RESEARCH—HUMAN—CLINICAL STUDIES

TOPIC Research—Human—Clinical Studies

Language and Motor Mapping During Resection of

Brain Arteriovenous Malformations: Indications,
Feasibility, and Utility
Andreu Gabarrós, MD*§
William L. Young, MD*†‡
Michael W. McDermott, MD*
Michael T. Lawton, MD*‡
*Department of Neurological Surgery,
University of California, San Francisco,
California; †Department of Anesthesia
and Perioperative Care, University of
California, San Francisco, California;
‡Center for Cerebrovascular Research,
University of California, San Francisco,
California; §Department of Neurological
Surgery, University Hospital of Bellvitge,
University of Barcelona, Spain
Correspondence:
Michael T. Lawton, MD,
Department of Neurological Surgery,
University of California,
San Francisco,
505 Parnassus Avenue,
M780, Box 0112,
San Francisco, CA 94143-0112.
E-mail: lawtonm@neurosurg.ucsf.edu
Received, December 14, 2009.
Accepted, August 23, 2010.
Copyright ª 2011 by the
Congress of Neurological Surgeons
BACKGROUND: Microsurgical resection of arteriovenous malformations (AVMs) located
in the language and motor cortex is associated with the risk of neurological de-
terioration, yet electrocortical stimulation mapping has not been widely used.
OBJECTIVE: To demonstrate the usefulness of intraoperative mapping with language/
motor AVMs.
METHODS: During an 11-year period, mapping was used in 12 of 431 patients (2.8%)
undergoing AVM resection (5 patients with language and 7 patients with motor AVMs).
Language mapping was performed under awake anesthesia and motor mapping under
general anesthesia.
RESULTS: Identification of a functional cortex enabled its preservation in 11 patients
(92%), guided dissection through overlying sulci down to the nidus in 3 patients (25%),
and influenced the extent of resection in 4 patients (33%). Eight patients (67%) had
complete resections. Four patients (33%) had incomplete resections, with circum-
ferentially dissected and subtotally disconnected AVMs left in situ, attached to areas of
eloquence and with preserved venous drainage. All were subsequently treated with
radiosurgery. At follow-up, 6 patients recovered completely, 3 patients were neuro-
logically improved, and 3 patients had new neurological deficits.
CONCLUSION: Indications for intraoperative mapping include preoperative functional
imaging that identifies the language/motor cortex adjacent to the AVM; larger AVMs
with higher Spetzler-Martin grades; and patients presenting with unruptured AVMs
without deficits. Mapping identified the functional cortex, promoted careful tissue
handling, and preserved function. Mapping may guide dissection to AVMs beneath
the cortical surface, and it may impact the decision to resect the AVM completely. More
conservative, subtotal circumdissections followed by radiosurgery may be an alternative
to observation or radiosurgery alone in patients with larger language/motor cortex AVMs.
KEY WORDS: Arteriovenous malformation, Electrocortical stimulation mapping, Language mapping, Micro-
surgery, Motor mapping

Neurosurgery 68:744–752, 2011 DOI: 10.1227/NEU.0b013e318207a9a7 www.neurosurgery-online.com

A
rteriovenous malformations (AVMs)
located in the motor and language cortex
are associated with significant risk of
neurological deterioration, especially in patients
presenting with unruptured AVMs and minimal
or no neurological deficits.1,2 Consequently,
ABBREVIATIONS: AVM, arteriovenous malforma-
tion; fMRI, functional magnetic resonance imaging;
mRS, modified Rankin Scale; MSI, magnetic source
imaging
these patients are typically managed conserva-
tively or treated radiosurgically.3,4 Motor or
language function may shift in the presence of an
AVM from its anatomic location to an adjacent
gyrus or even to the contralateral hemisphere.1,5-11
Increasingly, sophisticated neuroradiological
imaging can detect these shifts in functional
localization with functional magnetic resonance
imaging (fMRI), magnetoencephalography, mag-
netic source imaging (MSI), positron emission
tomography (PET), and Wada testing. Finding an
unexpected separation between eloquent cortex

744 | VOLUME 68 | NUMBER 3 | MARCH 2011 www.neurosurgery-online.com

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.


and AVM nidus on one of these imaging modalities may encourage
more aggressive intervention with surgical resection, but the reso-
lution of these studies is often low and significant shifts in func-
tional localization occur infrequently. Therefore, preoperative
functional imaging serves only as a screening test for major re-
organization of the eloquent cortex.
Electrocortical stimulation mapping is widely used with intrinsic
brain tumors, facilitating more extensive resections than otherwise
possible and increasing patient survival. However, electrocortical
stimulation mapping has not been widely used with AVMs,
probably because compact AVMs do not contain brain tissue
within the nidus, and they typically have a distinct plane between
their borders and adjacent brain. In addition, AVM resections can
be more technical, tedious, and prone to bleeding than glioma
resections, which makes them more daunting in awake AVM
patients than in awake glioma patients. In reality, a significant
portion of the dissection around an AVM is not in a subarachnoid
plane and instead is in a parenchymal plane beyond pia mater.
Furthermore, a significant number of AVMs have a border or
borders that are diffuse rather than compact, forcing the dissection
into brain parenchyma to completely encircle the nidus. Even
minor pial violations and parenchymal invasions can be costly
in the motor and language cortex. Stimulation mapping provides
an intraoperative technique for identifying and preserving func-
tional cortex around motor and language area AVMs. Nonetheless,
published experiences are few and small in size.
Intraoperative brain mapping requires a coordinated team of
neurosurgeons, neuroanesthesiologists, and neurophysiologists
that does not exist at many centers. The high volume of intrinsic
brain tumors treated with intraoperative mapping at the
University of California, San Francisco makes this team readily
available for AVM patients. We applied these mapping techni-
ques to selected motor and language area AVMs to evaluate their
indications, feasibility, and utility. We are confident operating
on small AVMs in these areas, but have advocated caution with
medium-sized AVMs, the so-called grade III+ AVMs.12 We
hypothesized that intraoperative stimulation mapping may be one
method to improve outcomes in these and higher Spetzler-Martin
grade AVMs in these eloquent territories.
PATIENTS AND METHODS
The study was approved by the University of California, San Francisco
Committee on Human Research and conducted in compliance with
Health Insurance Portability and Accountability Act regulations. During
an 11-year period, 431 patients underwent microsurgical resection of
their AVMs by a single neurosurgeon (M.T.L.), and intraoperative brain
mapping was used in 12 of these patients (2.8%). Data were obtained
from an ongoing registry of AVM patients treated at our institution,
maintained prospectively as part of the University of California, San
Francisco, Brain Arteriovenous Malformation Study Project.
Preoperative Evaluation
Preoperatively, all patients underwent complete neurological exami-
nation and a baseline modified Rankin Scale score was assigned. All
NEUROSURGERY
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
BRAIN MAPPING OF AVMS
patients underwent CT scanning, MR imaging, catheter angiography,
and fMRI and/or MSI. Patients with AVMs in the language cortex
underwent language testing (naming, counting, reading, and verbal
fluency tests) by a neuropsychologist. Patients with AVMs less than
10 mm from the motor or language cortex, as determined by fMRI and/
or MSI, and with normal neuropsychologic language testing (.90%
proficiency), were selected for intraoperative brain mapping. Of the
12 patients in this study, 7 had motor and 5 had language area AVMs.
Anesthetic Considerations
Patients with motor cortex AVMs were operated on with general
anesthesia; and patients with language cortex AVMs were operated on
‘‘awake.’’ During craniotomy and surgical exposure for awake cases,
patients were sedated with intravenous agents that included propofol
(50-100 mg
kg21
min21) and remifentanil (0.05-0.2 mg
kg21
min21).
Local anesthesia with lidocaine was used for scalp and temporal muscle.
After the bone flap was removed, dura was locally infiltrated with li-
docaine. Before mapping, sedatives were discontinued and patients were
awakened under the care of the anesthesiologist. In patients operated on
with general anesthesia, muscle relaxants were not used other than during
endotracheal intubation, and a low dose of potent agent, eg, desflurane
(2-3%) and intravenous propofol infusion (50-125 mg
kg21
min21) was
maintained until the end of the surgery.
Cortical Stimulation Mapping
The cortical surface was widely exposed and stimulation was per-
formed using standard, bipolar electrodes (5 mm). Constant current
generator was set to deliver biphasic square wave pulses of 4-ms duration
at 60 Hz. Stimuli were initiated at 1.5 mA and increased to a maximum
of 8 mA. Intraoperative monitoring using somatosensory evoked po-
tentials was also used in all cases. Sterile numbered labels were used to
identify the stimulation sites.
During awake surgery, patients were asked to repeat 3 language tasks
tested preoperatively: counting numbers from 1 to 50, naming objects on
a computer screen, and reading single words projected sequentially on
a computer screen. Cortical sites were stimulated 3 times for 3-second
durations, and language deficits (eg, counting errors, speech arrest, or
alexia) during at least 2 of the 3 stimulations in the same site identified
a functional speech area. A neuropsychologist conducted all preoperative
and intraoperative language testing. Corticography was routinely used
during awake surgery to avoid language errors owing to subclinical
seizure activity. Once the sites were identified, sedative drugs were
restarted for patient comfort. In the general anesthesia group (asleep),
stimulated sites on the motor strip showed positive motor responses
in muscles of the contralateral side of the body (eg, foot, leg, arm,
hand, or face). Intraoperative seizures induced by stimulation were
treated with topical ice-cold Ringer’s solution and intravenous propofol
(1 mg/kg bolus).
Patient Outcomes
Neurological outcome was assessed using the modified Rankin Scale
(mRS). A single clinical nurse, under the supervision of a neurologist,
performed all clinical assessments before any treatment, preoperatively,
at 6 months postoperatively, and during the follow-up period. All
patients had follow-up data within 6 months of analysis. Patients with
AVM-related seizures that were controlled with medications, and
without any other neurological symptoms, were assigned a mRS score
of 0. Alternatively, neurologically intact patients with refractory
VOLUME 68 | NUMBER 3 | MARCH 2011 | 745
GABARRÓS ET AL

AVM-related seizures that interfered with daily activities (working,
school, driving, etc) were assigned a score of 2. Surgical morbidity was
assessed by comparing mRS 6 months postoperatively with the mRS
immediately before surgery.
RESULTS
Patients and AVM Characteristics
Twelve patients with AVMs located in the motor or language
cortex had a mean age of 31 years (Table 1). There were 5 men
and 7 women. Only 3 patients (25%) presented with hemor-
rhage. AVM sizes ranged from 1.5 cm to 6.0 cm (mean size,
3.7 cm). Patients with Spetzler-Martin grade III+ AVMs were the
largest subgroup (5 patients), and 3 patients had high-grade AVMs.
Ten patients underwent preoperative AVM embolization,
and 3 of these patients were embolized twice. Four patients
had undergone radiosurgery previously, with incompletely
obliterated AVMs.
In 3 patients (25%), the AVM was deep to the cortical surface
and mapping guided the dissection through overlying sulci down
to the nidus. In addition, a small lobule of cortex was resected in
each of these cases and mapping helped avoid the language cortex
in 2 cases and the motor cortex in 1 case (Table 2).
In 4 patients (33%), brain mapping influenced the attitude
toward the extent of resection. Two patients had AVM borders
that were too intimately associated with the functional cortex to
complete the resection safely. Two patients had deep, conical
extensions of nidus into white matter near the internal capsule
and deep perforating artery supply from lenticulostriate arteries
and branches of the posterior cerebral artery. Further dissection
across these areas was aborted. In these cases of incomplete re-
section, the nidus was circumferentially dissected in all planes
except those associated with functional sites, disconnecting
most arterial feeders while preserving the major draining vein.
Dearterialized AVMs were left in situ with the intent to treat
with radiosurgery.

Utility of Brain Mapping Surgical Complications

Cortical mapping was considered useful in all 12 patients. The
language cortex was identified in all 5 patients undergoing awake
craniotomy (Figures 1 and 2), and the motor cortex was also
identified in one of these patients (Figure 3) (Table 2). The motor
cortex was identified in 6 of 7 patients undergoing asleep cra-
niotomy. Identification of the functional cortex enabled its
preservation during AVM resection in 11 patients (92%), with
care during dissection not to transgress pial planes. The inability
to identify the motor cortex in one patient was still considered
useful because it enabled confident dissection in the parenchymal
planes around the nidus. A strip electrode was placed posteriorly
in this patient to try to identify the motor cortex, but this was
unsuccessful.
Two patients, both asleep, had intraoperative seizures during
motor mapping and both were easily controlled with cold
Ringer’s solution and intravenous propofol. Two patients (17%)
had hemorrhagic complications (Table 3). An intracerebral
hemorrhage in the resection cavity indicated a small piece of
residual AVM that was resected at the time of hematoma evac-
uation. The other patient had an epidural hematoma that was
evacuated urgently without neurological sequelae. Median length
of stay was 9 days (range, 4-30 days).
Patient Outcomes
Eight patients (67%) had a complete AVM resection con-
firmed by angiography postoperatively (Table 3). Four patients

TABLE 1. Patient Demographics and AVM Characteristicsa

Patients Age, y/Sex Presentation Location Size, mm S/M Gradeb Prior Radiosurgery Embolization
Awake
1 28/F Seizures L Frontoparietal 47 IV (S2V1E1) None 0
2 15/F Hemorrhage L Insular 28 III2 (S1V1E1) Yes 1
3 39/M Seizures L Temporal, Insular 47 III+ (S2V0E1) None 1
4 30/F Headache L Temporal 48 III+ (S2V0E1) None 1
5 43/M Seizures L Frontal 31 III+ (S2V0E1) None 1
Asleep
6 38/F Headache L Precentral 50 IV (S2V1E1) None 1
7 35/M Hemorrhage R Rolandic 50 III+ (S2V0E1) Yes (2) 2
8 33/F Headache L Rolandic 17 III2 (S1V1E1) None 2
9 24/M Headache R Postcentral 33 III+ (S2V0E1) None 2
10 17/F Headache L Rolandic 15 II (S1V0E1) Yes 0
11 42/M Hemorrhage, seizures L Rolandic 21 III2 (S1V1E1) None 1
12 32/F Seizures L Precentral 60 V (S3V1E1) Yes 1
a
S/M, Spetzer-Martin; F, female; M, male; L, left; R, right.
b
According to modified Spetzler-Martin scale for grade III AVMs.12

746 | VOLUME 68 | NUMBER 3 | MARCH 2011 www.neurosurgery-online.com

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.

 BRAIN MAPPING OF AVMS

FIGURE 1. Patient 4. Axial T2-weighted magnetic resonance image (A) and sagittal T1-weighted MR image (B) demonstrated
this left temporal AVM adjacent to Wernicke’s and Broca’s areas. Digital subtraction angiography [left internal carotid artery
injection, anteroposterior (C) and lateral (D) views] demonstrated a Spetzler-Martin grade III+ AVM. AVM, arteriovenous
malformation.

(33%) had incomplete resections, with circumferentially dis-
sected and subtotally disconnected AVMs left in situ, attached to
areas of functional eloquence and with preserved venous drainage.
Incompletely resected AVMs included 2 Spetzler-Martin grade
IV lesions and one grade V lesion, as well as a modified grade III-
lesion. Residual AVM was confirmed on postoperative angiog-
raphy in these patients, and all were treated subsequently with
Gamma knife radiosurgery. Two of these patients are within the
3-year latency period after radiosurgery. One patient’s AVM was
successfully obliterated with radiosurgery. One patient (patient 1)
had a new hemorrhage 1 year after radiosurgery, with associated
deterioration consisting of hemiparesis and expressive aphasia.
She underwent embolization and complete resection.
Nine patients deteriorated postoperatively, reflecting the high
Spetzler-Martin grades of these AVMs and their proximity to
eloquent cortex (Table 3). Two patients were unchanged and
1 patient with a ruptured AVM improved. The mean post-
operative mRS was 2.9, compared with a mean preoperative mRS
score of 1.3. However, much of this deterioration was transient,
with a mean mRS score of 1.3 three months after surgery. At late
follow-up (mean follow-up duration, 3.3 y), 6 patients (50%)
recovered completely and 3 patients (25%) were neurologically

NEUROSURGERY VOLUME 68 | NUMBER 3 | MARCH 2011 | 747

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.

GABARRÓS ET AL

FIGURE 2. Patient 4. A, intraoperative mapping identified Wernicke’s area (48) and Broca’s area (40), both adjacent to the AVM. B, a distinct subarachnoid dissection
plane enabled the nidus to separate from Broca’s area. C, the dissection plane was less distinct with Wernicke’s area (behind the retractor blade). D, the nidus was
circumferentially dissected and the draining vein darkened. E, angular artery was skeletonized to preserve normal en passage arteries to distal territories. AVM, arteriovenous
malformation.

improved relative to their preoperative mRS score. Three patients
(25%) remained with new neurological deficits. Nine patients
(75%) had good outcomes (living independently with mRS # 2).
There was no mortality in this series.
Of the 5 patients presenting with seizures, only 1 patient had
seizures at late follow-up evaluation. Three patients had seizures
during the perioperative period and none had seizures at the
late follow-up.
DISCUSSION
The safety of electrocortical stimulation mapping during
surgery for brain AVMs has been reported previously,13-19 and
our small experience with 12 AVMs in motor and language areas
confirms its feasibility and safety. There were no complications
associated with its use. Motor mapping required minimal alter-
ation in the usual operative routine, but the neurosurgeon must
be prepared to control intraoperative seizures. Language mapping
required significant alteration in the usual operative routine, but
there were no awkward circumstances for the neurosurgeon and
none of the patients experienced any significant discomfort
or complained about awake anesthesia. We did not develop
novel mapping techniques, but our experience with established
techniques has encouraged us to be more surgically minded in
our management of carefully selected language and motor
area AVMs.
Indications for Mapping With AVMs
Preoperative functional imaging was performed in all 12 of our
patients and it demonstrated proximity of the AVM to the
language or motor cortex. Many functional imaging options are
available, and we have used PET scans, MSI, and fMRI. Our
current imaging modality of choice is fMRI because it is easy to
acquire, is convenient to review on our hospital network, and
enables excellent correlation with brain and AVM anatomy.
fMRI is based on changes in imaged blood flow with functional
activation, and we have found, like others, that abnormal
hemodynamics of AVMs interfere with functional imaging,
overestimating the area of functional cortex.6,7,20,21 However, the
information from fMRI is sufficient to apply the algorithm that

748 | VOLUME 68 | NUMBER 3 | MARCH 2011 www.neurosurgery-online.com

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.

 BRAIN MAPPING OF AVMS

entire nidus, and dissection rarely stops short of complete re-

TABLE 2. Utility of Language and Motor Mapping During AVM section. In contrast, high-grade AVMs are large, can have diffuse
Resectiona borders with unclear dissection planes, or can extend deep into
Cortical Extent of sensitive white matter. Mapping is useful with these larger AVMs
Patients Mapping Preservation Approachb Resectionc because an adjacent lobule of brain may need to be removed to
better visualize a dissection plane and, in some cases, resection
Awake
1 Language Yes Yes Yes
must be stopped to avoid surgical morbidity.
2 Language Yes No No Patient presentation is another important indication for
3 Language Yes No No mapping. Patients that present with hemorrhage and associated
4 Language Yes No No hemiplegia or aphasia were not mapped because many of these
5 Language, motor Yes Yes No patients were comatose and not suitable for intraoperative speech
Asleep testing. In addition, many of these patients’ functional cortex or
6 Motor Yes No Yes
associated white matter tracts may have been injured by the prior
7 Motor No No No
8 Motor Yes No Yes hemorrhage. In contrast, patients that present with unruptured
9 Motor Yes No No AVMs and seizures without deficits can cooperate with intra-
10 Motor Yes Yes No operative speech testing and appreciate the importance of co-
11 Motor Yes No No operation. The surgical risk, therefore, in these nonhemorrhagic
12 Motor Yes No Yes patients is considerable and mapping is more informative.
a The 12 patients who underwent language or motor mapping
AVM, arteriovenous malformation
b
‘‘Approach’’ refers to mapping’s utility in guiding the dissection through overlying were selected from a larger group of 36 patients with AVMs in or
sulci or brain to deep AVMs. around the language cortex and 28 patients with AVMs in or
c
‘‘Extent of resection’’ refers to mapping’s utility in deciding whether to completely around the motor cortex. The 52 patients who were not mapped
resect the AVM or abort resection after subtotal circumferential dissection around
the AVM.
had either Canestra group I lesions, or Canestra group III lesions
that were small, compact, and/or associated with significant
neurological deficits from hemorrhage. Excluded patients were
sufficiently different from included patients that a direct comparison
Canestra et al14 developed for peri-Sylvian AVMs: group I of outcomes would not demonstrate the benefits of mapping.
patients (minimal risk) have AVMs at least one gyrus removed
from language activation and are operated on asleep; group II Usefulness of Mapping With AVMs
(high-risk) patients have AVMs intimately associated with lan- We judged mapping to be useful in all 12 cases of AVM
guage function and are considered for Gamma knife radiosurgery resection. Even when mapping fails to identify the motor cortex
or observation; and group III patients (undetermined risk) have (as in one of our patients), it enables confident dissection around
AVMs adjacent to the language function and the risk of language the nidus. Mapping had 3 specific effects during AVM surgery.
deficit cannot be predicted with fMRI alone. The patients in our First, it identified the functional cortex, promoted careful
study were group III patients and were operated on awake with handling during the dissection, and thereby preserved function
cortical stimulation. The Canestra algorithm applies to language (11 patients). Second, mapping guided the dissection to AVMs
mapping, but we adopted a similar scheme for the motor cortex. beneath the cortical surface, identifying safe sulci away from
Patients with AVMs 1 gyrus removed from the motor cortex or functional cortex and overlying lobules whose removal might
with major reorganization of the motor function were operated improve access to the nidus (3 patients). Third, mapping im-
without motor mapping; patients with AVMs intimately asso- pacted the decision to resect the AVM completely, encouraging
ciated with motor function were considered for Gamma knife more conservative, subtotal circumdissections in 4 patients.
radiosurgery or observation; and patients with AVMs adjacent to Schaller et al1 refer to the precentral cortex, basal ganglia,
the motor function were operated on with motor mapping. and brainstem as areas of ‘‘high eloquence.’’ Knowing the exact
Therefore, we recommend preoperative functional imaging to location of the language and motor cortex, and the impact of
help select AVM patients for mapping. a deficit in intact patients, clearly makes the neurosurgeon more
We applied mapping techniques to larger AVMs with higher conservative. In this experience with 12 patients, the rate of
Spetzler-Martin grades. AVMs in the language and motor cortex subtotal resection was 25%; in the senior author’s overall AVM
are, by definition, eloquent, but many of them lack deep venous experience in the other 419 AVM patients, the rate of subtotal
drainage. Therefore, medium or large size is a common feature in resection is less than 2%. An informed attitude to these eloquent
these mapped AVMs. Low-grade AVMs tend to be more com- AVMs is appropriate, as reflected by a surgical morbidity rate of
pact, with clearer dissection planes and better separation from the 25% in a challenging group of patients. It is often difficult to halt
eloquent brain. The usefulness of mapping with these AVMs is the dissection short of complete resection, but the neurosurgeon
less because approaches to small AVMs are more direct, adjacent must allow the mapping to inform intraoperative decisions and
lobules of the brain do not need to be removed to encircle the occasionally stop operating. The small size of this experience and

NEUROSURGERY VOLUME 68 | NUMBER 3 | MARCH 2011 | 749

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.

GABARRÓS ET AL

FIGURE 3. Patient 10. Digital subtraction angiography [left internal carotid artery injection, lateral (A) and anteroposterior (B) views] demonstrated a Spetzler-Martin
grade II AVM adjacent to motor cortex. C, the AVM was located in the central sulcus, and motor mapping confirmed motor function immediately anterior to the nidus (10).
D, the superficial dissection was in subarachnoid spaces, but deep dissection was in parenchymal planes. The nidus was encircled (E) and resected completely (F). AVM,
arteriovenous malformation.

the absence of a control group of patients with AVMs operated on
without mapping make it impossible to conclude that brain
mapping with language and motor AVMs improves outcomes.
However, mapping does allow for safe surgical dissection that
resulted in more complete resections in these 12 patients than
might have otherwise been achieved.
There are, however, substantial risks to resecting an AVM in an
awake patient that should not be underestimated. Loss of airway
patency, significant hemorrhage, and hemodynamic instability
can present daunting challenges in a patient in rigid pin fixation
with an unsecured airway. Even mild hypoxemia or CO2
retention can impair brain relaxation and surgical exposure.
Securing the airway may interrupt surgery at a critical juncture.
Intraoperative anesthetic management should include vigilance
for sudden hemodynamic changes that might require mas-
sive fluid resuscitation and pharmacological manipulation of the
circulation. The anesthesiologist must be prepared with large bore
intravenous access, numerous intravenous sites for multifaceted
resuscitation, vasoactive agents, and warm fluids.
An Alternative Strategy for Language and Motor
Cortex AVMs
The AVMs in 4 patients treated with incomplete circum-
dissection (dearterializing the nidus while preserving the

750 | VOLUME 68 | NUMBER 3 | MARCH 2011 www.neurosurgery-online.com

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.

 BRAIN MAPPING OF AVMS

TABLE 3. Surgical Results and Patient Outcomes After AVM Surgery With Brain Mappinga
Seizures mRS
Additional F/U,
Patients AVM Resection Complications Postop 3 mo Treatment Preop Postop Last y Obliteration
Awake
1 Incomplete None No No Radiosurgery 1 2 3 1.0 Hemorrhage, embolization,
complete resection
2 Total None No No None 2 3 3 0.8
3 Total None Temporal Temporal None 1 1 1 6.0
4 Total ICH No No None 1 3 1 1.0
5 Total EDH No No None 1 3 0 2.0
Asleep
6 Incomplete None No No Radiosurgery 1 4 3 Within latency period
7 Total None No No None 3 4 1 11.0
8 Incomplete Intraop seizure Focal motor No Radiosurgery 1 4 1 2.0 Obliterated
9 Total Intraop seizure No No None 1 1 0 2.0
10 Total None No No None 0 4 0 2.0
11 Total None No No None 3 1 1 1.0
12 Incomplete None Generalized No Radiosurgery 1 3 1 8.0 Within latency period
a
AVM, arteriovenous malformation; mRS, modified Rankin Scale; EDH, epidural hematoma; ICH, intracerebral hematoma; Preop, preoperative; Postop, postoperative;
F/U, follow-up; Intraop, intraoperative.

draining vein and adjacent functional cortex) were left in situ
and treated with radiosurgery. We hypothesize that AVMs in
this reduced flow state might respond more favorably to ra-
diosurgery. The outcomes in this small group are mixed, with 1
hemorrhage during the latency period, 1 AVM obliteration, and
2 patients still in the latency period. These results confirm that
significant flow reduction does not protect against hemorrhage,
and treatment must continue until the AVM is obliterated
completely. The sequence of treatment modalities is reversed
from other high-risk AVMs. We typically treat these patients
with stereotactic radiosurgery first, either in a single stage or 2 to
3 volume stages. We then reevaluate AVM resectability and
therapeutic risks after the latency period or after intervening
hemorrhage. AVMs that are more favorable for surgery, either
because of radiosurgical reduction in nidus size or neurological
deterioration after hemorrhage, are treated with embolization
and resection.
In contrast to this standard strategy, an alternative strategy
would be conservative surgery first with intraoperative mapping
as a guide, followed by radiosurgery and possibly a second sur-
gery. Although one of our patients had an obliterative response to
radiosurgery after incomplete resection, additional patients are
needed to determine whether this strategy is effective, and
whether obliteration rates after radiosurgery are favorably influ-
enced by subtotal circumdissection.
Disclosure
The authors have no personal financial or institutional interest in any of the
drugs, materials, or devices described in this article.
REFERENCES
1. Schaller C, Schramm J, Haun D. Significance of factors contributing to surgical
complications and to late outcome after elective surgery of cerebral arteriovenous
malformations. J Neurol Neurosurg Psychiatry. 1998;65(4):547-554.
2. Morgan MK, Johnston IH, Hallinan JM, Weber NC. Complications of surgery
for arteriovenous malformations of the brain. J Neurosurg. 1993;78(2):176-182.
3. Friedman WA, Blatt DL, Bova FJ, Buatti JM, Mendenhall WM, Kubilis PS. The
risk of hemorrhage after radiosurgery for arteriovenous malformations. J Neurosurg.
1996;84(6):912-919.
4. Yamamoto M, Jimbo M, Hara M, Saito I, Mori K. Gamma knife radiosurgery
for arteriovenous malformations: long-term follow-up results focusing on com-
plications occurring more than 5 years after irradiation. Neurosurgery. 1996;38(5):
906-914.
5. Alkadhi H, Kollias SS, Crelier GR, Golay X, Hepp-Reymond MC, Valavanis A.
Plasticity of the human motor cortex in patients with arteriovenous malformations:
a functional MR imaging study. AJNR Am J Neuroradiol. 2000;21(8):1423-1433.
6. Lehericy S, Biondi A, Sourour N, et al. Arteriovenous brain malformations: is
functional MR imaging reliable for studying language reorganization in patients?
Initial observations. Radiology. 2002;223(3):672-682.
7. Thickbroom GW, Byrnes ML, Morris IT, Fallon MJ, Knuckey NW, Mastaglia
FL. Functional MRI near vascular anomalies: comparison of cavernoma and
arteriovenous malformation. J Clin Neurosci. 2004;11(8):845-848.
8. Ulmer JL, Hacein-Bey L, Mathews VP, et al. Lesion-induced pseudo-dominance
at functional magnetic resonance imaging: implications for preoperative assess-
ments. Neurosurgery. 2004;55(3):569-579.
9. Vates GE, Lawton MT, Wilson CB, et al. Magnetic source imaging demonstrates
altered cortical distribution of function in patients with arteriovenous malfor-
mations. Neurosurgery. 2002;51(3):614-623.
10. Lazar RM, Marshall RS, Pile-Spellman J, et al. Anterior translocation of language
in patients with left cerebral arteriovenous malformation. Neurology. 1997;49(3):
802-808.
11. Lazar RM, Marshall RS, Pile-Spellman J, et al. Interhemispheric transfer of
language in patients with left frontal cerebral arteriovenous malformation.
Neuropsychologia. 2000;38(10):1325-1332.
12. Lawton MT. Spetzler-Martin Grade III arteriovenous malformations: surgical
results and a modification of the grading scale. Neurosurgery. 2003;52(4):740-748.

NEUROSURGERY VOLUME 68 | NUMBER 3 | MARCH 2011 | 751

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.

GABARRÓS ET AL
13. Burchiel KJ, Clarke H, Ojemann GA, Dacey RG, Winn HR. Use of stim-
ulation mapping and corticography in the excision of arteriovenous malfor-
mations in sensorimotor and language-related neocortex. Neurosurgery.
1989;24(3):322-327.
14. Cannestra AF, Pouratian N, Forage J, Bookheimer SY, Martin NA, Toga AW.
Functional magnetic resonance imaging and optical imaging for dominant-
hemisphere perisylvian arteriovenous malformations. Neurosurgery. 2004;55(4):
804-812.
15. Duffau H, Lopes M, Arthuis F, et al. Contribution of intraoperative electrical
stimulations in surgery of low grade gliomas: a comparative study between two
series without (1985-96) and with (1996-2003) functional mapping in the same
institution. J Neurol Neurosurg Psychiatry. 2005;76(6):845-851.
16. Ebeling U, Schmid UD, Ying H, Reulen HJ. Safe surgery of lesions near the motor
cortex using intra-operative mapping techniques: a report on 50 patients. Acta
Neurochir (Wien). 1992;119(1-4):23-28.
752 | VOLUME 68 | NUMBER 3 | MARCH 2011
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
17. Gorecki JP, Smith RR, Wee AS. Excision of arteriovenous malformation in
sensorimotor and language-related neocortex using stimulation mapping and
corticography under local anesthesia. Stereotact Funct Neurosurg. 1992;58(1-4):89.
18. Kombos T, Suess O, Funk T, Kern BC, Brock M. Intra-operative mapping of the
motor cortex during surgery in and around the motor cortex. Acta Neurochir
(Wien). 2000;142(3):263-268.
19. Zamorano L, Matter A, Saenz A, Portillo G, Diaz F. Interactive image-guided
surgical resection of intracranial arteriovenous malformations. Comput Aided Surg.
1998;3(2):57-63.
20. Maldjian J, Atlas SW, Howard RS, et al. Functional magnetic resonance imaging
of regional brain activity in patients with intracerebral arteriovenous malformations
before surgical or endovascular therapy. J Neurosurg. 1996;84(3):477-483.
21. Shimamura N, Ohkuma H, Ogane K, et al. Displacement of central sulcus in
cerebral arteriovenous malformation situated in the peri-motor cortex as assessed by
magnetoencephalographic study. Acta Neurochir (Wien). 2004;146(4):363-368.
www.neurosurgery-online.com