R#{233}myL. Beaujeux, MD Daniel R#{252}fenacht, MD2

#{149} Daniel #{149} Mazen

Endovascular ofVertebral

patients

were

asymptom-

atic, with vertebral AVF discovered at routine clinical examination. Specific symptoms at presentation in the other patients were tinnitus (n = 21), vertigo (n = 6), neurologic deficit (n = 3), and pain (n = 2). Of the 46 AVFs, 19 (41%) were caused by trauma and 27 (59%) were spontaneous. The fistula was found at C-i to C-2 in 21 (46%) cases, at C-2 to C-5 in five (11%), and below C-5 in 20 (44%). Thirty-four

patients

MD MD

#{149} Alfredo

Casasco,

V

lam branches) These rare

and lesions

a neighboring vein. (1,2) can be either or spontaneous in origin.

traumatic Lesions with a known the result of penetrating

trauma

cause may or blunt

or of iatrogenic spontaneous

genital

origin lesions

or associated

(3,4).

are

with

con-

of the

dys(6). usuat

neck.

in minimal

therapy vertebral

morbidity.

is the treatment AVF.

Endovascular

of choice

for

Index terms: Arteries, therapeutic blockade, 901.1299 #{149}Fistula, arteriovenous, 901.179, 901.494 #{149} Fistula, therapeutic blockade, 901.1299 #{149} Veins, therapeutic blockade, 907.1299

Radiology

1992;

183:361-367

myelopa-

thy (4). Patients with fistulas from penetrating trauma can present with either massive hemorrhage, neurologic deficit, or pulsatile hematoma (3). Treatment is aimed at occlusion of the fistula and preservation of the adjacent artery. It is rarely successful by surgical means. With the development of endovasculam techniques, especially balloon occlusion, embolization has become the standard treatment (4,7).

In this

article,

we review

rience

in management

AVFs

in 45 patients,

our

expe-

of 46 vertebral

with

the

goals

of

reviewing the characteristics of this infrequent condition and determining the advantages of endovascular treatment, on the basis of angiographic

and I From the Department of Neuroradiology and Therapeutic Angiography, H#{244}pitalLariboisi#{232}re, Paris. Received August 30, 1991; revision requested October 1 1; revision received December 20; accepted December 30. Address reprint requests to RUB., Department of Radiology, H#{244}pitalCivil, CHRU de Strasbourg, 1 Place de l’H#{244}pital, 67091 Strasbourg, France. 2 Current address: Department of Radiology, Mayo Clinic, Rochester, Minn. ( RSNA, 1992

and

clinical

Selection

to intervene

could

using

the

(ii

=

patients years

21 male

(53%). (range,

patients

The 6-75

(47%)

average years).

and

age

endovascular

6;

13%),

34 patreated

technique.

neurologic

Angiographic

deficit

such

as

Evaluation

Initial global injection of contrast medium was followed by selective catheterization of the cervical vessels. Injections were then performed to assess the site and the feeding pedicles, flow pattern, venous drainage, and hemodynamic effects (vertebrobasilar steal) of the vertebral AVF. In three

cases

of vertebral

AVF,

used to slow the flow rate ization of the fistula site.

Treatment

sia

were

achieved

vertebrobasilar insufficiency (n = I; 2%), spinal cord symptoms (n = 2; 4%), or cervical radiculopathy (n = 2; 4%). Fourteen patients (30%) were asymptomatic and had their vertebral AVF discovered only at routine auscultation. Color Doppler flow ultrasound (US) was used to confirm the diagnosis of vertebral AVF in 12 patients.

(general

and

route

were

under

light

anesthesia

patient)

neurologic

was

a balloon

was

allow

local-

and

Protocol

procedures

6-year-old

We retrospectively reviewed the charts of 45 patients evaluated for vertebral AVF at our hospital from i978 to i99i. There

be

Presentation

The

METHODS

not

referred for surgery. Overall, with 35 vertebral AVFs were

the patients

AND

abstention).

Patients most commonly presented with tinnitus (n = 2i; 46%). Others had vertigo

results.

PATIENTS

(therapeutic

Three had clinical and angiographic improvement after simple diagnostic angiography and required no further intervenlion. Another two decided against any intervention. Early in our experience (prior to 1981), four patients in whom

Clinical

Pa-

brobasilar

insufficiency

not

by

in complete

in 32 cases

MD

The 45 patients had a total of 46 vertebral AVFs. Two patients had small asymptomatic vertebral AVFs, and it was decided

were tients

tients with vertebral AVFs may also present with bruits, tinnitus, vertigo, and cervical radiculopathy. In rare circumstances, vertebral AVFs can be associated with symptoms of verte-

occlusion

Patient

catheterization

abnormal

vasculature, as in fibromuscular plasia (5) and neurofibromatosis Clinically, vertebral AVFs are ally incidental lesions discovered

auscultation

be

AVFs) were treated with the endovascular technique. Embolization was performed with latex balloons filled with contrast medium in most cases. Endovascular therapy resulted (91%) and partial occlusion in three (9%). The vertebral artery could not be preserved in three patients. Endovascular balloon treatment of vertebral AVFs is effective in occluding the shunt, avoids general anesthesia and surgical intervention, and results

MD Merland,

#{149} Jean-Jacques

Fistula’

arteriovenous fistulas (AVFs) are abnormal shunts between the extracranial vertebral amtery (or one of its muscular or radicu-

Some

Aymard,

#{149} Armand

Rich#{233},MD

ERTEBRAL

routine

(35 vertebral

MD

#{149} Marie-Claire

Treatment Arterlovenous

The clinical and angiographic feahires of 46 vertebral arteriovenous fistulas (AVFs) seen during a 12-year period (45 patients) were reviewed. Fourteen

C. Reizine, H. Khayata,

performed

with

sedative

anesthe-

was

to allow

used

close

monitoring.

used

The

in 34 of 35 (97%)

in one

clinical arterial

of the

24 female

was

44 Abbreviation:

AVF

=

arteriovenous

fistula.

361

a.

b.

Figure tus.

1. Case (a) Vertebral

tem into gram

1.

Spontaneous AVF (*) at C-5

the vertebral

shows

patent

vein

(curved

vertebral

artery

arrow) and

and epidural shunt

d.

C.

vertebral AVF with fibromuscular dysplasia and to C-6 level has vertebral and epidural drainage.

veins

occluded

(straight

arrow).

balloon

(arrows).

by the

complex venous drainage (b) Schematic representation

(c) Balloon

placement

in a 50-year-old shows the

at fistula

woman complex

with drainage

site. (d) Postembolization

tinnisys-

angio-

VA

a.

b.

Figure 2. Case 28. tion) that occurred

and

an associated

clavian

cases

artery,

treated.

artery

artery

after

was

(9%),

of the vertebral

jugular

(arrow).

of C-6 in 57-year-old

vein

puncture.

(b) Representative

obtained

after

treatment

woman

(a) Angiogram

diagram with

of drainage a detachable

without obtained

tinnitus with

system. balloon

VA (encircled

left =

but with

a cervical

bruit

(heard

at ausculta-

subclavian injection shows the fistula site vertebral artery, VV = vertebral vein. (c) Subarea). Occlusion of the fistula, with a patent

achieved.

through

(83%), or

C.

AVF at level

internal

angiogram

has been

Access

#{149} Radiology

vertebral

false aneurysm

in 29 cases

in three

puncture 362

2 weeks

postembolization

vertebral

oral

Traumatic

the

through

artery

the

fem-

axillary direct

above

previously ligated artery in two (6%). loon occlusion of the fistula was used 33 (94%)

a

Balin

cases.

In the embolization

medium-filled

Balt, 0.3

procedure,

contrast

latex

Montmorency, x 0.5-mm

Habbia,

Teflon

Sartrouville,

balloons

France)

(no.

are

I or 2;

tied

catheter

(Technofluor

France),

and

on a

this May

1992

combination tern

is mounted

with

a 3-F

as a coaxial

catheter

(Balt)

for

sys-

ter embolization, vals.

detach-

ment. A 7-F guiding catheter is positioned at the origin of the vertebral artery to allow progression of the coaxial system. The coaxial system is continuously flushed with

saline

into

the

solution. venous

side

ing fluoroscopic eral views) and inflated

The

with

balloon

of the

is passed

fistula,

by

and

Control

diluted

at yearly were

inter-

available

for 29 of the 35 patients (83%). Doppler performed 1 month after the procedure

and then

yearly

thereafter

in six patients (17%) period of the study.

US

was employed

in the more

recent

contrast

me-

Of the 46 vertebral AVFs, 27 (59%) were spontaneous in origin and 19 (41%) were the result of trauma. Four of the spontaneous vertebral AVFs

angiogram

cular

embolization Some

end

of each

procedure.

cases

tions

at the

(up

ries,

required

to four)

several

over

one

embolizaor

multiple

ses-

sions to completely eliminate the fistula. Several sessions are sometimes needed in cases of huge fistulas, to avoid a breakthrough syndrome (3), and also when the embolization procedure lasts too long for

the patient.

Forty-five

required

to treat

occasion, multiple were needed. In three cases, and

not

duced

AVFs.

(up

associated

On

to five)

of the fistula

of the

vertebral

with

the arterial

with

Of the

15 (79%)

were

fibromus-

traumatic

inju-

iatnogenic:

These

were

caused by jugular vein puncture (n = 10), carotid artery puncture (n = 3), or surgery (n = 2). The memaining four (21%) were caused by hyperextension injury (n = 3) or gunshot wound (n = 1).

trast media the fistula

artery

balloons

intro-

side. In one case,

(59%)

of the

vente-

bral AVFs were on the right vertebral artery, and 19 (41%) were on the left. The

fistula

was

found

on

the

follow-

ings

diminished

therapy immediate

into

with

ing

a balloon

artery

sacrificed

at the

level

was

of the

fistula

and

occluding the artery; at control angiography, there was a persistent fistula that

was

occluded

vertebral

vein

clusion. paresis the

vertebral

artery)

by catheterizing

the

and performing

In another and multiple

neurologic

balloon

patient, venous

condition

oc-

with quadridrainage, deteriorated

into

a perimedullary

previously

fistula

was

with

then

n-butyl

2

recovery

initial

balloon

balloon

had

immediatedly

embolized

Paris),

was

and

rapid. by

deflation

subse-

In another

embolization

followed

not

(Histoacryl;

Bruneau,

quent

was

that

The remaining

cyanoacrylate

Laboratoire

AVF

vein

treated.

case,

of a vertebral

a relapse

secondary

because

to neck

tion. A Gianturco coil (Cook, Paris) then used to occlude the fistula.

of

rotawas

A plain 15, and

Evaluation radiograph

at i, 8,

occlusion,

to observe

of the balat 3-4

weeks).

3 years;

after

obtained

whether the deflation timing loon was normal (ie, deflation clinically, Doppler

30 days

was

The patients

range,

6 months

clinical condition and were usually assessed

Volume

were

angiographically, flow US (mean

183

#{149} Number

followed

and follow-up

up

with color period,

to 8 years).

The

angiographic results at I or 2 months af2

and

at C-2

Spontaneous

om above

vertebral

or above in 63% than C-S in 58%

The vertebral AVF was supplied by the vertebral artery itself in 39 (85%) of the cases. Other feeding pedicles muscular

branches

in four

(9%) and radicular (6%). The venous

branches drainage

in three was out-

side

the

transverse

foramen

in 19

cases

(41%), inside in 12 (26%), and both in 15 (33%). Drainage was ascending in six cases (13%), descending in 27 (59%), and in both directions in 13 (28%). Drainage was into the vertebral vein in 15 cases (33%), internal jugular vein in 14 (30%), epidural venous

plexus

in 12 (26%)

(Fig

associated

with

false

aneurysms

in seven (15%) of the patients (Fig 2); six were on the venous side and one was on the arterial side. Angiographic

treatfindafter

delayed

was

of the

(several

patients

days)

quadriparesis.

in the

At long-term

follow-up evaluation (3 years), the vertebral AVFs were found to have remained occluded, and the clinical improvement has been sustained. One of the two patients with quadriparesis has neurologic

experienced recovery,

a complete the other

and

has

markedly improved. In the three patients with only partial endovascular occlusion of the vertebral AVF, the results were more variable. One patient with tinnitus had relief of his symptoms, one with radiculopathy

required

surgery,

one previously asymptomatic remained unchanged.

and

patient

Complications

1), epi-

dural venous plexus and internal jugular vein in four (9%), and penimedullary veins in one (2%). Vertebral AVFs were

Follow-up

(11%),

AVFs were at C-2 (n = 17) and lower (n = 11).

included

days after placement of the balloon in the vertebral vein. Emergency angiography showed that the fistula was draining only

been

in five

in 21 (46%).

by locat-

thereby

(fed by the contralateral

un-

1). Improvement

in 30 (94%)

was

vertebral

remained

or disappeared

(Table

two

The

other

In all 32 patients completely ed, the objective and subjective

and

intentionally

the

Results

C-S

fistula.

At final control anvertebral artery was

were sufficient to occlude and relieve symptoms.

bosed and changed.

the vertebral artery was of small diameter, and the balloon could not be introduced the

treated.

Residual false aneurysms were found in two cases after embolization. At follow-up examination 1 year later, one of these had completely throm-

ing portions of the vertebral artery: below C-S in 20 (44%), between C-2 to

therefore

surgically

patent in 32 (91%) of the cases treated. In four patients with small fistulas (two after percutaneous puncture), catheterization and injection of con-

Clinical

Topography Twenty-seven

occlusion

be achieved

from

were

dysplasia.

finally

were

35 vertebral

balloons

preservation

could

procedures

the

(15%)

was

them intervention. giography, the

of AVF

diurn (one part contrast medium, two parts saline solution) to totally obliterate the fistula. Once selective occlusion of the fistula is confirmed while preserving the vertebral artery, the balloon is detached by use of gentle coaxial traction. A control is obtained

of the occipital and vertebral arteries entering a large false aneurysm, The remaining two AVFs are being followed up clinically, without fur-

RESULTS Causes

posttraumatic AVF on muscular C-2 territory and was fed by multiple pedicles, including muscular branches

which

us-

guidance (frontal and latroad-mapping, and it is

enough

then

angiograrns

Results

Total occlusion of the fistula was achieved in 32 (91%) of the 35 patients treated (Table 1). In the remaining three cases (9%), the occlusion was partial, with persistent venous drainage. One of these three was a

Theme were no clinically symptomatic complications in any of the 35 patients treated. The vertebral artery was found angiographically to be occluded

in three

clusions tions.

caused One

(9%)

cases;

no clinical

occlusion

was

these

oc-

manifestaperformed

intentionally because of the small diameter of the distal artery and the impossibility to achieve clusion. In another case,

selective octhe balloon

remained bulging in the vertebral antery lumen, and the artery was found to be occluded on subsequent angiograms.

one

Finally,

patient

symptoms

resolved

of fistula

after

a failed

Radiology

in

at#{149} 363

Table 1 Angiographic

and

Clinical

Results

Embolization

Site (Vertebral Artery/Location)

Case No./Age(y)/Sex 1/50/F

R/C-5 to C-6

S

R/C-5 R/C-i R/C-6

to C-6 to C-2

V

5/34/M

L/C-5

to C-6

6/71/M

L/C-5

7/75/M

L/C-6

8/51/M 9/50/F

V

Tinnitus

Total

Asymptomatic

V

Asymptomatic

Partial

Asymptomatic

S

Vertigo

Total

Asymptomatic

R/C-6

V

L/C-2 L/C-1

S S

Asymptomatic Tinnitus

Asymptomatic Asymptomatic Asymptomatic Asymptomatic

to C-7

Tinnitus

Total

Asymptomatic

13/40/M 14/31/F

R/C-4 R/C-1

Tinnitus Cervical

Total Partial

Asymptomatic Asymptomatic

15/53/M

L/C-5 R/C-6 R/C-1 R/C-1 L/C-1 L/C-1 L/C-1 R/C-1 R/C-6 L/C-i

T ‘II ill

Tinnitus Vertigo Tinnitus Tinnitus Tinnitus Tinnitus Quadriparesis Asymptomatic Tinnitus; pain Tinnitus Tinnitus

Total Total Partial Total Total Total Total Total Total Total Total

Asymptomatic Asymptomatic Asymptomatic Asymptomatic Asymptomatic

Note.-L

to C-2 to C-6

to C-2

to C-7

Ttt S

to to to to to

C-2 C-2 C-2 C-2 C-2

R/C-6 L/C-6 PlC-i

to C-2

L/C-6 L/C-1 R/C-1 L/C-1 R/C-1 L/C-1

S

right,

=

to C-2 to C-2

spontaneous,

=

T

=

radiculopathy

Asymptomatic

Some improvement Asymptomatic Asymptomatic Asymptomatic Asymptomatic

Quadriparesis

Total

Marked

S Tt

Asymptomatic Asymptomatic

Total

Asymptomatic

S

Asymptomatic

Total Total

V

Cervical

Total

S S S S S

Tinnitus; pain Vertigo Asymptomatic Asymptomatic Tinnitus

Total Total Total

Asymptomatic Asymptomatic Asymptomatic Asymptomatic Asymptomatic Asymptomatic

Total

Asymptomatic

Total

Asymptomatic

radioculopathy

improvement

traumatic.

dysplasia.

Fibromuscular

of the internaljugular

vein.

following car accident. I Sequelae of direct puncture of vessels for angiography. 0 Cases 18 and 19 involved two vertebral AVFs in the same ** Head and neck surgery. tt Gunshot wound.

U Also underwent H underwent

embolization

with n-butyl-cyanoacrylate.

embolization

with

tempt at embolization; the was later found to actually thrombosis

of the

a Cianturco

patient have had

vertebral

Vertebral

AVFs

are

either

in origin.

from

of Halbach

(59%)

spontane-

Our

series

et al (3),

of the AVFs

in our

spontaneous. Some AVFs are congenital, in young patients on those

who are become ditions balance,

asymptomatic. These fistulas apparent during certain conthat change the hemodynamic such as pregnancy or hyper-

tension

(2).

are

to fragile

spontaneous

abnormal

(fibromuscular

require #{149} Radiology

only

become coughing,

symptomatic or sneezing)

Traumatic

were

Other

a

vertebral

(trivial (2).

AVFs

trauma,

are most

commonly though plication

or traumatic that

patient.

coil.

artery.

DISCUSSION

most

Asymptomatic

11 5* 5* 5 T S S S T**

R/C-6

left, R

=

to C-2

L/C-3

t Attempts at catheterization t Spine surgery. * Hyperextension injury

364

Asymptomatic Asymptomatic Asymptomatic Asymptomatic

Tinnitus

28/57/F 29/15/F 30/68/M 31/30/F 32/24/F 33/52/M 34/49/F 35/6/F

and

Total Total Total Total

S

27/6/MU

due

S S

Tinnitus Asymptomatic Asymptomatic Asymptomatic

R/C-6

23/66/F 24/47/M 25/26/M 26/27/MU

structure

Clinical Outcome

12/20/F

22/27/M

patients spontaneous especially

Symptoms

S

18/66/F* 19/66/F# 20/51/F 21/58/M

since

Degree of Occlusion at Angiography

R/C-2

17/53/F

differs

Evaluation

Total Total Total Total

16/66/F

ous

and at Last Follow-up

Cause

2/53/M 3/15/F 4/12/M

10/23/F 11/56/F

*

Balloon

after

a minimal

AVFs

vascular dysplasia)

insult

to

of iatrogenic cause. Althis type of AVF is a rare comof internal jugular venous puncture (8,9), 53% (n = 10) of our traumatic fistulas were a result of such puncture. Other iatrogenic causes include surgery in the neck

(10). Penetrating injuries accounted for only one (5%) of our cases, versus 60% in Halbach’s series of 20 cases (3). In our series, we found that posttraumatic vertebral AVFs affected the lower portion of the vertebral artery

0

elow C-5), while spontaneous ones mainly involved the upper portion (at or higher than C-2) (Table 2), which was similar to findings in other series (4-7). Nagashima et al found that

most AVFs involve the left vertebral artery (1). In our series, however, most (59%) of the fistulas affected the right vertebral artery; this was probably related to the techniques used in the placement of central catheters in patients with AVF of iatrogenic ongin. Clinical symptoms produced by vertebral AVFs are related to flow mate

in the shunt, and

venous

factors

can

chronicity drainage

further

rounding ondary

structures arterial steal

tension

(7).

Vertebral

tomatic

AVFs

and

at routine the cervical by turbulent

of the pattern.

influence

lesion, These

sum-

by way of secor venous hyperare

often

are sometimes

asymp-

discovered

examination, as a bruit in region. Tinnitus, caused flow, is another common May

1992

bral

AVFs

from

mations

anteniovenous

(13). In certain

angiography simple attempt

alone (three at selective

ization

cases)

(four

occlude provement.

the

cient

cases) or a catheter-

was

sufficient

to

of the

altered

fistula

after

any effect related to the viscosity of contrast media. The goal of treatment is occlusion of the fistula site and preservation of

methods

symptom of vertebral AVF and seems to be related to the ascending venous drainage to the jugular foramen. Neurologic complications are rare with vertebral AVFs. Brain stem signs (ver-

tigo, ventebrobasilar rare and probably

insufficiency) are are due to arterial

steal. Some cases opathy are related

of cervical to direct

sion veins

of the nerve (11). Spinal

roots cord

rare,

and can be caused

nect

compression

dilated

and

of the

pulsatile

fect) or by spinal

spinal

veins

cord

di-

cord

(mass

ischemia

nal

to what

dural

venous ciency

is observed

fistula

with

drainage in young

with

of the

tnalateral mon and

vertebral, external

eral

and

cervical

should injection

artery ing

ipsilateral

and

costocervical

of the

is

means

anastomosis

such

the

vertebral artery (10). The exact fistula site must be well visualized. The fistula is sometimes better visualized by

fistula

(26).

filled

balloons

ment

of vertebral

in our

series

(Fig

approach

warranted

The

classic

3).

stiffness

to AVFs

in most

is no

cincum-

surgical

arterial

feeders.

bypass

and

preservation

artery.

It is the

The

today,

and

has

now

main

fistulas. is

and

verteof

a high

benememaand

few hospitalization days. They even be used in cases of failure

With

smaller With

can of

series

with different ment, a very high occlusion of AVFs

After enced toms.

underwent general

flow

Volume

the

Balloon most

this

series

is reduced.

Selective catheterization sionally helpful to distinguish 183

#{149} Number

2

is occa-

verte-

occlusion commonly (24,25).

tomatic

of

which

use

of the fistula used method Balloons

were

was in intro-

et al, preser-

Our

patients

recovery tinnitus,

performed

with

and in 91% results

patient recruitsuccess rate for safe has been achieved.

symptoms

Of the one and

a substantial

only occurred

two

patients

is indepenthe other amount

complication

vertebral

experi-

from their sympvertigo, verteand even disappeared

insufficiency,

recovered

series, boy)

(23).

fistula artery These

embolization, total Bruits,

of a proximal

treatment anesthesia

even

vation of the vertebral artery in 12 of 20 (60%) cases (3). Thus, in two large

is then

Injection

the

are similar to those of Halbach who achieved occlusion and

means

blood

agents

of catheters,

to occlude

strength.

for continuous

of the

related to of small

embolic

types

sia is important

the

un-

for failure

various

tion of the fistula site requires tempomary and partial occlusion of the involved ipsilateral vertebral artery by

balloon.

investigation

preserve the vertebral (32 of 35) of the cases.

brobasilar

monitoring. In our patient (a 6-year-old

with

fistulas can be occluded. the above methods, it was

prior surgical treatment (10,22). In cases of complete transection of the vertebral artery, a trapping procedure

neurologic only one

the

route have in catheterization

quadniparesis, ambulatory

where

the

agent

of previously

reasons

different

possible

mode

allow

with dently

opacifivisualiza-

avoid

be navigated with variable

embolization

endovasculan difficulty

portion

of the primary

and

with two balloons is effective for elimmating the shunt and the hemorrhage (3). The light sedative anesthe-

vertebral good

and

embolic

it cannot Catheters

treatment.

cases,

medium-

for treat-

attempts with the methods can be balloon through (27), particle emboli-

liquid

after

contralateral In some

AVFs

in

reachable shunts. Coil embolization through the contralateral vertebral artery has also been described (30).

proce-

The

(3,28),

cord

using

Contrast

are effective

spinal

cation.

occurred

use of a balloon inflated with a polymerizing substance such as hydmoxyethyl methacrylate. If only a partial occlusion is accom-

small that a balloon.

fit-to-cost ratio. These procedures quire only minimal embolization terial, light sedative anesthesia,

of lookwith

two

surgical

treatment

occipital

emnous

postoper-

access

were

Endovascular treatment of ventebral fistula is now the simplest, most reliable method for occluding AVFs. It is very effective in elimination of the bral

arteries

is an important for the usual

cases

shunt

con-

(4). Sometimes,

ipsilateral

ative

of the fis-

primary

then combined with endovascular embolization of the remaining of the vertebral artery (15).

ascending

be performed

of the

There

several

ipsilateral comcarotid, and ipsilat-

contralateral

loss

be fol-

the subclavian to the distal vertebral artery for large, complex fistulas with

AVFs series.

and

should

It is always

dures involved direct surgical ligature of punctiform fistulas (16,20,21) and the use of a bypass venous graft from

perimedullary

Angiography is the standard radiologic examination for vertebral AVF. Assessment of the vascular anatomy necessary before embolization is attempted. Detailed assessment of the

anatomy

ligation

by revascularization

stances.

spi-

aneurysms

not

after balloon occlusion of the fistula only two cases, which differs from previous experience with carotid-cay-

zation

(1,16,17).

vessel

and

longer

(12). Cardiac insuffipatients (2) or mas-

sive hemorrhage in traumatic (3) was not observed in our

difficult (1,15,18-20).

The

ated by arterial steal or venous hypertension. Venous hypertension results in neumologic symptoms in a manner

similar

tula

ef-

gener-

ap-

was

(glue or autologous clot), and coils (29) have been used to complete the occlusion of the shunt. These are especially useful when the shunt is so

always

route. by

yen-

false

artery

(3). This

and

lowed is

to obliterate

be used series.

plished, multiple same or different used. A detachable the venous route

avoided

by the dilated involvement

used

artery. of endovassurgical

is present,

vertebmal

tebral AVFs (14,15). Because of the many arteriolized venous channels and the direct inaccessibility of many of the fistulas, surgical treatment of AVFs has often been unsatisfactory

Proximal

radiculcompres-

by either

were

femoral artery or by with road-mapping, occluded vertebral

If a transection

Residual

or second-

the patency of the vertebral Prior to the development cular techniques, various

(4).

proach can used in our

suffi-

occlusion,

artery

manip-

hemodynamics

to produce

duced via the direct puncture, into a previously

the contnalateral

fistula and cause imThis may be related to

thrombosis

ulation,

malfor-

cases,

was

artery in only

reported

with

asymp-

occlusion, three

one of these, occlusion of the was intended to also occlude shunt. Potential complications

ously

has of

cases;

in

artery the previ-

endovasculan Radiolotrv

#{149} 365

L

fl-

) ACA ::::j

a.

b.

Figure means

3. Case 10. of proximal

persistent vertebral dural venous drainage (wide straight arrows) representation

ACA

=

treatment

through include this

artery

ischemic

media-related

normal-perfusion

(3),

puncture

cervical

include

of the

were

vertebral

events,

pressure

of which

4.

and

breakevents and emboli observed

for this condiwith close in conscious morbidity and

matic

arteriovenous

artery 2.

3.

T, Kawanuma

6.

spinal

Halbach Treatment

A, Susuki fistula of the cord symptoms.

VV, Higashida of vertebral

#{149} Radiology

treatment

12.

Hieshima GB, Cahan LD, Mehringer CM, Bentson JR. Spontaneous arteriovenous fistulas of cerebral vessels in association fibromuscular dysplasia. 1986; 18:454-458.

Deans

J Neuro-

GB. fis-

Neurosur-

WR, Block 5, Leibrock

13.

L, Berman fistula in Radiology

7. Reizine

D, Laouiti M, Guimaraens L, Rich#{233} 14. MC, Merland JJ. Les fistules art#{233}rioveineuses vert#{233}brales:aspects cliniques,

K. Trauvertebral

RT, Hieshima arteriovenous

AJNR 1987; 8:1121-1128.

from

BM, Skultety FM. Arteriovenous patients with neurofibromatosis. 1982; 144:103-107.

5,

surg 1977; 46:681-687. De Bray JM, Bertrand P, Bertrand F, Jeanvoine H. Les fistules art#{233}rio-veineuses spontan#{233}esde l’art#{232}re vert#{233}brale:a propos d’un cas, revue de Ia litt#{233}rature. Rev Med mt 1986; 7:133-139.

tulas.

366

with

ligature

treated injection

by shows

artery (straight arrows). The dilated epiarrow) angiogram. The left vertebral artery The fistula site can be seen (*). (c) Schematic of the vertebral AVF. VA = vertebral artery,

JJ, Reizine D, Riche MC, et al. 11. Traitement endovasculaire des fistules art#{233}rio-veineuses vert#{233}brales:a propos de vingt-deux cas. Ann Chir Vasc 1986; 1:73-

et traitement

endovasculaire-a

A, Kamisasa

Sakaguchi

the

bruit. AVF had been surgically with right subclavian artery

Merland

with gery

8.

C, Iwasaki

above

angiographiques,

References Nagashima

(arrow)

78. S.

in

occlusion of vertebral reliable, cost-effective, method that is now

the primary treatment tion. It can be performed neurologic monitoring patients, with minimal no mortality. U

1.

artery

series.

Endovascular AVFs is a safe, and reproducible

with obtained

(Fig 3 continues).

reaction,

(3,31-34). Ischemic spasm, dissection,

none

to C-2 in 23-year-old woman recurred. (a) Angiogram

the contralateral (ligated) vertebral is well visualized (curved arrow). (b) Left ascending cervical artery (curved is supplied by collaterals from muscular branches (narrow straight arrows).

of direct

ascending

contrast

C.

Spontaneous vertebral AVF at level of C-i ligature of the vertebral artery, but symptoms AVF (*). The right vertebral artery supplies

9.

par

voie

de vingt-cinq

cas. Ann Radiol 1985; 28:425-438. Hayward R, Swanton H, Treasure quired arteriovenous communication: plication of cannulation of internal vein. Br Med J 1984; 288:1195-1196.

15.

T.

Accomjugular 16.

terne. 164. Johnson

apres

cath#{233}t#{233}risme jugulaire

Ann Fr Anesth CE, Russell

Resolution pseudotumor

EJ, Huckman

1986;

in5:162-

Neuroradiology

vertebral

17.

MS.

of spinal epidural vascular following balloon occlusion

of a postoperative

fistula.

Reanim

arteriovenous 1990; 31 :529-532.

18.

HL Jr. Svien a complication

bral angiography-a

report

HJ.

Arteof verte-

of a case.

Neurosurg 1963; 20:73-75. Merland JJ, Reizine D. Malformations vasculaires vert#{233}brom#{233}dullaires. Paper no. 31671 Gb, radiodiagnostic II. Paris: Encycl Med Chir, 1987. Lawdahl RB, Routh WD, VitekJJ, McDowell HA, Gross GM, Keller FS. Chronic arteriovenous fistulas masquerading as arteriovenous malformations: diagnostic considerations and therapeutic implications. Radiology 1989; 170:1011-1015. Mathey 1’ Cormier JM. Fistule art#{233}rioveineuse cong#{233}nitale du segment cervical des vaisseaux vert#{233}braux. Ann Chir 1957; 11:307-314. George B, Laurian C. Techniques of vertebral artery surgery: arteriovenous malformations. In: George B, Laurian C, ed. The

vertebral

Verri#{232}res D, Bernard C, Dacheux J, Reizine D, Echter E. Fistules art#{233}rio-veineuses cervicales

10.

propos

Olson RW, Baker rio-venous fistula:

artery:

pathology

and surgery.

Vienna: Springer-Verlag, 1987; 167-170. Schumacker HB, Campbell RL, Heimburger RF. Operative treatment of vertebral arteriovenous fistulas. J Trauma 1966; 6:3-19.

Berguer R, Feldman Cure by combination tachable intravascular 1982; Bartal

AJ, Wilner

HI, Lazo A.

of operation

balloon.

Ann

and deSurg

196:65-68. AD, Levy

J. Excision of a congenital suboccipital vertebral arteriovenous fistula. J Neurosurg 1972; 34:452-456.

May

1992

d. Figure

(d) Angiogram

e.

tery

supplies

3 (continued). (straight arrows)

(curved uistula

19.

20.

21.

22.

23.

24.

arrows).

the

The intracranial

is occluded

with

a balloon

obtained fistula

JL, Couffinhal

(*) at the and

and occlusion

of major

cerebral

vessels.

interspace.

is not seen

intracranial

JP,

Andreassian B. Fistule art#{233}rio-veineuse vert#{233}brale: technique chirurgicale par voie post#{233}rieure para-m#{233}diane. Neurochirurgie 1982; 28:315-318. Chou SN, French LA. Arteriovenous fistula of the vertebral vessels in the neck: case reports. J Neurosurg 1965; 22:77-80. Russell EJ, Goldblatt D, Levy JM, et al. Percutaneous obliteration of a postoperatively persistent vertebral arteriovenous fistula. AJNR 1989; 10:196-200. Sadasivan B, Mehta B, Dujovny M, Malik GM, Ausman JI. Balloon embolization of nontraumatic vertebral arteriovenous fistulae in children. Surg Neurol 1989; 32: 126-130. Serbinenko FA. Balloon catheterization

of the left vertebral

cervical

system the

25.

and

JC, Coquillaud

injection first

vertebrobasilar (arrows),

Markham JW. Spontaneous arteriovenous fistula of the vertebral artery vein. J Neurosurg 1969; 31:220-223.

Berthelot

with

26.

Neurosurg 1974; Debrun C, Legre

JP.

tebral

flstulae

fistula.

29.

30.

after direct drains

into

(e) Control

puncture angiogram

31.

41:125-145.

by detachable

balloons

of 54 traumatic

blood

with

flow. 32.

of arteriovenous 1980;

Morelli RJ. of vertebral

An angiographic arteriovenous

Neurosurg

Psychiatry

complication fistula. J Neurol

1967; 30:264-266.

of a traumatic Neuroradiology

33.

20:203-205.

Rossi PL, Passariello R, Simonetti G. Control of a traumatic arteriovenous fistula by a modified Gianturco coil embolus system. AJR 1978; 131 :331-333. Freitag HJ, Grzyska U, Zeumer H. The use of the “cross over technique” in the

management bral fistula.

obtained

The vertebral

and

epidural

after

ar-

veins

embolization.

The

Halbach VV, Higashida RT, Hieshima GB, Norman D. Normal perfusion pressure breakthrough occurring during treatment of carotid and vertebral fistulas. AJNR 1987; 8:751-756. Higashida RT, Halbach VV, Tsai FY, et al.

Interventional neurovascular treatment of traumatic carotid and vertebral artery lesions: results in 234 cases. AJR 1989; 153:

F, et al.

carotid-cavern-

occlusion

Neuroradiology

view).

cervical,

injected.

J, Kasbarian M, Tapias PL, Endovascular occlusion of ver-

balloon

(lateral

suboccipital,

J Neurosurg 1981; 55:678-692. B, Hoare R. Percutaneous trans-

Kendall venous

28.

are

conservation of the vertebral Radiology 1979; 130:141-147. Debrun C, Lacour P, Vinuela Treatment ous fistulas.

27.

fistula

on this angiogram.

vessels

Caron

artery

The

577-582. Spetzler Normal theory.

34.

Kondoh

RF, Wilson CB, Weinstein P. et al. perfusion pressure breakthrough Clin Neurosurg 1978; 25:651-672. T, Tamaki N, Takeda N, Suyama

T, Oi S, Matsumoto hemorrhage extracranial

J Neurosurg

S.

Fatal intracranial

after balloon occlusion vertebral arteriovenous 1988; 69:945-948.

of an fistula.

vertebro-verte1989; 31:174-

176.

Volume

183

#{149} Number

2

Radiology

#{149} 367