CLINICAL STUDIES

Stereotactic Radiosurgery and Particulate Embolization for
Cavernous Sinus Dural Arteriovenous Fistulae
Bruce E. Pollock, M.D., Douglas A. Nichols, M.D.,
James A. Garrity, M.D., Deborah A. Gorman, R.N.,
Scott L. Stafford, M.D.
Departments of Neurological Surgery (BEP, DAG), Diagnostic Radiology
(DAN), Ophthalmology (JAG), and Radiation Oncology (SLS),
Mayo Clinic and Foundation, Rochester, Minnesota

OBJECTIVE: To evaluate the safety and efficacy of stereotactic radiosurgery, either with or without transarterial
embolization, in the treatment of patients with dural arteriovenous fistulae (DAVFs) of the cavernous sinus.
METHODS: W e reviewed the findings, from a prospectively established database, for 20 patients with cavernous
sinus DAVFs who were treated with either radiosurgery alone (n = 7) or radiosurgery and transarterial
*
embolization (n = 13) in a 7-year period. The median follow-up period after radiosurgery was 36 months (range,
4-59 mo).
‘ RESULTS: Nineteen of 20 patients (9 5 % ) experienced improvement of their clinical symptoms. Fourteen of 15
;aj
patients (93%) experienced either total (n = 13) or nearly total (n = 1) obliteration of their DAVFs, as
documented by angiography performed a median of 12 months after radiosurgery. No patient experienced a
ilk
recurrence of symptoms after angiography showed DAVF obliteration. Two patients developed new neurological
|
deficits after embolization procedures. One patient exhibited temporary aphasia secondary to a venous infarc­
tion; another patient exhibited permanent Vlth cranial nerve weakness related to acute cavernous sinus throm­
bosis. Two patients experienced recurrent symptoms and underwent repeat transarterial embolization at 7 and
12 months; both patients achieved clinical and angiographic cures (5 and 10 mo later, respectively). One patient
experienced recurrent visual symptoms and underwent transvenous embolization 4 months after radiosurgery.
CONCLUSION: Staged radiosurgery and transarterial embolization provided both rapid symptom relief and long­
term cures for patients with cavernous sinus DAVFs. Radiosurgery alone was effective for patients with DAVFs
whose arterial supply was not accessible via a transarterial approach, although the time course of symptom
improvement was longer, compared with patients who also underwent embolization.

J

(Neurosurgery 45:459-467, 1999)
Keywords: Cavernous sinus, Dural arteriovenous fistula, Embolization, Gamma knife, Stereotactic radiosurgery

fis tu la e (D A V F s )
a v e r n o u s s in u s d u r a l ^ ' ^ ° m u n ic a .r o n s b e t w e e n
a re a b n o rm a l a rte n o v e n o u s
(|O V ian d
•
b ra n c h e s o f th e in te r n a l c a r o h
s in u s . B a rro w e
m l c a r o t id a r t e r y ( E C A ) a n d
u s s in u s fis tu la e
a l. ( 3 ) c l a s s i f i e d c a r o t i d a r t e r y a rte n al " J p f
C

fo u r a n g i o g r a p h i c t y p e s , o n * ^
f^
e e n th e IC A a n d th e
T y p e A fis tu la e a r e d ir e c t s h u n t s b e w ^
d aV F
Type A
ca v ern o u s s in u s a n d d o n o t r e P
u s u a lly d e v e o p
^
fis tu la e a r e h i g h - f l o w
tr e a ,m e n t b ecau se o _ P ^
tr a u m a a n d o f t e n r e q u i r e
g deterioration. T y p e
' '
g r e s s iv e v i s u a l o r n e u r o l o g
b r a n c h e s o f th e
D rep resen t s h u n ts b e tw e e n u r

B), dural branches of the EC A (Type C), or dural branches of
both the IC A and EC A (Type D) and the cavernous sinus.
These fistulae arise spontaneously and generally affect
middle-aged women. Cavernous sinus D A V Fs occur less fre­
quently than transverse/sigmoid sinus D A V Fs (1, 4, 6) and
have a low risk of hemorrhage, unless cortical venous drain­
age is present (1). Patients with cavernous sinus D A V Fs may
develop an alarming symptom complex consisting of propto­
sis, chemosis, extraocular muscle palsies, and, sometimes,
threatened visual loss resulting from either increased intraoc­
ular pressure or reduced ocular perfusion pressure. For these
patients, rapid curative treatment is necessary to prevent vi-

Neurosurgery, Vol. 45, No. 3, Septem ber 1999

Downloaded from https://academic.oup.com/neurosurgery/article-abstract/45/3/459/2838461
by Tufts University user
on 17 March 2018

459

460

Pollock et al.

sual deterioration, in addition to cosmetic and other sym p­
tomatic considerations.
Management of cavernous sinus D A V Fs has included ob­
servation (3), intermittent manual compression of the IC A
(16), microsurgery (9, 13, 23, 29), transarterial or transvenous
embolization (3, 10, 12, 14, 16, 21, 25-27, 30, 31), and stereo­
tactic radiosurgery (2, 6, 11, 19). Endovascular techniques
have become the most common treatment for patients with
symptomatic cavernous sinus D A V Fs that do not spontane­
ously resolve, because of the high rates of cure after emboli­
zation (20). However, embolization of cavernous sinus
D A VFs has been reported to cause at least temporary oph­
thalmoplegia in 8 to 50% of patients (16, 21, 25, 31), and
recanalization of embolized D A V Fs is well documented (7,12,
16). Direct surgical resection has been advocated for patients
who experience failure of endovascular attempts at oblitera­
tion (9, 29), but such surgery is difficult and is associated with
high rates of postoperative cranial nerve deficits. Also, major
morbidity related to occlusion of the IC A has been reported
after direct surgical treatment of cavernous sinus D A V Fs (9).
W e previously reported our experience with 29 patients,
using a staged approach to intracranial D A V F treatment that
involved radiosurgery followed by planned transarterial em­
bolization (19). This staging allows complete radiosurgical
coverage of the involved dural sinus, for progressive long­
term obliteration of all abnormal fistulous connections. Sub­
sequent embolization provides rapid symptom relief and/or
elimination of cortical venous drainage. This report outlines
the results of this staged approach for 20 patients with sym p­
tomatic Type B, C, or D cavernous sinus D A V Fs who were
treated at our center during a 7-year period.

PATIENTS AND METHODS
Patient population
Between March 1992 and October 1997, 20 patients with
cavernous sinus D A V Fs underwent radiosurgery, using a
Leksell Gamma Unit (Elekta Instruments, Atlanta, G A ), at the
M ayo Clinic (Rochester, M N ). A ll clinical, neuroimaging, and
follow-up information for the patients was entered into a
prospectively maintained computer database. Our patients
included 17 women and 3 men. The clinical and angiographic
characteristics of the patients are presented in Table 1. The
median patient age was 67 years (range, 34-80 yr). A ll patients
exhibited symptoms resulting from the fistulae and had ex­
perienced failure of conservative management. No patient
had experienced intracranial hemorrhage before the proce­
dure. For four patients, cortical venous drainage was noted
during preoperative angiography.

Patient evaluation
A ll patients underwent a thorough ophthalmological exam­
ination, by a neuro-ophthalmologist, before radiosurgery. Ob­
jective tests of visual acuity and visual fields were performed
to provide baseline data for comparison during follow-up
monitoring. Because of the complexity of the D A V F cases
referred to our institution, the available imaging results were

TABLE 1. Clinical and Angiographic Characteristics of
20 Patients with Cavernous Sinus Dural Arteriovenous
Fistulae before Radiosurgery
No. of Patients (%)
Characteristic

Radiosurgery
Alone
(n = 7)

Radiosurgery and
Embolization
(n = 13)

Chemosis/proptosis
Diplopia
Decreased vision
Tinnitus
Headache/eye pain
Prior embolization
Type of D AVFa (Barrow)''
B
C
D
Type of DAVF (Cognard)r
I
I la
I la + b
III

6 (86)
4 (57)
2 (29)
2 (29)
2 (29)
0(0)

12 (92)
9 (69)
6(46)
5 (38)
4(31)
2 (15)

1 (14)
1 (14)
5 (71)

0(0)
2 (15)
11 (85)

1 (14)
6 (86)
0(0)
0(0)

0(0)
9 (69)
3 (23)rf
1 (8)

a DAVF, dural arteriovenous fistula.
b According to the classification system of Barrow et al. (3).
c According to the classification system of Cognard et al. (7).
d One patient underwent preradiosurgical embolization, with elim­
ination of cortical venous drainage, before radiosurgery.

reviewed by an interventional neuroradiologist before man­
agement decisions were made. Bilateral selective ICA, ECA,
and vertebral artery angiographic series were essential for
diagnosis and identification of all arterial supplies to the
fistulae. Some patients required superselective b r a n c h - v e s s e l
angiography to precisely establish the exact site of a f i s t u l a . In
fact, for several patients we thought it was necessary to repeat
angiography at our institution before decisions could b e m ade
regarding management. The final decision to proceed with
radiosurgery and transarterial embolization was made only
after thorough analysis of the clinical and imaging f i n d i n g s in
each case.

Radiosurgical dosimetry
Conventional biplanar angiography, using high-resolution
magnification and photographic subtraction techniques, was
performed in all cases. Selective IC A , EC A , and/or v e r te b r a l
artery injections were performed as required to d e m o n s t r a t e
the target for dose-planning. Superselective a n g i o g r a p h y of
EC A branches was also performed when necessary, in un­
usual cases. The Leksell Gamma Plan System (used after earh
1993) permits scanning of the angiographic films, together
with stereotactic magnetic resonance imaging results (via di­
rect network data transfer), into the dose-planning c o m p u t e r The ability to integrate the stereotactic angiographic and mag'
netic resonance imaging results allows conformal doseplanning for the fistula and maintains the ability to o b s e n e
the adjacent temporal lobe, pituitary gland, and optic a p p a'

Neurosurgery, Vol. 45, No. 3, September 1999

Downloaded from https://academic.oup.com/neurosurgery/article-abstract/45/3/459/2838461
by Tufts University user
on 17 March 2018

Dural Arteriovenous Fistula Radiosurgery

3:

ratus. The goal was to include the entire area of the involved
cavernous sinus in the radiosurgical dose plan.
A median of three isocenters of radiation (range, 1-10 iso­
centers) were used to cover the D A V Fs of the patients in this
series. All patients were treated at the 50% or greater isodose
line. Because the locations of the D A V Fs were usually in the
inferior or posterior part of the cavernous sinus, the radiation
dose to the adjacent optic apparatus was less than 10 G y in all
cases {Fig. 2). The median prescription isodose volum e was 2.8
cm3(range, 0.7-7.5 cm 3). The median marginal radiation dose
was 20 Gy (range, 18-20 G y). The median maximal radiation
dose was 40 G y (range, 22.2-40 G y).

:
d

9'
3!

Transarterial embolization technique
Because radiosurgery involves a latency interval before
obliteration of the fistula, the aim of combining radiosurgery
with transarterial embolization was to provide rapid symp­
tom relief and to eliminate cortical venous drainage, if
present. Two patients underwent embolization procedures
before radiosurgery. One patient remained symptomatic de­
spite both transvenous and transarterial embolizations and
was referred for further evaluation and care. She was treated
with radiosurgery and repeat transarterial embolization at
our center. Another patient underwent preradiosurgical em­
bolization. This patient had a bilateral cavernous sinus D A V F

461

with such extensive cortical venous drainage that adequate
angiographic imaging of the fistula itself was impossible w ith­
out reduction of the blood flow through the fistula.
Particulate embolization was used as a planned adjunct
after radiosurgery for 12 patients, in one (n = 10) or two (n = 2)
sessions. Seven patients underwent radiosurgery alone because
transarterial embolization was thought not to be technically pos­
sible with acceptable risk. Polyvinyl alcohol particles were the
primary embolization material used in the embolization proce­
dures. In some cases, other materials (silk sutures, Gelfoam, or
fibered microcoils) were used in conjunction with the polyvinyl
alcohol particles. Cortical venous drainage was eliminated in all
four patients with this angiographic finding, using either pre- or
postradiosurgical embolization procedures. Most patients were
monitored in the hospital overnight after their radiosurgical and
embolization procedures.

Follow-up monitoring
Yearly clinical and angiographic assessments were at­
tempted for all patients. Routine follow-up magnetic reso­
nance imaging evaluations were not performed unless the
patients developed symptoms thought to be related to radia­
tion injury of the adjacent brainstem or temporal lobe, because
magnetic resonance imaging results are mostly noninformative for this patient group. Additional treatment was consid-

7)

it

e

-<£' ■£

i)

e
'C

li

8
3
I
'fI
I
I

n
a

u

)1

N e u ro su rg e ry , Vol.

Downloaded from https://academic.oup.com/neurosurgery/article-abstract/45/3/459/2838461
by Tufts University user
on 17 March 2018

45,No. 3, September 7

FIG URE 1.
Radiosurgical dose
plan for a 51-year-old
woman with progres­
sive eye pain and dip­
lopia related to a left
cavernous sinus
DAVF. Five 8-mm iso­
centers of radiation
were used to treat the
fistula (volume, 2.3
cm3) with a marginal
dose of 20 Gy (maxi­
mal dose, 40 Gy).
Note that the optic
nerves and chiasm
were outside the 20 %
isodose line and thus
received less than 8
Gy of radiation.

462

Pollock et al

ered if the patients remained symptomatic after the initial
procedures. Clinical follow-up data were available for all
patients for a median of 36 months after radiosurgery (range,
4-59 mo).

Statistical analyses
Nonparametric variables were analyzed using Fisher's ex­
act test; parametric variables were compared using Student's
t test.

RESULTS
Patient symptoms
Ocular symptoms related to venous congestion (chemosis
or proptosis) improved significantly (n = 2) or resolved com­
pletely (n = 15) for 17 of 18 affected patients (94%). Seven of
eight patients (88%) with decreased visual fields or acuity
regained normal vision. One patient experienced decreased
vision 11 months after the procedure but was found to have
choroidal detachment with secondary-angle glaucoma. Ten of
13 patients (77%) with preoperative diplopia experienced ei­
ther normalization (n = 8) or significant improvements (n =
2). Three patients with preoperative Vlth cranial nerve weak­
ness exhibited no change in diplopia during follow-up mon­
itoring. Tinnitus (n = 7) and eye pain/headache (n = 6) were
resolved for all patients. Overall, improvement of some symp­
toms was experienced by 19 of 20 patients (95%). No differ­
ence in symptom improvement was noted between patients
treated using radiosurgery alone (100%) and patients treated
using radiosurgery and transarterial embolization (92%) (P =
1.0). However, the median time to patient improvement was
shorter for the patients who underwent both radiosurgery
and embolization, compared with the patients who under­
went only radiosurgery (2 w k versus 6 mo, P < 0.001). No
patient experienced intracranial hemorrhage during the
follow-up period.

Complications and additional treatments
Complications related to either stereotactic angiography
(n = 1) or the embolization procedures (n = 2) occurred in
three patients (15%). One patient became lethargic, with left
hemiparesis, during stereotactic angiography; embolic occlu­
sion of the M l segment of the right m iddle cerebral artery was
discovered (Fig. 2). The patient underwent immediate intraarterial thrombolytic therapy, and the neurological status of
the patient returned to baseline. Radiosurgery and two em­
bolization procedures were completed in the next several
days. Fourteen months later, the chemosis and diplopia of the
patient have resolved and she remains neurologically intact.
One patient developed new Vlth cranial nerve paresis during
preradiosurgical embolization, as a result of acute cavernous
sinus thrombosis. Fifty-four months later, her diplopia has
improved but she still requires prism glasses. One patient
developed a small venous infarction during postradiosurgical
embolization, resulting in transient aphasia, which improved
in the months after the procedure.
Three patients who initially showed improvement after
radiosurgery and transarterial embolization underwent addi­
tional embolization procedures because of recurrent symp­
toms. Two patients underwent repeat transarterial particulate
embolization procedures at our center, 7 and 12 months after
the initial procedures. The symptoms of both patients again
improved, and the patients demonstrated complete oblitera­
tion of their D A V Fs in later follow-up angiograms. One pa­
tient exhibited increasing intraocular pressure and worsening
of her vision, as noted 4 months after staged radiosurgery and
embolization. The patient underwent transvenous coil occlu­
sion of the cavernous sinus via a transfemoral and superior
ophthalmic vein approach, at another center, with rapid im­
provement in her chemosis and ocular pressure. The visual
acuity of the patient has returned to normal, but she continues
to exhibit some diplopia. One patient underwent medial rec-

Neuroimaging outcomes
Thirteen of 15 patients (87%) experienced complete obliter­
ation of their D AVFs, as demonstrated in follow-up angio­
grams obtained a median of 12 months after radiosurgery
(range, 4-33 mo). One patient with a Type D fistula (3) expe­
rienced complete resolution of her chemosis/proptosis and
exhibited improved vision 13 months after radiosurgery. A n ­
giography performed at that time showed that the E C A sup­
ply was gone but a small IC A supply to the fistula persisted
(more than 95% obliteration). The patient remains asymptom­
atic 33 months after radiosurgery and refuses to undergo
further angiography. One patient exhibited recurrent visual
symptoms 4 months after staged radiosurgery and emboliza­
tion. Angiography performed at the time of transvenous em­
bolization revealed recanalization of her D A V F. The remain­
ing five patients have refused follow-up angiography but are
clinically cured.

FIG URE 2. Right anteroposterior carotid artery angiograms
for a 71-year-old woman with chemosis, tinnitus, and diplo­
pia resulting from bilateral cavernous sinus DAVFs. A, ICA
injection before radiosurgery, showing embolic occlusion ot
the M1 segment. The patient underwent immediate intraarterial thrombolytic therapy, and her neurological status
returned to baseline. B, ICA injection during embolization
3 days after radiosurgery, showing restoration of vessel
patency.

Neurosurgery, Vol. 45, No. 3, September 1999

Downloaded from https://academic.oup.com/neurosurgery/article-abstract/45/3/459/2838461
by Tufts University user
on 17 March 2018

Dural Arteriovenous Fistula Radiosurgery
tus muscle resection 34 months after radiosurgery, to improve
persistent diplopia.

463

B

ILLUSTRATIVE c a s e s
Patient 1
A 67-year-old man presented with chemosis, proptosis, diplopia,
and eye pain resulting from a left cavernous sinus DAVF {Fig. 3).
The Type D fistula received arterial supply from both the right and
left ICA and ECA (3). Venous outflow was retrograde through the left
superior ophthalmic vein, with no cortical venous drainage (Type IIa)
(7). The patient underwent radiosurgery alone because it was not
thought that significant reduction of the arteriovenous shunting could
be accomplished by transfemoral embolization with acceptable risk.
Two months after radiosurgery, the pain, chemosis, and diplopia of
the patient had improved. One year later, the examination results
were normal, and angiography showed that the DAVF had been
completely obliterated.

j

/

Patient 2
A 34-year-old woman presented with chemosis, proptosis, diplo­
pia, and eye pain resulting from a left cavernous sinus DAVF {Fig. 4).
Despite previous transarterial and transvenous embolization proce­
dures performed at another center, the patient remained sym ptom ­
atic. The arterial supply was from the left ECA and bilateral ICAs
(Type D) (3). Venous outflow was through the left superior ophthal­
mic vein, without cortical venous drainage (Type Ila) (7). The patient
underwent radiosurgery, followed by transarterial embolization 1

B
}

c

r

\

D

FIGURE 3. Patient 1. Right carotid artery angiograms for a
67-year-old man with chemosis/proptosis, diplopia, and eye
pain resulting from a left cavernous sinus DAVF. The fistula
also received arterial supply from the left ICA and ECA.
4 and B, lateral (A) and anteroposterior {B) ECA injections
Performed at the time of radiosurgery. C and D , lateral (C)
and anteroposterior (D) common carotid artery injections
Performed 14 months after radiosurgery, showing complete
Alteration of the DAVF.

FIGURE 4. Patient 2. Left lateral carotid artery angiograms
for a 34-year-old woman with proptosis/chemosis, diplopia,
and eye pain resulting from a left cavernous sinus DAVF. The
patient had undergone an unsuccessful attempt at trans­
venous embolization 1 month earlier, at another institution.
A, common carotid injection performed at the time of radio­
surgery. Bf ICA injection performed 11 months after radio­
surgery and transarterial embolization, showing persistence
of the DAVF. The patient underwent repeat embolization of
the residual left ECA supply to the DAVF. C, ICA injection
performed 22 months after radiosurgery, showing complete
obliteration of the DAVF.

day later. The patient showed initial improvement but remaine<
symptomatic 12 months later. Repeat angiography showed oblitera
tion of the right ICA supply to the fistula and reduction of the left I O
and ECA supply. The patient underwent repeat embolization of th
remaining left ECA arterial supply. At 20 months after radiosurger\
the patient experienced deterioration in her visual fields and acuit\
and angiography showed that the arterial supply was again de
creased; however, the patient exhibited nearly complete thrombosi
of her inferior and superior ophthalmic veins and significant venou
congestion of her entire orbit. The patient began to receive aspirii
treatment and subcutaneous low-molecular weight heparin injec
tions. After 2 months of antiplatelet and anticoagulation therapy, th
redness and swelling in her left eye were much improved, and he
visual acuity and fields had returned to normal. Repeat radiosurger
was planned; however, stereotactic angiography showed that th
DAVF had been completely obliterated, and the head-frame wa
removed. Because the patient continued to exhibit partial thrombosi
of her ophthalmic veins, she is still undergoing antiplatelet therapy

N e u ro su rg e ry , Vol. 45, No. 3, September 1999

Downloaded from https://academic.oup.com/neurosurgery/article-abstract/45/3/459/2838461
by Tufts University user
on 17 March 2018

464

Pollock et al

DISCUSSION
Although the clinical presentation of patients with cavern­
ous sinus D A V Fs is often dramatic, the risk of intracranial
hemorrhage for such patients is low (1, 3, 4, 6-8). Aw ad et al.
(1), in their meta-analysis of reported D A VFs, found that only
6 of 45 patients (13%) with cavernous sinus D A V Fs described
in the literature exhibited either intracranial hemorrhage or
nonhemorrhagic focal neurological deficits other than oph­
thalmoplegia. Likewise, Davies et al. (8) found no patient with
a cavernous sinus D A V F (among a group of 30 patients) who
presented with intracranial hemorrhage or a nonhemorrhagic
neurological deficit. This is consistent with our experience, in
that none of our patients experienced intracranial hemorrhage
before our evaluation. Supporting the perception that cavern­
ous sinus D A VFs are clinically benign lesions is the fact that
some patients experience complete remission of their sym p­
toms without any intervention. Barrow et al. (3) reported that
six of seven patients (86%) with Type B to D fistulae who were
conservatively treated experienced spontaneous resolution at
a median of 6 years (range, 1-16 yr). However, patients with
Type B, C, or D cavernous sinus D A V Fs who do not respond
to conservative measures often require treatment to preserve
visual function and reverse the cosmetic effects of orbital
venous congestion.
The rationale for performing radiosurgery and transarterial
embolization in this staged fashion for patients with D A V Fs
was based on a number of considerations that together max­
imize patient outcomes. First, radiosurgery was expected to
produce long-term obliteration of most D A V Fs in the months
or years after the procedure (2, 5, 11, 18, 19). Barcia-Salorio et
al. (2) reported 20 patients with low-flow cavernous sinus
fistulae who underwent stereotactic radiosurgery. In that se­
ries, 18 of 20 patients (90%) experienced complete closure of
their fistulae, and symptomatic improvement occurred in the
first 1 to 2 months after radiosurgery. Second, when radiosur­
gery was performed before any embolization procedures, the
entire D A V F could be demonstrated angiographically, for
optimal dose-planning. This avoided the problem of partial
irradiation of the fistula because of a component being tem­
porarily occluded by embolization and thus not included in
the radiosurgical plan (24). Third, patients with disabling
symptoms or cortical venous drainage underwent particulate
embolization after radiosurgery whenever possible, to pro­
vide rapid symptom improvement and early protection
against hemorrhage during the latency period before
radiation-induced obliteration. Sim ilar to the experience of
Grossman et al. (10), we found that complete relief of sym p­
toms was usually obtained after partial embolization of the
arterial supply to these complex lesions. Fourth, because
radiation-induced obliteration of the D A V F was ongoing after
radiosurgery, it was thought that the number of patients who
would experience symptomatic recurrence because of recan­
alization of the embolized vessels would be low. In fact, no
patient in our series has experienced symptomatic recurrence
during follow-up periods extending up to 5 years after radio­
surgery. W e think that this staged approach is very effective

in providing both rapid improvement of symptoms and long­
term cure of D AVFs.
W ith an approach of staged radiosurgery and transarterial
embolization, or radiosurgery alone when transarterial embo­
lization was deemed not possible or of little benefit, 19 of 20
patients (95%) with cavernous sinus D A V Fs experienced sig­
nificant symptom improvement within weeks to several
months after the procedures. This interval to symptom im­
provement is much shorter than the natural history of spon­
taneous symptom resolution for patients treated conserva­
tively (3). No patient has experienced a recurrence of
symptoms during the follow-up period. Total or nearly total
(more than 95%) obliteration was observed for all patients
(n = 14) who underwent follow-up angiography 12 months or
more after radiosurgery. The rates of temporary and perma­
nent neurological deficits related to this management strategy
were 10% and 5%, respectively. These results are quite similar
to the recently published report by Guo et al. (11). Among
their series of patients treated using gamma knife radiosur­
gery alone, complete obliteration was observed in follow-up
angiograms, obtained 6 to 15 months after radiosurgery, for
12 of 15 patients (80%) w ith cavernous sinus DAVFs. No
patient experienced either symptom recurrence or radiationrelated complications after radiosurgery.
Delayed oculomotor or trigeminal nerve deficits have been
documented for approximately 10% of patients undergoing
radiosurgery for treatment of lesions of the cavernous sinus
(22, 28). In our series, no patient developed diplopia or facial
numbness, despite the use of a median marginal radiation
dose of 20 Gy. This is probably because of the small number
of patients in our study, and we certainly do not think that the
risk of such deficits is zero for this group of patients. Also, no
patient developed a new visual deficit related to the radio­
surgical procedure. Because the fistulae were usually located
in the inferior or posterior part of the cavernous sinus, we
were able to lim it the radiation dose to the adjacent optic
apparatus to less than 10 G y by using m ultiple (median, three)
small isocenters of radiation to conform ally cover the DAVF.
Our group and others have shown that the risk of delayed
optic neuropathy after radiosurgery is very low with this level
of radiation exposure (17, 22). Lastly, although we have not
observed a case of radiation-induced arterial occlusion after
more than 1200 radiosurgical procedures, Yamamoto et al.
(33) recently reported an asymptomatic patient who was
found to have stenosis of the m iddle cerebral artery after
radiosurgical treatment of an adjacent arteriovenous malfor­
mation. Therefore, although there may be some degree of
injury to the IC A after radiosurgical treatment of c a v e r n o u s
sinus lesions, the risk of symptomatic IC A occlusion is very
low.
Transfemoral transvenous embolization is increasingly be'
ing recognized as an important therapeutic alternative for
cavernous sinus D A V Fs (12, 25). W e regard transfemoral
transvenous embolization as a viable management option
when the pertinent venous anatomic features are favorable.
However, none of the patients in this series had an angi(V
graphically patent inferior petrosal sinus ipsilateral to the
cavernous sinus D A V F. In addition, transfemoral transvenous

Neurosurgery, Vol. 45, No. 3, September 1999

Downloaded from https://academic.oup.com/neurosurgery/article-abstract/45/3/459/2838461
by Tufts University user
on 17 March 2018

Dural Arteriovenous Fistula Radiosurgery
embolization does involve a risk of vascular perforation and
cerebral venous infarction (15, 32). Moreover, the long-term
effects of iatrogenic cavernous sinus occlusion are unknown.
We think that our results compare quite favorably with
recent studies of both endovascular and microsurgical treat­
ment of patients with cavernous sinus D AVFs. M iller et al.
(21) reported the results of 10 patients with Type B to D
cavernous sinus D A V Fs who underwent transvenous embo­
lization via the superior ophthalmic vein. W ith a median
follow-up period of 6 years, all patients experienced symptom
improvement. One patient required an additional transarte­
rial particulate embolization procedure for complete closure
of the fistula. Several patients (number not specified) experi­
enced "transient worsening of ocular m isalignm ent"; one pa­
tient developed new Vlth cranial nerve weakness, which did
not resolve and necessitated later strabismus surgery. Roy
and Raymond (25) treated 12 patients with cavernous sinus
DAVFs, using either a transarterial (n = 1), transvenous (n =
9), or combined endovascular (n = 2) approach. Eleven of 12
patients (92%) were cured at a median time of 6 months after
embolization. Transient V lth cranial nerve weakness was ob­
served for five patients (42%); one patient experienced per­
manent abducens palsy. Day and Fukushima (9) performed
microsurgery for nine patients with Type D cavernous sinus
DAVFs who experienced failure of prior embolization at­
tempts. All patients experienced resolution of their symp­
toms, and complete obliteration was confirmed by postoper­
ative angiography. Transient diplopia and facial numbness
occurred in all patients but resolved by 6 months after sur­
gery. One patient developed permanent hemiparesis related
to occlusion of the IC A . A second patient experienced post­
operative stenosis of the IC A , which caused temporary hemi­
paresis. No patient experienced recurrence of symptoms dur­
ing follow-up periods that extended from 1 to 14 years. Tu et
al- (29) surgically treated three patients w ith either Type C
(n = 1) or Type D (n = 2) cavernous sinus D A V Fs that could
not be successfully embolized. The D A V Fs were obliterated
lor all three patients, and no patient experienced symptom
recurrence during an average follow-up period of 27 months.

CONCLUSION
Staged radiosurgery and transarterial embolization, when­
ever possible, proved to be a safe and effective management
strategy for patients with cavernous sinus D AVFs. This ap­
proach provides rapid relief of symptoms and long-term an­
giographic cure and is w ell tolerated by patients.

a c k n o w led g m en ts

We thank the late Dr. Thoralf M. Sundt, jr., for envisioning
this novel approach to the treatment of patients with DAVFs,
ased on his vast clinical experience, and Dr. Robert J. Coffey
[0r contributing to the M ayo Clinic gamma knife program
between 1990 and 1997.

Received, January 15, 1999.
Accepted, May 14, 1999.
Reprint requests: Bruce E. Pollock, M.D., Department of Neurological
Surgery, Mayo Clinic, 200 First Street, SW, Rochester, MN 55905.

REFERENCES
1. Awad I A, Little JR, Akrawi WP, Ahl J: Intracranial dural arterio­
venous malformations: Factors predisposing to an aggressive
neurological course. J Neurosurg 72:839-850, 1990.
2. Barcia-Salorio JL, Soler F, Hernandez G, Barcia JA: Radiosurgical
treatment of low flow carotid-cavernous fistulae. Acta Neurochir
(Wien) 52:93-95, 1991.
3. Barrow DL, Spector RH, Braun IF, Landman JA, Tindall SC,
Tindall GT: Classification and treatment of spontaneous carotidcavernous sinus fistulas. J Neurosurg 62:248-256, 1985.
4. Brown RD, Wiebers DO, Nichols DA: Intracranial dural arterio­
venous fistulae: Angiographic predictors of intracranial hem or­
rhage and clinical outcome in nonsurgical patients. J Neurosurg
81:531-538, 1994.
5. Chandler HC Jr, Friedman WA: Successful radiosurgical treat­
ment of a dural arteriovenous malformation: Case report. Neuro­
surgery 33:139-142, 1993.
6. Coffey RJ, Link MJ, Nichols DA, Pollock BE, Gorman DA: Gamma
knife radiosurgery and particulate embolization of dural arterio­
venous fistulas, with a special emphasis on the cavernous sinus,
in Lunsford LD, Kondziolka D, Flickinger JC (eds): Gamma Knife
Brain Surgery. Basel, Karger, 1998, pp 60-77.
7. Cognard C, Gobin YP, Pierot L, Bailly A, Houdart E, Casasco A,
Chiras J, Merland J: Cerebral dural arteriovenous fistulas: Clinical
and angiographic correlation with a revised classification of ve­
nous drainage. Neuroradiology 194:671-680, 1995.
8. Davies MA, TerBrugge K, Willinsky R, Coyne T, Saleh J, Wallace
MC: The validity of classification for the clinical presentation of
intracranial dural arteriovenous fistulas. J Neurosurg 85:830-837,
1996.
9. Day JD, Fukushima T: Direct microsurgery of dural arteriovenous
malformation type carotid-cavernous sinus fistulas: Indications,
technique, and results. Neurosurgery 41:1119-1126, 1997.
10. Grossman RI, Sergott RC, Goldberg HI, Savino PJ, Zimmerman
RA, Bilaniuk LT, Schatz NJ, Bosely TM: Dural malformations with
ophthalmic manifestations: Results of particulate embolization in
seven patients. A JN R Am J Neuroradiol 6:809-813, 1985.
11. Guo W, Pan DHC, Wu H, Chung W, Shiau C, Wang L, Chiou H,
Yen M, Teng MMH: Radiosurgery as a treatment alternative for
dural arteriovenous fistulas of the cavernous sinus. A JN R Am J
Neuroradiol 19:1081-1087, 1998.
12. Halbach VV, Higashida RT, Hieshima GB, Hardin CW, Pribram
H: Transvenous embolization of dural fistulas involving the cav­
ernous sinus. AJNR Am J Neuroradiol 10:377-383, 1989.
13. Hamby WB: Carotid-cavernous fistula: Report of 32 surgically
treated cases and suggestions for definitive operation. J Neuro­
surg 21:859-865, 1964.
14. Ishimori S, Hattori M, Shibata Y, Shizawa H, Fujinaga R: Treatment
of carotid-cavernous fistula by Gelfoam embolization. J Neurosurg
27:315-319, 1967.
15. King WA, Hieshima GB, Martin NA: Venous rupture during
transvenous approach to a carotid-cavernous fistula. J Neurosurg
71:133-137, 1989.
16. Kupersmith MJ, Berenstein A, Choi IS, Warren F, Flamm E: M an­
agement of nontraumatic vascular shunts involving the cavern­
ous sinus. Ophthalm ology 95:121-130, 1988.

N e u ro su rg e ry , Vol. 45, No. 3, September 1999

Downloaded from https://academic.oup.com/neurosurgery/article-abstract/45/3/459/2838461
by Tufts University user
on 17 March 2018

465

466

Pollock et al

17. Leber KA, Bergloff J, Pendl G: Dose-response of the visual path­
ways and cranial nerves of the cavernous sinus to stereotactic
radiosurgery. J Neurosurg 88:43-50, 1998.
18. Lewis Al, Tomsick TA, Tew JM Jr: M anagement of tentorial
arteriovenous malformations: Transarterial embolization com ­
bined with stereotactic radiation or surgery. J Neurosurg 81:851859, 1994.
19. Link MJ, Coffey RJ, Nichols DA, Gorman DA: The role of radio­
surgery and particulate embolization in the treatment of dural
arteriovenous fistulas. J Neurosurg 84:804-809, 1996.
20. Lucas CP, Zabramski JM, Spetzler RF, Jacobowitz R: Treatment
for intracranial dural arteriovenous malformations: A meta­
analysis from the English language literature. Neurosurgery 40:
1119-1132, 1997.
21. Miller NR, Monsein LH, Debrun GM, Tamargo RJ, Nauta HJW:
Treatment of carotid-cavernous sinus fistulas using a superior
ophthalmic vein approach. J Neurosurg 83:838-842, 1998.
22. Morita A, Coffey RJ, Foote RL, Schiff D, Gorman DA: Risk of
injury to cranial nerves after gamma knife radiosurgery for skull
base meningiomas: Experience in 88 patients. J Neurosurg 90:4249, 1999.
23. Mullan S: Treatment of carotid-cavernous fistulas by cavernous
sinus occlusion. J Neurosurg 50:131-144, 1979.
24. Pollock BE, Kondziolka D, Lunsford LD, Bissonette DJ, Flickinger
JC: Repeat stereotactic radiosurgery of arteriovenous malforma­
tions: Factors associated with incomplete obliteration. Neuro­
surgery 38:318-324, 1996.
25. Roy D, Raymond J: The role of transvenous embolization in the
treatm ent of intracranial dural arteriovenous fistulas. N eu ro ­
surgery 40:1133-1144, 1997.
26. Takahashi A, Yoshimoto T, Kawakami K, Sugawara T, Suzuki J:
Transvenous copper wire insertion for dural arteriovenous mal­
formations of cavernous sinus. J Neurosurg 70:751-754, 1989.
27. Teng MM, Guo WY, Huang Cl, Wu CC, Chang T: Occlusion
of arteriovenous malformations of the cavernous sinus via the
superior ophthalmic vein. A JN R Am J Neuroradiol 9:539-546,
1988.
28. Tishler RB, Loeffler JS, Lunsford LD, Duma C, Alexander E, Kooy
HM, Flickinger JC: Tolerance of cranial nerves of the cavernous
sinus to radiosurgery. Int J Radiat Oncol Biol Phys 27:215-221,
1993.
29. Tu YK, Liu HM, Hu SC: Direct surgery of carotid cavernous fistulae
and dural arteriovenous m alform ations of the cavernous sinus.
Neurosurgery 41:798-806, 1997.
30. Uflacker R, Lima S, Ribas GC, Piske RL: Carotid-cavernous fistu­
las: Embolization through the superior ophthalmic vein ap­
proach. Radiology 159:175-179, 1986.
31. Vinuela F, Fox AJ, Debrun GM, Peerless SJ, Drake CG: Spontane­
ous carotid-cavernous fistulas: Clinical, radiological, and thera­
peutic considerations— Experience with 20 cases. J Neurosurg
60:976-984, 1984.
32. Watanabe A, Hirano K, Suzuki Y, Kamada M, Mohri Y, Okamura
H, Ishii R: Venous congestion of the insular cortex after transvenous
coil embolization of a dural carotid-cavernous sinus fistula, in Taki
W, Picard L, Kikuci H (eds): Advances in Interventional Neuroradiologi/
and Intravascular Neurosurgen/. Amsterdam, Elsevier Science, 1996,
pp 271-274.
33. Yamamoto M, Ide M, Jimbo M, Ono Y: Middle cerebral artery
stenosis caused by relatively low-dose irradiation with stereotac­
tic radiosurgery for cerebral arteriovenous malformations: Case
report. Neurosurgery 41:474^478, 1997.

CO M M EN TS
Pollock et al. describe 13 patients who were treated using
radiosurgery and then embolization. I do not completely un­
derstand the rationale for this study. There are no conclusive
data indicating that reductions in flow accelerate occlusion after
radiosurgery. In addition, the purpose of embolization be­
fore radiosurgical intervention should be to reduce nidus size,
so that the volume is decreased and maximal radiation doses can
be achieved. It is also known that patients with cortical venous
drainage should be treated as soon as possible, to prevent
subarachnoid hemorrhage. Patients 1 and 2 had very large
superior ophthalmic veins and would be excellent candidates
for the transvenous approach, which is associated with rela­
tively low m orbidity rates. In addition, embolization was
associated with a 15% complication rate for these patients,
which would be considered unacceptably high in most cen­
ters. I agree that transarterial embolization of Type C and D
lesions is often futile, and a transvenous approach is often
required. In cases in which the transvenous approach has
been attempted and has proven unsuccessful, radiosurgical
intervention is certainly reasonable.
Robert H. Rosenwasser
Philadelphia, Pennsylvania
Pollock et al. report 20 patients with cavernous sinus dural
arteriovenous fistulae. These lesions are different from trau­
matic high-flow fistulae and are associated w ith spontaneous
resolution in 10 to 60% of cases.
A ll patients in this study were symptomatic and had expe­
rienced failure of conservative treatment. This article pro­
poses tailored radiosurgery for the entire cavernous sinus, on
the basis of stereotactic angiography and stereotactic mag­
netic resonance imaging. Because radiosurgery involves a
latency interval before closure of the fistula, the goal of com­
bining radiosurgery with embolization has been previously
addressed by this group (2). The concept of using radiosur­
gery before embolization is elegant, and this approach can
facilitate delineation of the target if the target volume is
appropriate for radiosurgery. W hen the combination of the
two techniques was possible, the time to symptom relief was
significantly shorter (median, 2 w k versus 6 mo for radiosur­
gery alone). Nineteen of 20 patients (95%) experienced signiticant symptom improvement; although the natural course is
favorable, this interval to symptom improvement seems to be
much shorter. The rates of temporary and permanent neuro­
logical deficits related to this management strategy were 10
and 5%, respectively, and deficits were correlated with the
endovascular procedures. For each patient, these management risks should be compared w ith the risk of causing
permanent morbidity with a conservative approach.
The results from this study are promising but, because these
lesions tend to resolve spontaneously and neurological com­
plications were observed for 15% of the patients, it is wise to
reserve treatment for patients with visual deterioration, ob­
trusive diplopia, "m alignant" proptosis, or intolerable bruit,
as suggested by Barrow et al. (1). These indications are not
absolute, however, and depend on the general physical con

Neurosurgery, Vol. 45, No. J, September 1999

Downloaded from https://academic.oup.com/neurosurgery/article-abstract/45/3/459/2838461
by Tufts University user
on 17 March 2018

Dural Arteriovenous Fistula Radiosurgery
dition of the patient, the severity of the symptoms, and the
anatomic features of the fistula.
Lars Kihlstrom
Iver A. Langmoen
Stockholm, Sweden

1. Barrow DL, Spector RH, Braun IF, Landman JA, Tindall SC,
Tindall GT: Classification and treatment of spontaneous carotidcavernous sinus fistulas. J Neurosurg 62:248-256, 1985.
2. Link MJ, Coffey RJ, Nichols DA, Gorm an DA: The role of radio­
surgery and particulate embolization in the treatment of dural
arteriovenous fistulas. J Neurosurg 84:804-809, 1996.

Pollock et al. report 20 patients who underwent multimo­
dality treatment of dural arteriovenous malformations of the
cavernous sinus. Some underwent radiosurgery alone and
some underwent both radiosurgery and transarterial emboli­
zation. The vast majority of patients experienced improve­
ment of their clinical symptoms, and many of the patients
who were eligible for follow-up angiography showed success­
ful obliteration of their malformations. Risks were relatively

low, in terms of both new neurological symptoms and recur­
rence of the original symptoms. M any of the transient sym p­
toms developed as a result of embolization techniques used in
adjuvant treatment in the pre- or postradiosurgical periods. This
multimodality approach, sometimes with both radiosurgery and
embolization and sometimes with radiosurgery alone, seems to
provide low-risk management for these difficult lesions, espe­
cially for patients with lesions that are not easily treated or cured
by embolization alone. Radiosurgical pretreatment followed by
embolization for symptomatic patients, during the wait for the
development of radiosurgical benefits, seems to be the proper
staging technique. This approach obviates the problems that
develop because of the inability to define the target volume
when embolization precedes radiosurgery. Radiosurgery seems
to provide permanent obliteration, whereas embolization pro­
vides only temporary obliteration in certain cases. The symp­
toms can be temporarily controlled by embolization during the
wait for the development of the full benefits of radiosurgery.
This report provides longer-term documentation of the sus­
tained benefits of this multimodality approach.
L. Dade Lunsford
Pittsburgh, Pennsylvania

Future Meetings— Congress of Neurological Surgeons
The following are the planned sites and dates for future annual meetings of the
Congress of Neurological Surgeons:

1999
2000
2001
2002

Boston, M A
San Antonio, TX
San Diego, C A
Philadelphia, PA

October 30-November 4
September 23-28
September 29-October 4
September 21-26

Future Meetings— American Association of
Neurological Surgeons
The following are the planned sites and dates for future annual meetings of the
American Association of Neurological Surgeons:
2000
2001
2002
2003

San Francisco, C A
Toronto, Ontario, Canada
Chicago, IL
San Diego, C A

April 8-13
April 21-26
A pril 6-11
April 26-May 1

N e u ro su rg e ry , Vol. 45, No. 3, September 1999

Downloaded from https://academic.oup.com/neurosurgery/article-abstract/45/3/459/2838461
by Tufts University user
on 17 March 2018

467