Neurosurg Rev (2013) 36:541–549
DOI 10.1007/s10143-013-0470-1

ORIGINAL ARTICLE

Sylvian fissure arteriovenous malformations: long-term
prognosis and risk factors
Lingtong Liu & Hao Li & Jian Zheng & Shuo Wang &
Jizong Zhao & Yong Cao

Received: 16 July 2012 / Revised: 10 January 2013 / Accepted: 10 March 2013 / Published online: 5 May 2013
# Springer-Verlag Berlin Heidelberg 2013

Abstract Background and objective: Sylvian fissure arteriovenous malformations (AVMs) are among the most challenging AVMs to manage surgically. The estimates of their risk
factors and prediction of their long-term prognosis are crucial
for clinical decision-making. The authors conducted a retrospective review to patients with sylvian AVMs treated
microsurgically to evaluate the risk factors associated with
long-term prognosis. Methods: Forty-one patients with
sylvian fissure AVMs treated microsurgically between June
2009 and December 2011 were retrospectively reviewed with
a mean follow-up time of 23 months (range 6–35 months).
Chi-square test was utilized to compare proportions and ranksum test to compare ordinal materials. Odds ratios (ORs) were
used to assess risk factors associated with postoperative shortterm outcome and long-term adverse outcome (mRS scores 3–
6). Results: One patient died in 1 month after surgery. Eighteen (43.9 %) patients had postoperative transient neurological
deterioration. Good outcomes (mRS scores 0–2) were observed in 29 (72.5 %) patients in the follow-up. From the
multiple logistic regression, a history of AVM bleeding and
L. Liu : H. Li : J. Zheng : S. Wang : J. Zhao : Y. Cao (*)
Department of Neurosurgery, Beijing Tiantan Hospital,
Capital Medical University, 6 Tiantan XiliChongwen District
Beijing 100050, China
e-mail: caoyong6@hotmail.com
L. Liu
e-mail: liulingtong21@yahoo.com.cn
H. Li
e-mail: handanlihao-1@163.com
J. Zheng
e-mail: neurosurgeon.zheng@gmail.com
S. Wang
e-mail: captain9858@vip.sina.com
J. Zhao
e-mail: zhaojz205@hotmail.com

deep venous drainage increase the risk of postoperative transient neurological deterioration, with OR=8.211 and OR=
4.869, respectively. A history of AVM bleeding was a risk
factor for long-term adverse outcome, with OR=7. Conclusions: Despite different Sugita classification subtypes, sylvian
fissure AVMs’ long-term results with microsurgical resection
are better than expected; a history of AVM bleeding is a risk
factor for postoperative temporary neurological deterioration
and for long-term adverse outcome, while the AVM deep
venous drainage is a risk factor only for temporary neurological deterioration.
Keywords Arteriovenous malformation . Sylvian fissure .
Prognosis . Risk factors

Introduction
Sylvian fissure arteriovenous malformations (AVMs) were
first described by Kenichiro Sugita in 1987 [1], which is
supposed to account for approximate 8.92–11.1 % of all
brain AVMs [7, 13]. The reported incidence rates of newly
diagnosed AVMs have varied in different population-based
studies from 0.89 to 1.34 cases per 100,000 person-years [1,
2, 4, 5, 12]. Based on the data, the incidence rates of newly
diagnosed sylvian AVMs varied about 0.09 to 0.14 cases per
100,000 person-years. The sylvian AVMs are among of the
most challenging AVMs to manage surgically. They involve
the middle cerebral artery and are surrounded by critical
structures such as insular cortex, basal ganglia, internal
capsule along the deep margin, and Broca’s areas along
the lateral margins in the dominant hemisphere. They require substantial subarachnoid dissection to open the sylvian
fissure, expose the nidus, and identify feeding arteries and
draining veins. Transit arteries that continue beyond the
AVM to supply normal brain can resemble feeding arteries

542

and must also be preserved. Draining veins often cover the
nidus and need to be preserved throughout the resection.
Smaller arteries such as the lenticulostriate and anterior
choroidal arteries contribute to these AVMs and can be
difficult to expose and control [7]. For its rarity and surgical
difficulty, this disease is confined to small series and few
studies has yet been able to demonstrate long-term prognosis and associated risk factors [7, 13, 15, 16]. We retrospectively reviewed 41 patients suffering from sylvian fissure
AVMs in our hospital and identified their long-term prognosis and prognostic risk factors.

Neurosurg Rev (2013) 36:541–549

1.56:1, with 25 males (61 %) and 16 females (39 %). Sixteen
(39 %) patients presented with previous hemorrhage before
surgery, 16 (39 %) with seizures, 7 (17.1 %) with headaches,
and 2 (4.9 %) with hemiparesis. The detail of every patient
was shown in the Table 1. In our series, only one patient
(patient 40) had ever received endovascular embolization for
right lateral sylvian fissure AVM found by intracerebral hemorrhage 8 years ago. She suffered recurrent hemorrhage on the
same location before admitted. No other therapy was applied
preoperatively and postoperatively in our patients.
AVM characteristics

Subjects and methods
Study population
A total of 355 patients with AVMs were treated with microsurgical resection by a single neurosurgeon (SW) at the Beijing Tiantan Hospital, Capital Medical University between
June 2009 and December 2011. Of these patients, 41 were
identified as having sylvian fissure AVMs. They accounted for
11.55 % of all AVM patients. The study was approved by the
Ethics Committee of the Beijing Tiantan Hospital.
Cerebral angiography, computed tomography (CT),
and/or magnetic resonance imaging (MRI) was performed
in all patients to confirm the AVM diagnosis. AVMs were
categorized by the operating neurosurgeon as pure, medial,
lateral, and deep types, according to the Sugita classification
[13]. CT scans, MRI scans, angiograms, intraoperative photographs, and operative reports were reviewed before making these determinations.
Follow-up checkups were in 3 to 6 months after the
surgery and then at yearly intervals. The most recent checkup was done in the June 2012; mean follow-up time was
23 months (range 6–35 months). Neurological assessments
were performed by a neurosurgeon resident, under the supervision of an attending neurosurgeon, preoperatively,
postoperatively, and during the follow-up period. Newly
emerging neurological dysfunction or deterioration of original neurological deficit appeared shortly after operation
(within 1 week) was defined as bad short-term result; no
new neurological dysfunction or mitigation of original neurological deficit after surgery was regarded as good shortterm result. The modified Rankin Scale (mRS) [14] was
used to grade long-term outcomes. The mRS scores 3–6
was named as bad long-term outcome; mRS scores 0–2
were named as good long-term outcome.
Patient characteristics
Of 41 patients with sylvian fissure AVMs, the average age was
36.1 years (range, 6–56 years) and the male/female ratio was

According to the Spetzler–Martin grading system, 4 patients
(9.8 %) were grade I AVMs, 15 (36.6 %) were grade II AVMs,
19 (46.3 %) were grade III AVMs, 2 (4.9 %) were grade IV
AVMs, and 1 (2.4 %) were grade VAVMs. The mean diameter of
AVMs was 39.7 mm (range, 5–70 mm). Twenty (48.8 %) AVMs
located in the eloquent area like Broca gyrus, basal ganglia, and
Wernicke’s areas. Thirteen (31.7 %) had deep venous drainages.
According to the Sugita classification system, 7 AVMs (17.1 %)
were pure AVMs, 17 (41.5 %) were lateral AVMs, 12 (29.3 %)
were medial AVMs, and 5 (12.2 %) were deep AVMs.
Statistical analyses
Frequencies were used for prevalence data. Chi-square test
was utilized for categorized data and rank-sum test for ordinal
materials. The level of statistical significance for univariate
analysis was set at 0.05. The P values for entry and removal
during multivariate logistic regression (forward LR) were 0.10
and 0.05, respectively. As for multivariate logistic regression
models, we assigned bad result (short or long term) as digit
one and good result (short or long term) as digit zero. Odds
ratios (ORs) were used to evaluate the strength of association
between risk factors and short-term or long-term outcome. All
statistical analyses were performed with SPSS software (version: PASW Statistics 18, 2009).

Results
Surgical outcome
All AVMs were exposed with standard pterional craniotomies. Complete removal of the AVM was documented by
intraoperative ultrasound and/or confirmed by postoperative
angiography in 40 patients, except 1 patient had tight adhesion between the nidus and the middle cerebral artery and
failed to be removed completely. One patient suffered from
postoperative surgical site intracerebral hemorrhage and
then evacuation of hematoma was performed. On the 15th
day after operation, this patient died from brain stem

Age
(year)
/sex

54/M
48/M
13/F
27/F
31/M
52/M
35/M
47/M
52/M
45/M
38/M
52/M
47/M
36/M
31/M
48/M
34/F
40/F

7/M
11/M
34/F
34/M
18/F
9/M
43/F
23/F
35/F
19/F
7/F
20/M
42/M
48/F
11/F
20/M

Patient NO.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34

Hemorrhage
Hemorrhage
Hemorrhage
Seizure
Hemorrhage
Hemorrhage
Hemiparesis
Seizure
Headache
Hemorrhage
Hemorrhage
Seizure
Hemiparesis
Headache
Hemorrhage
Headache

Hemorrhage
Seizure
Hemorrhage
Hemorrhage
Seizure
Seizure
Headache
Headache
Seizure
Seizure
Seizure
Seizure
Seizure
Seizure
Hemorrhage
Hemorrhage
Headache
Seizure

Presentation

Table 1 Summary of patients

Deep
Medial
Deep
Pure
Medial
Lateral
Pure
Medial
Medial
Pure
Deep
Lateral
Lateral
Lateral
Deep
Medial

Pure
Lateral
Deep
Lateral
Lateral
Medial
Pure
Medial
Lateral
Lateral
Lateral
Lateral
Lateral
Medial
Lateral
Medial
Lateral
Pure

Sugita

R
L
L
R
R
R
L
R
R
R
R
L
L
L
R
R

L
L
l
L
L
L
R
L
R
L
R
R
R
L
R
L
R
R

Side

2
2
2
2
2
2
2
1
2
2
1
1
2
2
2
2

2
2
2
1
2
2
1
3
2
2
2
2
2
2
3
2
2
2

Size

1
0
0
0
1
0
1
0
0
1
1
0
0
0
0
0

0
1
1
1
0
0
0
1
0
0
0
1
0
0
0
1
0
1

Drainage

1
1
1
0
1
1
0
0
0
0
1
1
1
1
1
0

0
0
1
1
0
1
1
1
0
0
0
0
0
1
0
1
0
0

Eloquence

4
3
3
2
3
3
3
1
2
3
3
2
3
2
3
2

1
3
3
3
2
3
2
5
2
2
1
2
2
4
3
3
2
3

Grade

3
2
0
1
4
0
1
1
1
1
3
1
0
1
1
1

4
1
3
3
1
1
1
1
1
1
0
1
1
1
1
0
1
0

Preoperative mRS

3
0
3
0
4
1
0
0
0
0
13
0
3
1
4
0

3
1
0
3
0
1
3
0
0
0
1
13
0
6
0
3
1
3

mRS(last
follow-up)

16
32
14
26
15
33
15
29
12
34
15
32
26
32
18
31

15
11.5
29
15
12
28
26
35
32
32
6
35
9
14
18
20
30
9

Duration of
follow-up
(month)

1
1
1
0
1
0
0
0
0
1
1
0
1
0
1
0

1
0
1
1
0
0
1
0
0
0
0
1
0
1
0
1
0
1

Status of postop

1
0
1
0
1
0
0
0
0
0
1
0
1
0
1
0

1
0
0
1
0
0
1
0
0
0
0
1
0
Death
0
1
0
1

Status of
follow-up

Neurosurg Rev (2013) 36:541–549
543

infarction. The patient was included in the postoperative
short-term evaluation and was excluded from the followup evaluation. Eighteen patients (43.9 %) had bad shortterm result and 23 (56.1 %) had good short-term result.
Eleven patients (27.5 %) had bad long-term result and 29
(72.5 %) had good long-term result (Table 1).
Illustrative case one
History and examination This 34-year-old man with no
significant medical history had suffered from medical intractable seizures for 1 year and was referred to our department. On admission the neurological examination disclosed
nothing positive. The patient underwent angiography studies, which revealed an AVM bordering the right sylvian
fissure (Fig. 1a, b). On orthotropic and lateral view, the right
ICA angiogram showed a Spetzler–Martin grade II AVM
(size, 2; deep venous drainage, 0; eloquence, 0) fed by a
dilated middle cerebral artery. The nidus drained into the
superior sagittal sinus.
Operation and postoperative course A standard right
pterional approach was performed and a large angioid lesion
(Fig. 1c) was encountered. A complete removal was
performed assisted by intraoperative real-time ultrasound.
Postoperative course of the patient was uneventful and no
neurological deficit occurred. Postoperative angiography
was performed 1 week after operation (Fig. 1 d, e), and
complete obliteration of the AVM was confirmed. At 1 year’s
follow-up, he was seizure-free and had recovered well.
Sugita Sugita Classification, L left, R right, mRS the modified Rankin Scale score

Grade: refers to the Spetzler–Martin grading system, with points assigned according to arteriovenous malformation diameter or size, deep venous drainage, and eloquence. Status of postop.—1
means new neurological dysfunction or deterioration of original neurological deficit appeared shortly after operation, 0 means no neurological dysfunction and mitigation of original neurological
deficit during 1 week after surgery

0
1
0
1
1
0
6
31
12
2
3
3
Lateral
Lateral
Lateral
15/M
17/F
8/M
39
40
41

Seizure
Hemorrhage
Hemorrhage

R
L
L

2
2
2

0
1
0

0
0
1

1
1
0

1
4
0

0
0
0
0
29
8
8
24
3
1
2
2
Pure
Medial
Medial
Medial
31/M
46/M
26/F
32/M
35
36
37
38

Seizure
Hemorrhage
Headache
Seizure

L
R
L
L

2
1
2
2

1
0
0
0

0
0
1
1

1
1
1
1

0
1
1
1

Status of postop
Duration of
follow-up
(month)
mRS(last
follow-up)
Preoperative mRS
Grade
Eloquence
Drainage
Size
Side
Sugita
Presentation
Age
(year)
/sex
Patient NO.

Table 1 (continued)

0
0
0
0

Neurosurg Rev (2013) 36:541–549
Status of
follow-up

544

Illustrative case two
History and examination This 35-year-old man with no
significant medical history had suffered from intermittent
headache for 7 years and aggravated in the recent 3 years.
On admission, the neurological examination disclosed nothing positive. The patient underwent angiography exam,
which revealed a right pure sylvian fissure AVM (Fig. 2a,
b). On orthotopic and lateral view, the right ICA angiogram
showed a Spetzler–Martin grade II AVM (size, 1; deep
venous drainage, 0; eloquence, 1) fed by a dilated middle
cerebral artery. The nidus drained into the superior sagittal
sinus and inferior petrosal sinus.
Operation and postoperative course A standard right
pterional approach was performed and a large angioid lesion
was encountered. A complete removal was performed
assisted by intraoperative real-time ultrasound. The patient
had left hemiplegia after surgery and did not recover at
discharge. Postoperative MRI was performed 1 week after
operation (Fig. 2 c, d). At 2 years’ follow-up, he still had left

Neurosurg Rev (2013) 36:541–549

545

Fig. 1 a, b The right ICA
angiogram showed a Spetzler–
Martin grade II AVM fed by a
dilated middle cerebral artery. c
Intraoperative screenshot
obtained after opening the
sylvian fissure through the right
pterional approach showed the
nidus was in the sylvian fissure.
d, e Postoperative angiography
confirmed complete obliteration
of the AVM

upper extremity hemiplegia and the mRS scores of followup were 3.

Risk factors associated with short-term and long-term result
in the univariate analysis
Compared with the preoperative mRS score of the patients
with and without previous bleeding by the rank-sum test, there
was no statistically significant difference between the two
groups (P value 0.09). The significance of each factor affecting the short-term and long-term result is shown in Table 2.

Previous bleeding, eloquence of area involved, AVMs with
deep venous drainage, age≤18, Sugita classification, and
Spetzler–Martin grade classification were significant risk factors for short-term result by univariate analysis. Previous
bleeding, eloquence of area involved, and Spetzler–Martin
grade classification were significant risk factors for longterm result by univariate analysis.
Multivariate logistic regression
The multivariate logistic regression was on the base of results
from the univariate analysis of risk factors associated with

546

Neurosurg Rev (2013) 36:541–549

Fig. 2 a, b The right ICA
angiogram shows a Spetzler–
Martin grade II AVM fed by a
dilated middle cerebral artery. c,
d Postoperative MRI with
enhancement confirmed
complete removal of the AVM
and showed infarction on the
right basal ganglia area

short-term and long-term result. Seven factors, including age,
sex, previous bleeding, deep venous drainage, eloquence of
area involved, Sugita classification, and Spetzler–Martin classification, entered the mode of the parameter estimate associated with short-term result (Table 3). Previous bleeding (OR=
8.211, 95 % CI=1.760, 38.312) and deep venous drainage
(OR=4.869, 95 % CI=0.991, 23.92) remained significant risk
factors associated with short-term result. The incidence of
postoperative transient neurological deterioration for patients
with a ruptured AVM is as 8.211(95 % CI) times bigger as
patients with an unruptured AVM. Patients with deep venous
drainage were significantly more likely than those with superficial venous drainage to have postoperative transient neurological deterioration.
Sex, previous bleeding, eloquence of area involved, and
Spetzler–Martin classification entered the model of the parameter estimate associated with long-term result at the P<
0.1. Previous bleeding (OR = 7.000, 95 % CI = 1.476,
33.207) remained significant risk factors associated with
the long-term result in the multiple regression model
(Table 4), which manifested that the incidence of bad
long-term result for patients with a ruptured AVM is 7
(95 % CI) times bigger as compared to patients with an
unruptured AVM.

Discussion
Based on our results, despite different Sugita classification subtypes, sylvian fissure AVMs patients’ prognosis
with microsurgical resection are better than expected; a
history of AVM bleeding is a risk factor for postoperative temporary neurological deterioration and for longterm adverse outcome, while the AVM deep venous
drainage is a risk factor only for temporary neurological
deterioration.
Previous hemorrhage in common intracerebral AVMs
perhaps helps neurosurgeons to dissect the margin between
the nidus and surrounding parenchyma [6]. However,
sylvian fissure AVM which locates in or adjacent to the
sylvian fissure involves the middle cerebral artery and is
surrounded by critical structures such as the basal ganglia
and internal capsule. A prior bleeding from AVM frequently
cause hematoma or subarachnoid hemorrhage (SAH) in
sylvian fissure [10], which later always lead to adhesion
and blurred boundary between the nidus and middle cerebral
artery or these surrounding critical structures. These will
bring extra difficulty of visualization and dissection of the
AVM planes. Obscure anatomy might cause unexpected
injury to critical structure and accidental intraoperative

Neurosurg Rev (2013) 36:541–549

547

Table 2 Risk factors associated with short-term and long-term results (chi-square test)
Risk factors

Short-term result

Long-term mRS

Bad (No., %)

Good (No., %)

8 (80.0)
10 (32.3)

2 (20.0)
21 (67.7)

8 (32.0)

17 (68.0)

Female
10 (62.5)
Previous bleeding
Yes
12 (75.0)
No
6 (24.0)
Drainage
Deep
10 (71.4)
Superficial
8 (29.6)
Eloquence
Yes
12 (60.0)
No
6 (28.6)
Sugitaa

6 (37.5)

Agea
≤18
>18
Sex
Male

Pure
Lateral
Medial
Deep
Diameter
0–3 cm
3–6 cm

P

0–2 (No., %)

0.012

3.685

0.055

3–6 (No., %)

χ2

P

6 (60.0)
23 (76.7)

4 (40.0)
7 (23.3)

0.418

3.532

0.080

20 (83.3)

4 (16.7)

9 (56.3)

7 (43.8)

4 (25.0)
19 (76.0)

10.303

0.001

8 (50)
21 (87.5)

8 (50)
3 (12.5)

6.771

0.009

4 (28.6)
19 (70.4)

6.540

0.011

8 (61.5)
21 (77.8)

5 (38.5)
6 (22.2)

1.161

0.281

8 (40.0)
15 (71.4)

4.108

0.043

11 (57.9)
18 (85.7)

8 (42.1)
3 (14)

3.872

0.049

4 (57.1)
14 (82.4)
9 (81.8)
2 (20)

3 (42.9)
3 (17.6)
2 (18.2)
3 (60)

0.188

4
23

2
9

21.67
20.63

0.655

2

0

15.00

4 (57.1)
5 (29.4)
4 (33.3)
5 (100.0)

3 (42.9)
12 (70.6)
8 (66.7)
0 (0)

3
15

3
18

22.25
21.32

2

12.00

15
7
1

16.32
24.95
25.67

0.008

17
11
1

2
8
1

17.11
23.42
25.00

0.024

2
8
9
4

18.83
21.57
19.32
23.39

0.709

3
10
10
6

0
4
4
3

15.00
20.71
20.71
21.67

0.717

6–cm
0
Spetzler–Martin
I,II
4
III
12
IV,V
2
Bleeding in surgery (ml)
0–100
1
100–500
7
500–1,000
5
1,000õ
5
a

χ2

0.031

0.391

Fisher’s test

AVM rupture. In addition, hematoma and SAH would compress the surrounding important structures or cause following possible cerebral vasospasm, both possibly leading to
long-term neurological deficit [11]. These might explain
why a bleeding history was a risk factor for short-term and
long-term adverse prognosis for sylvian fissure AVMs.
Deep venous drainage is another risk prognosis factor for
sylvian fissure AVM in the short term. Sylvian fissure
AVMs with deep venous drainage are always located in
the deep brain and often share blood drainage with basal

Table 3 parameter estimate associated with short-term result
Parameter estimate S.E.
Bleedinga 2.105
Drainageb 1.583
Constant −1.640

P

OR

95 % CI

0.786 0.007 8.211 1.760,38.312
0.812 0.051 4.869 0.991,23.922
0.578 0.005 0.194

a

Without a history of symptomatic AVMs bleeding=0; with a history
of symptomatic AVMs bleeding=1

b

Superficial venous drainage=0; deep venous drainage=1

548

Neurosurg Rev (2013) 36:541–549

Table 4 Parameter estimate associated with long-term result
Parameter estimate S.E.
Bleedinga 1.946
Constant −1.946

P

OR

95 % CI

0.794 0.014 7.000 1.476, 33.207
0.617 0.002 0.143

a

Without a history of symptomatic AVMs bleeding=0; with a history
of symptomatic AVMs bleeding=1

ganglia and thalamus. Deep location of the AVM nidus [8]
and the presence of deep venous drainage [3] have been
regarded as risk factors for hemorrhage from an AVM.
Additionally, shortly after nidus was surgically removed,
vein drainage from surrounding structure might be also
affected causing transient neurological deficits. As time
passed by, the venous drainage might redistribute. This
might explain why the deep venous drainage had significant
effect on the short-term outcome.
Interestingly, some factors such as eloquent area involved, Spetzler–Martin grade classification, and Sugita
classification show significant (P<0.05) in the univariate
analysis; however, these factors did not present statistic
significance (P<0.05) in the multivariate logistic regression
model finally. The model always allows the most significant
factor to enter preferentially. Considering the interaction
among all factors, the three factors might not be powerful
enough to influence the prognosis and just manifest certain
tendency between this single factor and the unfavorable
prognosis. Their more detailed effects on the prognosis need
to be explored in a larger population.
Bleeding is more likely to occur to AVM patients with a
history of bleeding or deep venous drainage [9], which may
support choosing more invasive treatment for this kind of
patients. However, from our results, sylvian fissure AVM
patients with a prior bleeding or deep venous drainage might
have bad long-term or short-term outcomes if they select
microsurgery. It seems to be a dilemma. In this situation,
decision-making for surgery should be much more cautious.
Limitation
Indeed, previous AVM bleeding may cause preoperative neurological deficits and will affect the long-term prognosis. We
compared the preoperative mRS score of the patients with and
without previous bleeding; we found that there was no statistically significant difference between the two groups. However, the P value was 0.09, which is approaching the P value
0.05. It is also possible that the significant difference of
preoperative status of patients between two groups might
exist, which need a larger population to prove. At the same
time, we realized that previous bleeding might not only affect
preoperative patient’s status but also increase surgical difficulties in the operation. Both might be as a whole leading to

their bad short- or long-term surgical results. Additionally,
some factors such as arterial injury, venous infarction, vasospasm, etc., might play a very important role and affect the
postoperative outcome; these factors will be evaluated in a
future prospective study.
Statistical analysis only provide mathematical logic,
which cannot clarify the underlying reasons that a history
of AVM bleeding and AVM deep venous drainage is superior to other factors to affect the prognosis. Meanwhile, a
larger numbers of sylvian fissure AVM patients treated with
microsurgery with long-term outcomes are required for better prognosis evaluation in the future.

Conclusion
Despite different Sugita classification subtypes, sylvian fissure
AVMs’ results with microsurgical resection are better than
expected; a history of AVM bleeding is a risk factor for postoperative temporary neurological deterioration and for longterm adverse outcome, while the AVM deep venous drainage is
a risk factor only for temporary neurological deterioration.
Acknowledgments We are greatly indebted to Dr. Yang Wang for
their precious statistic assistance.
Funding This study is supported by a grant “National Science and
Technology Support Program (No: 2011BAI08B08)” (Principal Investigators, Professor Shuo Wang) from the Ministry of Health, China.

References
1. ApSimon HT, Reef H, Phadke RV, Popovic EA (2002) A
population-based study of brain arteriovenous malformation:longtreatment outcomes. Stroke 33(12):2794–800
2. Brown RD Jr, Wiebers DO, Torner JC, O’Fallon WM (1996) Incidence and prevalence of intracranial vascular malformations in Olmsted County, Minnesota, 1965 to 1992. Neurology 46(4):949–5
3. Duong DH, Young WL, Vang MC, Sciacca RR, Mast H,
Koennecke HC, Hartmann A, Joshi S, Mohr JP, Pile-Spellman J
(1998) Feeding artery pressure and venous drainage pattern are
primary determinants of hemorrhage from cerebral arteriovenous
malformations. Stroke 29:1167–1176
4. Hillman J (2001) Population-based analysis of arteriovenous malformation treatment. J Neurosurg 95(4):633–7
5. Laakso A, Hernesniemi J (2012) Arteriovenous malformations:epidemiology and clinical presentation. Neurosurg Clin N Am 23:1–6
6. Lawton MT, Du R, Tran MN, Achrol AS, McCulloch CE, Johnston
SC, Quinnine NJ, Young WL (2005) Effect of presenting hemorrhage
on outcome after microsurgical resection of brain.arteriovenous
malformations. Neurosurgery 56(3):485–9
7. Lawton MT, Lu DC, Young WL (2007) Sylvian fissure arteriovenous malformations:an application of the Sugita classification to
28 surgical patients. Neurosurgery 61(1):29–36
8. Mast H, Young WL, Koennecke HC, Sclacca RR, Osipov A,
Pile-Spellman J, Hacein-Bey L, Duong H, Stein BM, Mohr JP

Neurosurg Rev (2013) 36:541–549
(1997) Risk of spontaneous haemorrhage after diagnosis of
cerebral arteriovenous malformation. Lancet 350:1065–1068
9. Pollock BE, Flickinger JC, Lunsford LD, Bissonette DJ, Kondziolka
D (1996) Factors that predict the bleeding risk of cerebral arteriovenous malformations. Stroke 27(1):1–6
10. Qureshi AI, Mendelow AD, Hanley DF (2009) Intracerebral
haemorrhage. Lancet 373:1632–1644
11. Schmieder K, Moller F, Engelhardt M, Scholz M, Schregel W,
Christmann A, Harders A (2006) Dynamic cerebral autoregulation
in patients with ruptured and unruptured aneurysms after induction
of general anesthesia. Zentralbl Neurochir 67:81–87
12. Stapf C, Mast H, Sciacca RR, Berenstein A, Nelson PK, Gobin YP,
Pile-Spellman J, Mohr JP; New York Islands AVM Study Collaborators (2003) The New York Island’s AVM study: design, study
progress, and initial results. Stroke 34(5):e29–33
13. Sugita K, Takemae T, Kobayashi S (1987) Sylvian fissure
arterivenous malformations. Neurosurgery 21:7–14
14. van Swieten JC, Koudstaal PJ, Visser MC, Schouten HJ, van Gijn J
(1998) Interobserver agreement for the assessment of handicap in
stroke patients. Stroke 19(5):604–7
15. Wang S, Zhao HY, Zhao YL (2011) Microsurgical treatment for
sylvian fissure arteriovenous malformations. Zhonghua Yi Xue Za
Zhi 91:1609–12
16. Zimmerman G, Lewis A, John M, Tew JM Jr (2000) Pure sylvian
fissure arteriovenous malformations. J Neurosurg 92:39–44

Comments
Feres Chaddad, Jose Maria Campos Filho, Evandro de Oliveira, São
Paulo, Brazil
The authors provide an impressive series of sylvian fissure AVMs
in a short space of time, which constitute a challenging subgroup of
lesions.
The Spetzler–Martin grade has provided the orientation about the
surgical risk of AVMs. Its size, distribution along the sylvian fissure
and the relation between the nidus and the middle cerebral artery
clearly define the surgical morbidity. The correlation of AVM with
the insula lobe, the frontal, parietal and temporal cortex influences the
choice in treatment and patients´ outcome. Moreover, the AVMs supplied by the lateral lenticulostriate arteries show higher morbidity than
others because such arteries are related to the central core blood supply.
The deep venous drainage is a clear risk to neurological deterioration because the former indicates the impairment of the insula’s lobe
drainage veins.
The surgical management of AVMs demands the integrated approach to a specially trained neurosurgeon, an interventional neuroradiologist and a radiosurgeon working as a team. Each AVM is studied
in order to achieve the best treatment. Thus, they can be surgically
treated, a small number can be treated with radiosurgery and a combination of embolization can be used for some of them before surgery.
Hideyuki Yoshioka, Hiroyuki Kinouchi, Yamanashi, Japan
Arteriovenous malformations (AVMs) localized in the sylvian fissure constitute a challenging subgroup of lesions to manage surgically.
In this article, Liu et al. reported their surgical experience of 41 patients
with sylvian fissure AVM, which is one of the biggest series of this rare
entity of AVM so far, and presented the excellent surgical outcome.
Complete removal of AVM was achieved in all but one case, and
72.5 % of the patients showed good long-term outcome.
The authors retrospectively analyzed their short-term and long-term
surgical outcomes. In the multivariate logistic regression analysis,
previous bleeding and deep venous drainage were associated with
temporary neurological deterioration. For long-term outcome, previous
bleeding was related to poor outcome. It is of interest that eloquency of

549
the lesion, Spetzler–Martin grade, and Sugita classification were not a
significant risk factor in their analysis.
Although this article provided practically important information,
there are some limitations in the study. As the authors pointed out, the
factors regarding bleed or deep venous drainage affecting the postoperative poor outcome were not clarified. Brain parenchymal injury,
intraoperative AVM rupture, arterial injury, venous infarction, and
cerebral vasospasm were considered to worsen neurological symptoms
after AVM surgery1. Information about such factors in this large series
of sylvian fissure AVM are valuable for the neurosurgical community,
and such profound analyses would be encouraged in future.
References
1. Pik JH, Morgan MK (2000) Microsurgery for small arteriovenous malformations of the brain: results in 110 consecutive patients.
Neurosurgery 47:571–575
Thomas Kretschmer, Oldenburg, Germany
Sylvian fissure AVMs are a rare type (around 10 % of AVMs),
whose microsurgical removal demands vascular expertise. They have
been dealt with in the literature before (E.G. Lawton, Lu, Young,
University of California, San Francisco Brain Arteriovenous Malformation Study Project, 2007). Additional information with regard to
evaluation and risk assessment of natural history and surgical risk
assessment and outcome analysis is very useful and welcome. The
presented series of 355 AVM overall is rather large and has been
operated on within a remarkable short time intervall of only 2.5 years
by a single neurosurgeon.
The findings of this series with regard to short- and long-term
outcome therefore are precious. Statistical analysis underlines the
higher incidence of neurological deficits as compared to AVM treatment in other locations; however, it also confirms that prognosis after
surgery also was better than expected. Important results to me were:
one third had an unfavorable outcome. The incidence of postoperative
transient neurological deterioration with a ruptured sylvian fissure
AVM was 8 times higher as compared to unruptured ones. Ruptured
AVMs had a 7 times larger chance for a bad long-term result. The
mentioned other sylvian fissure series demonstrated slightly different
outcome results (Lawton et al. 2007).
I congratulate the authors for this remarkable series and report
thereof.
Yasser Orz, Alexandria, Egypt
The arteriovenous malformations (AVM) localized in the sylvian
fissure constitute a challenging subgroup of lesions. The authors reported
their experience with 41 patients harboring such lesions according to
Sugita’s Classification, out of 355 patients with AVM treated with microsurgical resection at their institute between June2009–Dec. 2011. They
analyzed their results stressing on risk factors responsible for short-term
and long-term bad outcome according to mRS.
In the univariate analysis, previous bleeding, eloquence of area
involved, and Spetzler–Martin grade were significant risk factors for
long-term bad outcome; but in multivariate logistic analysis, only the
previous bleeding was a risk factor for long-term bad outcome.
Sugita’s classification, deep venous drainage, and age are risk factors
for transient short-term bad outcome.
Although they showed excellent results with microsurgical resection where complete removals of AVM were achieved in all patients
except one and only one patient died, a number of points already
mentioned by the authors deserve reemphasis. The Sugita’s classification does not guide the surgical approach, predict long-term surgical
outcome, or affect the clinical decision process. Adjuvant treatment
modalities (embolization, radiosurgery) may be very useful for
targeting deep feeders or deep portions of AVMs thereby increasing
the safety of surgical resection.
The authors should be congratulated for such good surgical results
and thoughtful analysis of their work.