right frontotemporal region.
INTERVENTION: Cerebral angiography demonstrated a high-flow arteriovenous fistula between the junction of the petrous and laceral portions of
the right internal carotid artery and right IPS, with rapid filling of both
cavernous sinuses. Successful obliteration was obtained with Guglielmi
detachable coils, followed by complete resolution of the bilateral ocular
abnormality.
CONCLUSION: This is the first reported case of iatrogenic fistula formation
to either involve the IPS or to complicate FO telemetry. In addition,
symptomatology was anomalously delayed. This case highlights the importance of noting FO anatomic asymmetries before FO instrumentation
and of routinely inquiring for “swooshing” noises after electrode
withdrawal. (Neurosurgery 50:209–212, 2002)
Key words: Arteriovenous fistula, Foramen ovale telemetry

W

ieser et al. (17) first introduced
foramen ovale (FO) telemetry in
1985 to aid in the localization and lateralization of seizure onset where nonin-

vasive tests had previously incriminated one or both mesial temporal
lobes. Smooth multicontact electrodes
are inserted via an 18-gauge needle into
Neurosurgery, Vol. 50, No. 1, January 2002

209

the ambient cistern, such that they lie
between the mesial temporal lobe laterally and the brainstem medially (15–17).
Used in this way, FO telemetry represents a compromise between more informative—but more invasive—subdural or intracerebral depth electrode
monitoring and noninvasive—but less
informative—scalp telemetry or neuroimaging studies. In principle, FO telemetry permits definitive surgery to be undertaken against a background of
minimal preoperative investigative risk.
The senior author (CEP) has performed FO telemetry in more than 200
patients during the past 7 years. Complications were rare and concord in frequency with those reported by others
(10, 15, 16). Nevertheless, a unique complication not previously documented
was encountered that we herein report
in detail.

CASE REPORT
First admission
A 37-year-old man was admitted in
August 1998 for neurosurgical assessment of drug-resistant epilepsy. His first
seizure was a tonicoclonic convulsion at
the age of 17 years; since then, the majority of seizures were simple or complex partial in pattern, with only occasional secondary generalization. Attacks
occurred at an average of eight per
month and typically commenced with
an epigastric aura, then lip smacking or
other automatisms, before culminating
in a backward fall and urinary incontinence. Normal consciousness was restored after variable lengths of postictal
confusion. No episodes of status epilepticus had been recorded, and his longest
seizure-free interval had been 2 years;
however, a steady increase in seizure
frequency had occurred in recent years.
Therapy with a variety of anticonvulsants proved ineffective; at the time of
his investigation, he was taking lamotrigine and topiramate.
Magnetic resonance imaging scans
demonstrated decreased volume and
high T2 signal in the right hippocampus
consistent with mesial temporal sclero-

210

Marshman et al.

sis. Positron emission tomography with
[18F]fluorodeoxyglucose demonstrated
a hypometabolic zone in the right anterior temporal lobe, a finding also consistent with right mesial temporal sclerosis. Scalp telemetry, however, revealed
bilateral sharp waves in both anterior
temporal regions, a finding inconsistent
with a unilateral focus. Moreover, the
results of neuropsychological tests suggested that language dominance lateralized to the right hemisphere. Bilateral
FO telemetry was therefore undertaken
to aid lateralization of the primary
focus.
Bilateral six-contact Teflon-insulated
FO electrodes (0.11 mm in diameter)
were implanted under fluoroscopic control while the patient was under general
anesthesia. The electrodes were placed
into each ambient cistern through an
18-gauge needle as previously described (3, 15–17). No complication was
encountered during this procedure; in
particular, no excessive hemorrhage
was apparent. The patient awoke from
the anesthetic without obvious neurological change, and his anticonvulsant
medication was thereafter tapered. Postoperative cranial x-rays confirmed adequate electrode positioning. Subsequent
continuous 32-channel video electroencephalogram telemetry, performed between August 11 and August 20, 1998,
recorded five complex partial seizures,
each consistent with a right temporal
lobe focus.
Both electrodes were subsequently removed without event while the patient
was under general anesthesia, and he
was discharged home 2 days later. In
retrospect, however, a “swooshing”
noise had been subjectively audible immediately after electrode removal. No
dramatic cardiovascular fluctuation was
found retrospectively from further
study of his observation charts at either
FO needling or at the time of electrode
withdrawal.

Second admission
Four weeks after discharge, the patient suddenly reported ptosis and diplopia in the right eye, although visual
acuity remained preserved. Apart from
mild retro-orbital pain and tinnitus, no
other abnormality was noted. Examina-

tion disclosed a partial, pupil-sparing,
right oculomotor palsy that rapidly progressed within days to complete, bilateral, pupil-sparing oculomotor paralysis
with bilaterally complete ptosis. Both
eyes were red and proptosed, and a soft
systolic bruit was audible over the right
frontotemporal region. No other neurological abnormality was noted. A clinical diagnosis of caroticocavernous sinus
fistula was made, and the patient was
admitted for further investigation.
Cerebral angiography demonstrated
a high-flow arteriovenous fistula between the junction of the petrous and
laceral portions of the right internal carotid artery (ICA) and the right inferior
petrosal sinus (IPS) (Fig. 1). There was
rapid filling of both the cavernous sinus
(CVS) and inferior ophthalmic veins;
however, no cortical venous reflux was
demonstrated. Guglielmi detachable
coiling (GDC; Target Therapeutics, Fremont, CA) for embolization of the arteriovenous fistula was thereafter immediately instituted via the femoral artery.
After a 6-French guide catheter was positioned within the proximal right ICA,
a microcatheter was advanced through
the fistulous communication into the
right CVS. Through this microcatheter,
both the CVS and IPS were packed with
six GDC coils in retrograde fashion. The
subsequent embolization dramatically
reduced fistulous flow, with only minimal residual IPS filling (Fig. 2). After the
procedure, a rapid and complete resolution of the bilateral ophthalmoplegia
was obtained.

Third admission
The excellent clinical result remained
unchanged at 16 months’ review,
whereupon, at a third admission, a right

FIGURE 2. Posttreatment (GDC) right
ICA angiogram, frontal projection. Only
minor residual flow from the fistula
into the right IPS (arrowhead) is now
apparent. The GDC ball extends from
the right CVS and into the right IPS.
temporal lobectomy was successfully
performed, resulting in complete cessation of seizure activity. The patient experienced no further complications. The
resected specimen revealed features
typical of mesial temporal sclerosis.

DISCUSSION
FO telemetry is typically free of complications. Minor symptoms commonly
reported relate to transient ipsilateral trigeminal nerve damage. These include hypoesthesia or dysesthesia at one corner of
the mouth, brief contractions of the ipsilateral masseter (16), or lancinating ipsilateral dental pain after electrode withdrawal (10, 16). Occasionally, more
serious and permanent trigeminal damage has been reported (including, anecdotally, trigeminal avulsion) (10, 15–17).
Such complications probably relate to
multiple insertion attempts (the practice
of CEP is to limit such attempts to two).

FIGURE 1.
Pretreatment right
common carotid
angiograms. A,
frontal projection.
B, right anterior
oblique projection
(early arterial
phase). A highflow fistula is present between the junction of the petrous and lateral portions of
the right ICA and the cranial portion of the right IPS (arrowhead). Rapid shunting
is apparent into both CVSs and ophthalmic veins.
Neurosurgery, Vol. 50, No. 1, January 2002

Fistula after Foramen Ovale Telemetry
More rarely, oculomotor, trochlear, or
even abducens nerve damage has been
reported, as has meningitis (4, 7, 10, 15,
16).
Significant intracranial hemorrhage
(or ischemia) has also been reported
with FO instrumentation, including FO
telemetry (1, 9, 10, 12, 13, 15, 16). Furthermore, four cases of iatrogenic ICACVS fistula have also complicated percutaneous FO instrumentation (5, 6, 11)
(the case of Revuelta et al. [8] strictly
involved the external carotid artery).
Frank operative hemorrhage, then a
postoperative swooshing noise, typified
these fistulae. No fistulae, however,
have directly involved the IPS (Bartlow
and Penn’s [2] traumatic case indirectly
involved the superior petrosal sinus via
the CVS), and none has complicated FO
telemetry. Our case therefore seems to
be the first iatrogenic fistula either to
directly involve the IPS or to complicate
FO telemetry.
The FO and carotid canal are intimately related at the cranial base in the
middle cranial fossa. Anatomic variants—such as a foramen of Vesalius or
Arnold (14) or a primitive foramen
lacerum medius (11) (a defect in the foramen lacerum allowing communication with the FO)—predispose the patient to ICA needle puncture during FO
instrumentation. Although none of
these variants was present in the patient
we studied, a definite asymmetry in FO
contour was apparent, with the right FO
possessing a considerably thinner posterior wall than that of the left (Fig. 3).
This could have permitted ICA (and
subsequently rostral IPS) penetration.
FO asymmetry should therefore always
be borne in mind before contemplating
percutaneous instrumentation.
An unusual feature of the clinical presentation of this patient was the 4-week
delay in ocular symptomatology after
electrode removal. Most ICA-CVS fistulae
complicating FO instrumentation have
presented within 2 days of the causative
procedure (6, 11). Gokalp et al. (5), however, reported a case that presented 12
days after percutaneous trigeminal rhizotomy, whereas Bartlow and Penn (2) reported a traumatic fistula that developed
12 years after an automobile accident (although a swooshing noise had been subjectively audible transiently closer to the

211

named authors and was not supported
by any external funds. No conflict of
interest is present.
Received, December 13, 2000.
Accepted, August 24, 2001.
Reprint requests: Laurence A.G. Marshman,
F.R.C.S., Department of Neurosurgery, King’s
College Hospital, Denmark Hill, London SE5
9RS, England.
Email: l.a.g.marshman@btinternet.com

FIGURE 3. Fine section cranial base
computed tomographic scan (bone windows). The right FO is more slender
than the left and possesses a much thinner posterior wall. Such asymmetrical
anatomy could predispose to ICA penetration during FO instrumentation.
time of injury). The anomalous delay in
such cases could be explained by initial
pseudoaneurysm formation before fistula
formation. Alternatively, in the case of the
fistula we studied, the apparent delay
could have related to gradual IPS thrombosis that, with complete IPS occlusion,
resulted in fistulous flow being redirected
anteriorly into both CVS and ophthalmic
veins. Such a mechanism is suggested by
poor opacification of the ipsilateral IPS on
diagnostic angiography (Fig. 1).
FO telemetry successfully lateralized
the primary epileptogenic focus in the
patient we studied. In consequence, a
definitive right temporal lobectomy was
indicated—and
successfully
performed—with complete resolution of
seizure activity. The need for more invasive subdural (or intracerebral depth
electrode) telemetry was therefore obviated. Nevertheless, a serious complication occurred in a procedure ordinarily
considered only moderately invasive.
Its anticipation and early investigation
are emphasized—especially in instances
in which patients have reported a subjective swooshing noise after FO instrumentation (11). In addition, anatomic
asymmetry in FO contour should always be borne in mind whenever FO
instrumentation is being contemplated.

ACKNOWLEDGMENTS
We thank Devi Amin, Department of
Neurophysiology, for her kind help in
the production of the article. This article
was significantly contributed to by all
Neurosurgery, Vol. 50, No. 1, January 2002

REFERENCES
1. Apfelbaum RI: Subarachnoid hemorrhage and
“normal pressure hydrocephalus”: Fatal complication of percutaneous microcompression of
the Gasserian ganglion—Case report. Neurosurgery 22:150–151, 1988 (comment).
2. Bartlow B, Penn RD: Carotid-cavernous fistula
presenting as a posterior fossa mass: Case report. J Neurosurg 42:585–588, 1975.
3. Fernandez Torre JL, Alarcon G, Binnie CD,
Polkey CE: Comparison of sphenoidal, foramen
ovale and anterior temporal placements for detecting interictal epileptiform discharges in
presurgical assessment for temporal lobe epilepsy. Clin Neurophysiol 110:895–904, 1999.
4. Fujita Y, Nakazaki K: Complications of alcohol
injection into the mandibular nerve: Gasserian
palsy and oculomotor palsy [in German]. Reg
Anaesth 5:39–41, 1982.
5. Gokalp HZ, Kanpolat Y, Turner B: Carotidcavernous fistula following percutaneous trigeminal ganglion approach. Clin Neurol
Neurosurg 82:269–272, 1980.
6. Kuether TA, O’Neill OR, Nesbit GM, Barnwell SL:
Direct carotid cavernous fistula after trigeminal
balloon microcompression gangliolysis: Case report. Neurosurgery 39:853–856, 1996.
7. Markakis E: Transitory fourth cranial nerve
palsy due to foramen ovale electrode placement. Acta Neurochir (Wien) 139:789–790,
1997.
8. Revuelta R, Nathal E, Balderrama J, Tello A,
Zenteno M: External carotid artery fistula due
to compression of the gasserian ganglion for
relief of trigeminal neuralgia: Case report.
J Neurosurg 78:499–500, 1993.
9. Rish BL: Cerebrovascular accident after
percutaneous radiofrequency thermocoagulation of the trigeminal ganglion: Case report.
J Neurosurg 44:376–377, 1976.
10. Schüler P, Stefan H, Neubauer U, Schulemann H:
Vergleich von foramen ovale und subduralen
streifenelektroden zur präoperativen epilepsiediagnostik. Presented at the 36th Annual Meeting
German EEG-Society, 10–13 Celle, 1991, p 152
(abstr).
11. Sekhar LN, Heros RC, Kerber CW: Carotidcavernous fistula following percutaneous retrogasserian procedures: Report of two cases.
J Neurosurg 51:700–706, 1979.
12. Spaziante R, Cappabianca P, Peca C, de Divitiis
E: Subarachnoid hemorrhage and “normal pressure hydrocephalus”: Fatal complication of percutaneous microcompression of the Gasserian
ganglion—Case report. Neurosurgery 22:148–
151, 1988.

212

Marshman et al.

13. Sweet WH: Dangerous rises in blood pressure
upon heating of trigeminal rootlets: Increased
bleeding times in patients with trigeminal neuralgia. Neurosurgery 17:843–844, 1985.
14. Sweet WH: Treatment of trigeminal neuralgia
by percutaneous rhizotomy, in Youmans JR
(ed): Neurological Surgery: A Comprehensive Reference Guide to the Diagnosis and Management of
Neurological Problems. Philadelphia, W.B.
Saunders Co., 1990, vol 6, ed 3, pp 3888–3921.
15. Wieser HG: Semi-invasive EEG: Foramen ovale
electrodes, in Lüders H (ed): Epilepsy Surgery.
New York, Raven Press, 1991, pp 361–370.
16. Wieser HG, Hajek M: Foramen ovale and peg
electrodes. Acta Neurol Scand Suppl 152:33–
35, 1994.
17. Wieser HG, Elger CE, Stodiek SRG: The foramen ovale electrode: A new recording method
for the pre-operative evaluation of patients suffering from mesio-basal temporal lobe epilepsy. Electroencephalogr Clin Neurophysiol
661:314–322, 1985.

COMMENTS
The authors report an unusual case
of iatrogenic carotid-cavernous fistula
caused by an injury to the carotid artery
induced by foramen ovale (FO) electrodes. As the authors point out, the
intimate relationship between the FO
and the carotid canal creates the potential for this complication with any instrumentation involving the FO. An iatrogenic carotid-cavernous fistula has
been reported as a complication of percutaneous procedures for the treatment
of trigeminal neuralgia, but this is the
first report of this complication occurring from FO telemetry during an evaluation for epilepsy. The authors note
that asymmetry of the patient’s FO may
have contributed to this complication
because the foramen on the side ipsilateral to the injury was more slender and

had a thin posterior wall. Such asymmetry or other anatomic variants should be
analyzed before placement of FO instrumentation to reduce the likelihood of
this complication.
I suspect that the delayed onset of this
patient’s symptoms were due to gradual
thrombosis of the inferior petrosal sinus
and an alteration in the venous drainage
of the fistula, which is the primary determinant of symptoms in patients with
carotid cavernous fistulae.
Daniel L. Barrow
Atlanta, Georgia
The authors report an interesting and
rare complication of FO telemetry. An
internal carotid-inferior petrosal sinus
fistula developed after the FO electrodes
were removed. The fistula was successfully treated with coil embolization. The
authors stress the need for careful preoperative evaluation of anatomic asymmetry in FO and the importance of the
patient’s subjective complaints, such as
the presence of a “swooshing” noise after removal of the electrodes.
Demetrius K. Lopes
Chicago, Illinois
L. Nelson Hopkins
Buffalo, New York
In this article, Marshman et al. describe a single case of an arteriovenous
fistula occurring after FO electrodes
were placed for evaluation of epilepsy.
A few cases are described in the literature regarding such fistulae after trigeminal neuralgia procedures, but the
patient that the authors describe seems

to have some other features that may
have been a complication of the electrodes—and may even have come
from the act of removing the electrodes. The fistula involved the inferior
petrosal sinus and seemed to drain
secondarily into the cavernous sinus;
the patient presented 4 weeks after the
instrumentation.
This is a good description of a rare
complication of a common procedure.
The authors suggest that anatomic variants in the shape of the FO may play
a role in this predilection. They do
not provide much advice as to how to
avoid this complication if an abnormal
FO is identified. The treatment with
Guglielmi detachable coiling was successful, with only minimal residual. It is
not clear why the authors took 16
months after this procedure to proceed
with definitive treatment of the epilepsy, but the presence of the fistula did
not seem to affect the successful epilepsy treatment.
Probably the most pertinent information in this study is the suggestion that
clinical symptomatology may be delayed in appearance. This does not necessarily mean that the injury may not
present immediately, but pseudoaneurysm formation or delayed thrombosis
of an earlier fistula may predispose the
patient to delayed presentation. The article would have been improved by including a discussion of what steps may
be taken to avoid this complication if an
anatomic variant is recognized.
Robert J. Dempsey
Madison, Wisconsin

Congress of Neurological Surgeons’
Mission Statement
“The Congress of Neurological Surgeons exists for the purpose of promoting the public
welfare through the advancement of neurosurgery, by a commitment to excellence in
education, and by dedication to research and scientific knowledge. The Congress of
Neurological Surgeons maintains the vitality of our learned profession through the
altruistic volunteer efforts of our members and the development of leadership in service
to the public, to our colleagues in other disciplines, and to the special needs of our fellow
neurosurgeons throughout the world and at every stage of their professional lives.”

Neurosurgery, Vol. 50, No. 1, January 2002