usefulness of intraoperative electrical subcortical mapping during … · 2007. 5. 3. · resection...

ESPITE published opinions favoring surgery for low-grade gliomas,8,11,65,83,91,93,101,103,104,117 this treatmentstrategy remains controversial, particularly because

of the risk of inducing a permanent neurological deficit ifthe tumor is located near or within so-called eloquent brainareas.3,17,18,40,116,129,134 To decrease postoperative morbidity,functional mapping methods such as preoperative neuroim-aging1,4,19,41,44,46,55,58,61,62,67,75,77,87,89,96,110,119,125,128,136 (for example,PET scanning, fMR imaging, and magnetoencephalogra-phy) and intraoperative electrophysiological monitoring2,5,

15,18,25,26,37,38,54,60,68,69,72,79,82,85,86,94,102,111,112,114,121,124,127,131,135,140 havebeen extensively used in the past decade. Most authors havedescribed only their results with cortical mapping, usedboth at the time of preoperative planning (because neuro-functional imaging is incapable of mapping the white mat-ter tracts),1,4,19,41,44,46,55,58,61,62,67,75,77,87,89,96,110,119,125,128,136 and dur-

ing surgery by using preoperative functional data integrat-ed into a neuronavigational system,16,19,47,48,71,80,84,92,95,108,113,122

evoked potential recordings,2,5,18,45,47,48,54,80,82,86,87,111,114,135 and/orintraoperative direct electrical cortical stimulations.15,25,26,37,

38,41,46,55,58,60,68,69,71,72,79,80,84,85,89,94,102,111,112,121,124,125,127,131,140 As a rule,low-grade gliomas invade cortical as well as subcorticalstructures, and thus definitive deficits may occur becauseof surgical damage to pathways running in the white mat-ter.18,20,73,102

Currently, we present a consecutive series of 103 patientswho, during an operation for a low-grade glioma locatedin eloquent corticosubcortical areas, underwent real-timemapping of both sensorimotor and language cortical sitesand white fibers. Throughout these surgeries we performeddirect electrical stimulations as described by Berger andcolleagues.9,10 Clearly, the aim of our work was not to studythe impact of surgery on the natural history of a low-gradeglioma, but rather to detail both immediate and delayedpostoperative functional results after performing a maximalresection according to functional boundaries determinedwith the aid of intraoperative cortical and subcortical map-

J Neurosurg 98:764–778, 2003

764

Usefulness of intraoperative electrical subcortical mappingduring surgery for low-grade gliomas located withineloquent brain regions: functional results in a consecutiveseries of 103 patients

HUGUES DUFFAU, M.D., PH.D., LAURENT CAPELLE, M.D., DOMINIQUE DENVIL, M.D.,NICOLE SICHEZ, PEGGY GATIGNOL, S.T., LUC TAILLANDIER, M.D., MANUEL LOPES, M.D.,MARY-CHRISTINE MITCHELL, M.D., SABINE ROCHE, M.D., JEAN-CHARLES MULLER, M.D.,AHMAD BITAR, M.D., JEAN-PIERRE SICHEZ, M.D., AND RÉMY VAN EFFENTERRE, M.D.

Departments of Neurosurgery, Neurology, and Neuroanesthesiology, Hôpital de la Salpêtrière, Paris;and Department of Neurology, Centre Hospitalier Universitaire de Nancy, France

Object. Although a growing number of authors currently advocate surgery to treat low-grade gliomas, controversy stillpersists, especially because of the risk of inducing neurological sequelae when the tumor is located within eloquent brainareas. Many researchers performing preoperative neurofunctional imaging and intraoperative electrophysiological meth-ods have recently reported on the usefulness of cortical functional mapping. Despite the frequent involvement of subcor-tical structures by these gliomas, very few investigators have specifically raised the subject of fiber tracking. The authorsin this report describe the importance of mapping cortical and subcortical functional regions by using intraoperative real-time direct electrical stimulations during resection of low-grade gliomas.

Methods. Between 1996 and 2001, 103 patients harboring a corticosubcortical low-grade glioma in an eloquent area,with no or only mild deficit, had undergone surgery during which intraoperative electrical mapping of functional corticalsites and subcortical pathways was performed throughout the procedure.

Both eloquent cortical areas and corresponding white fibers were systematically detected and preserved, thus definingthe resection boundaries. Despite an 80% rate of immediate postoperative neurological worsening, 94% of patients recov-ered their preoperative status within 3 months—10% even improved—and then returned to a normal socioprofessional life.Eighty percent of resections were classified as total or subtotal based on control magnetic resonance images.

Conclusions. The use of functional mapping of the white matter together with cortical mapping allowed the authors tooptimize the benefit/risk ratio of surgery of low-grade glioma invading eloquent regions. Given that preoperative fibertracking with the aid of neuroimaging is not yet validated, we used intraoperative real-time cortical and subcortical stim-ulations as a valuable adjunct to the other mapping methods.

KEY WORDS • subcortical mapping • low-grade glioma • surgery •intraoperative stimulation • language • sensorimotor cortex

D

J. Neurosurg. / Volume 98 / April, 2003

Abbreviations used in this paper: DW = diffusion-weighted;fMR = functional magnetic resonance; KPS = Karnofsky Perfor-mance Scale; PET = positron emission tomography; SSEP = soma-tosensory evoked potential.

pings. After collecting these data, we reviewed the literatureon tumor surgery in eloquent areas, with particular focus ontechnical and functional considerations.

Clinical Material and Methods

Between October 1996 and October 2001, 103 patientsunderwent surgery for a corticosubcortical low-grade gli-oma located in eloquent brain areas with the aid of intraop-erative electrical mapping. Note that part of this experiencehas been previously described, in particular, the anatomico-functional organization of the language pathways but notthe detailed data regarding neurological outcome.34

The presenting symptoms and results of the preoperativeneurological examination were assessed by both the neu-rologists (D.D. and L.T.) and neurosurgeons (H.D., L.C.,M.L., A.B., J.P.S., and R.V.E.). The intensity of the motordeficit was rated using a standardized motor scale22 as fol-lows: 0, no deficit; 1, mild deficit (patient can use his or herlimb almost normally, that is, walking is possible, but thereis impairment of fine movements of the upper limb); 2,moderate deficit (movement is possible with help from theexaminer); and 3, severe deficit (no spontaneous movementagainst gravity). Speech functions were evaluated clinical-ly by neurologists, neurosurgeons, and/or a speech thera-pist (P.G.), who tested verbal comprehension, spontaneousspeech, naming, verbal fluency, narrative tasks, and repeti-tion. Moreover, a neuropsychological task performed bythe patient was assessed in the second half of the series (thatis, 1998) by a neuropsychologist (N.S.). Finally, we evalu-ated the KPS score for each patient.64

Hemispheric dominance was defined using a standard-ized questionnaire53 and fMR images obtained while the pa-tient performed semantic-fluency, covert sentence-repeti-tion, and story-listening tasks.76 Data were analyzed withthe aid of pixel-by-pixel autocorrelation and cross-correla-tion. The fMR imaging laterality indices were defined forseveral regions of interest as the ratio (L � R)/(L � R),where L is the number of activated voxels in the left hemi-sphere and R is that in the right hemisphere.76

The topography of the tumor was accurately demonstrat-ed on preoperative MR images (T1-weighted and spoiled-gradient images obtained before and after Gd enhancementin the three orthogonal planes, T2-weighted axial images,and fluid-attenuated inversion recovery axial images in thelast 3 years of the study).

During all surgical procedures, intraoperative real-timefunctional corticosubcortical mapping was performed—motor mapping in 39 patients after induction of general an-esthesia, and sensorimotor and language mappings in 64patients after administration of a local anesthetic agent—us-ing the technique of direct electrical stimulations, previous-ly detailed by us,28,31,33–35 and based on the methodology ofBerger and colleagues.9,10,97 The aim of this study was totrack and preserve eloquent structures during each momentand at each site of resection. Language tasks included theperformance of systematic counting, naming, and reading.Moreover, a calculation task was added if a patient harboreda lesion in the left angular and supramarginalis gyri, andrepetition and/or semantic tasks were added if a patient hada tumor within the left mid-posterior temporal lobe. Briefly,bipolar electrode tips spaced 5 mm apart and delivering a

biphasic current (pulse frequency 60 Hz, single-pulse phaseduration 1 msec, and amplitude 6–18 mA [general anesthe-sia] or 2–6 mA [local anesthesia]) (Ojemann cortical stim-ulator 1; Radionics, Inc., Burlington, MA) were applied tothe patient’s brain. First, before tumor removal, electricalmapping was completed at the cortical level to identify theessential eloquent sites that needed to be avoided and thusto define the superficial boundaries of resection accordingto functional data.15,25,26,37,38,41,46,55,58,60,68,69,71,72,79,80,84,85,89,94,102,111,112,

121,124,125,127,131,140 To define the intensity of the current, weprogressively increased the amplitude until a clinical mo-tor response was obtained (after general or local anesthesiahad been induced). We then used this intensity to performsomatosensory and language mappings after a local anes-thetic agent had been administered, without monitoring af-ter-discharge activity.

Second, direct stimulations, with the same electrical pa-rameters as those used at the cortical level, were continu-ously applied during glioma removal at the subcortical levelto detect sensorimotor and/or language pathways, whichrepresented the deep functional limits of resection.

With the patient in a state of general anesthesia, subcor-tical stimulations were applied continuously to avoid inter-ruption of the pyramidal fibers between the two brain stim-ulation trials. Despite repetitive stimulations, there was noconsideration of the eloquent structures until motor path-ways were encountered, that is, until a motor response wasinduced. Tumor removal was stopped once movement waselicited, and the same procedure was performed again inneighboring regions by closely following the corticospinaltracts with subcortical stimulations.

After having administered a local anesthetic and through-out the entire glioma removal procedure, the patient wasasked to perform tasks continuously, including motor open-ing and closing of the hand with the superior limb raised(with or without regular movements of the foot) as well aslanguage tasks. Most of the time these tests were controlledby the speech therapist (P.G.) or neuropsychologist (N.S.)or, more rarely, by the anesthesiology team, the membersof which had been trained in language testing (M.C.M.,S.R., and J.C.M.) and had received standardized instruc-tions detailed by the speech therapist. As soon as the slight-est limb weakness and/or dysphasia was noted, resectionwas stopped and subcortical stimulation was applied tocheck for any involuntary motor response and/or a speechdisturbance. If either of these occurred, glioma removal wasinterrupted at this site, and the same procedure was againperformed in the neighboring structures. If neither occurred,the patient was asked to rest, and when the tasks were nor-malized several minutes later, resection was resumed un-der the same conditions. In the case of parietal low-gradegliomas, subcortical stimulations were performed so thatthe awake patient could signal the induction of paresthesia,which would indicate that the resection had to cease be-cause thalamocortical pathways had been encountered.

At the end of the resection procedure, the integrity ofthe functional networks was systematically verified. Withthe patient in a state of general anesthesia, we stimulated themotor cortical sites. If the movements elicited were thesame as those evoked before tumor removal, the motorpathways were considered to be intact, thus meaning thatthe patient would recover despite the possibility of incur-ring an immediate postoperative deficit.29,33 Again after in-


Subcortical mapping in glioma surgery

765

H. Duffau, et al.

766 J. Neurosurg. / Volume 98 / April, 2003

TAB

LE

1C

lini

cal,

radi

olog

ical

, and

sur

gica

l ch

arac

teri

stic

s of

103

pat

ient

s w

ith

a co

rtic

osub

cort

ical

low

-gra

de g

liom

a in

an

eloq

uent

are

a*

Preo

pPr

eop

Pres

entin

gE

xam

inat

ion

KPS

Imm

edia

te P

osto

pD

elay

ed P

osto

pPo

stop

Ext

ent

Glio

ma

No.

of

Cas

esSy

mpt

oms

Res

ults

Scor

eSu

bcor

tical

Ana

tom

icof

unct

iona

l Bou

ndar

ies

Clin

ical

Res

ults

Clin

ical

Res

ults

KPS

of

Loc

atio

n(h

emis

pher

e)(n

o. o

f pa

tient

s)(n

o. o

f pa

tient

s)(n

o. o

f pa

tient

s)Id

entif

ied

on I

ntra

op S

timul

atio

ns(n

o. o

f pa

tient

s)(n

o. o

f pa

tient

s)Sc

ore

Res

ectio

n

fron

to-

40se

izur

es (

39no

rmal

(36

), m

o-10

0 (2

8)po

ster

iorl

y, p

yram

idal

pat

hway

s (a

ll ca

ses)

;no

rmal

(5)

,no

rmal

(38

100

(17)

13 T

,pr

ecen

tral

[8 P

RE

]),

tor

defi

cit (

2),

90 (

3)do

min

ant h

emis

pher

e: m

edia

lly, f

asci

culu

ssl

ight

def

icit

(12)

,[3

PR

E])

, mo-

90 (

18)

19 S

T,re

gion

ICH

(1)

spee

ch d

efic

it (2

)80

(9)

subc

allo

sal m

edia

lis w

/ or

w/o

hea

d of

cau

-m

oder

ate

defi

cit (

12),

tor

defi

cit (

1),

80 (

3)8

PF1

28 (

8 rt

, 20

lt)da

te n

ucle

us; m

ore

late

rally

, lan

guag

e pa

th-

seve

re d

efic

it (1

1)sp

eech

def

icit

(1)

70 (

2)F2

5 (1

rt,

4 lt)

way

s fr

om p

rem

otor

cor

tex;

aga

in m

ore

late

r-F3

7 (4

rt,

3 lt)

ally

, lan

guag

e pa

thw

ays

from

the

Bro

ca a

rea

para

limbi

c29

(20

rt,

seiz

ures

(27

norm

al (

27),

100

(11)

deep

ly, p

yram

idal

pat

hway

s w

ithin

inte

rnal

norm

al (

5),

norm

al (

2610

0 (6

)6

T,re

gion

9 lt)

[16

PRE

]),

spee

ch d

efic

it (2

)90

(5)

caps

ule;

dom

inan

t hem

isph

ere:

dee

ply,

arc

uate

slig

ht d

efic

it (1

3),

[4 P

RE

]), m

o-90

(16

)17

ST,

ICH

(2)

80 (

13)

fasc

icul

us w

ithin

ext

erna

l cap

sule

w/ o

rm

oder

ate

defi

cit (

4),

tor

defi

cit (

3)80

(4)

6 P

w/o

lat p

art o

f pu

tam

ense

vere

def

icit

(7)

70 (

3)te

mpo

ral

10 (

2 rt

,se

izur

es (

9no

rmal

(9)

,10

0 (8

)po

ster

iorl

y, la

ngua

ge p

athw

ays

corr

espo

ndin

g to

norm

al (

4),

norm

al (

1010

0 (7

)4

T,re

gion

8 lt)

[1 P

RE

]),

spee

ch d

efic

it (1

)80

(2)

ante

roin

feri

or p

art o

f ar

cuat

e fa

scic

ulus

slig

ht d

efic

it (4

),[1

PR

E])

90 (

2)5

ST,

(dom

inan

tIC

H (

1)(a

nter

ior-

mid

tem

pora

l glio

ma)

; ant

erio

rly,

mod

erat

e de

fici

t (1)

,80

(1)

1 P

hem

isph

ere)

lang

uage

pat

hway

s co

rres

pond

ing

to p

oste

ro-

seve

re d

efic

it (1

)in

feri

or p

art o

f ar

cuat

e fa

scic

ulus

(po

ster

o-te

mpo

ral g

liom

a)pa

riet

o-9

(9 lt

)se

izur

es (

9)no

rmal

(9)

100

(9)

deep

ly, p

oste

rosu

peri

or lo

op o

f ar

cuat

eno

rmal

(3)

,no

rmal

(9)

100

(7)

4 T,

occi

pito

-fa

scic

ulus

; mor

e su

perf

icia

lly, l

ocor

egio

nal

slig

ht d

efic

it (5

),90

(2)

4 ST

,te

mpo

ral

lang

uage

con

nect

ions

mod

erat

e de

fici

t (1)

1 P

junc

tion

(dom

inan

the

mis

pher

e)ro

land

ic8

(6 r

t, 2

lt)se

izur

es (

8no

rmal

(7)

, mo-

100

(5)

ante

rior

ly, p

yram

idal

pat

hway

s; p

oste

rior

ly,

norm

al (

3),

norm

al (

7), m

o-10

0 (4

)2

T,re

gion

[2 P

RE

])to

r de

fici

t (1)

80 (

3)so

mat

osen

sory

thal

amoc

ortic

al p

athw

ays;

slig

ht d

efic

it (3

),to

r de

fici

t (1)

90 (

3)4

ST,

dom

inan

t hem

isph

ere:

late

rally

, lan

guag

em

oder

ate

defi

cit (

1),

70 (

1)2

Ppa

thw

ays

com

ing

from

ven

tral

pre

mot

or &

seve

re d

efic

it (1

)fa

cial

pri

mar

y m

otor

cor

tices

pari

eto-

7 (4

rt,

3 lt)

seiz

ures

(6)

,no

rmal

(5)

, hyp

-10

0 (5

)an

teri

orly

, som

atos

enso

ry th

alam

ocor

tical

pat

h-no

rmal

(2)

,no

rmal

(7)

100

(4)

2 T,

retr

o-IC

H (

1)es

thes

ia (

2)80

(2)

way

s (a

ll ca

ses)

; dom

inan

t hem

isph

ere:

ant

ero-

slig

ht d

efic

it (4

),90

(3)

2 ST

,ce

ntra

lla

tera

lly (

deep

ly),

lang

uage

pat

hway

s is

sued

mod

erat

e de

fici

t (1)

3 P

regi

onfr

om s

upra

mar

gina

lis g

yrus

& jo

ined

the

post

eros

uper

ior

part

of

arcu

ate

fasc

icul

us

*F1

= s

uper

ior

fron

tal g

yrus

; F2

= m

iddl

e fr

onta

l gyr

us; F

3 =

infe

rior

fro

ntal

gyr

us; I

CH

= in

trac

rani

al h

yper

tens

ion;

P=

par

tial;

PRE

= p

harm

acol

ogic

ally

res

ista

nt e

pile

psy;

ST

= s

ubto

tal;

T=

tota

l.

duction of local anesthesia, we asked the patient to repeatall functional tasks. The absence of a deficit meant thatthe patient would recover despite an eventual immediatepostoperative sensorimotor and/or language disorder.36,43 In-traoperative glioma delineation was systematically deter-mined using ultrasonography or neuronavigation.

All patients were examined in the immediate postopera-tive period, at 3 months, and then every 6 months thereafter,by the same team of neurosurgeon, neurologist, and neuro-psychologist, who had assessed the same tasks performedpreoperatively. A control MR image was obtained in allcases immediately, 3 months, and then every 6 months post-surgery, to evaluate the quality of glioma removal accord-ing to the classification method reported on by Berger andcolleagues8 (that is, total resection indicated no residual sig-nal abnormality, subtotal resection indicated � 10 cm3 ofresidue, and partial resection indicated � 10 cm3 of resi-due). All MR images were systematically interpreted by atleast two observers.

Patient Population

This series included 48 men and 55 women, ranging inage from 17 to 63 years (mean 36 years). All but four pa-tients were right handed. Presenting symptoms includedseizures in 95% of cases and intracranial hypertension in5%. Pharmacologically resistant epilepsia was present in 27patients. Results of an initial detailed neurological exami-nation were normal in all patients except 10 who presentedwith a mild motor deficit (rated 1 on the standardized mo-tor scale; three patients), a slight hypesthesia (two patients),and a reduction in verbal fluency (five patients). Twenty-nine patients had a KPS score of 80, eight patients a scoreof 90 (able to work despite a mild deficit and/or chronic sei-zures), and 66 a score of 100. In patients who underwenta preoperative neuropsychological evaluation, although re-sults of the Mini-Mental State Examination were higherthan 28, all displayed slight working memory disorders.

Preoperative MR Imaging

According to results of MR imaging, lesion locationswere distributed as follows. Forty lesions were situated inthe frontoprecentral region: 28 within the superior frontalgyrus, with involvement of the supplementary motor area(eight on the right side and 20 on the left); five in the mid-dle frontal gyrus (one on the right side and four on the left);and seven in the inferior frontal gyrus (four on the right andthree on the left). Twenty-nine lesions were located in theparalimbic region, with involvement of the insular lobe, andwith or without the frontobasal and/or temporopolar regions(20 on the right and nine on the left). Ten gliomas were sit-uated in the temporal area, with involvement of the domi-nant lobe (two on the right and eight on the left). Nine gli-omas involved the left dominant parietooccipitotemporaljunction. There were eight lesions in the rolandic region (sixon the right and two on the left), and seven in the parieto-retrocentral area (four on the right and three on the left).

Results

The clinical, radiological, and surgical data of the 103 pa-tients are summarized in Table 1.

Surgical Findings

The initial cortical mapping allowed for the detection ofeloquent areas, which were then used as boundaries of re-section. The electrical subcortical mapping equally allowedfor the direct identification and preservation of the function-al fibers.

Low-Grade Gliomas in the Frontoprecentral Region. Theposterior limit of resection for lesions in the frontopre-central area was defined by the identification of the corti-cospinal pathways in all patients, for example, from medialto lateral directions, the motor tracts of the lower limb, up-per limb, and face. Thus, the entire corona radiata could beidentified, from the cortical surface to its junction with theinternal capsule, at the level of the rotation of the fibers(Fig. 1).

Moreover, in the dominant hemisphere, resection bound-aries were defined by language pathways (Fig. 2): medial-ly, the fasciculus subcallosal medialis, from the supplemen-tary motor area and cingulum to the head of the caudatenucleus, and from which were elicited transient symptomsof transcortical motor aphasia on stimulation (it was used asthe medial limit of resection in cases of lesions invading themiddle [with or without the inferior] frontal gyrus [four pa-tients]); more laterally, the language fibers coming from thepremotor cortex, and from which anomia and/or anarthriawas induced on stimulation (it was used as the lateral limitof resection for lesions invading the superior frontal gyrus[20 patients] and as the medial boundary for tumors involv-ing the inferior frontal gyrus [three patients]); again morelaterally, the language fibers coming from the Broca areaand connecting this region to the insula and primary motorcortex, from which was elicited speech arrest on stimulation(it was used as the lateral boundary of resection for low-grade gliomas involving the middle [with or without the su-perior] frontal gyrus [four patients]).

Low-Grade Gliomas in the Paralimbic Region. In the non-dominant hemisphere in 20 patients, the depth of resectionwas limited by the identification of the corticospinal path-ways in the posterior limb of the internal capsule (withor without its junction with the corona radiata for lesions inthe frontoinsular region, and its junction with the cerebralpeduncles for tumors in the temporoinsular region). No mo-tor response was induced by stimulation of the basal gan-glia, thus allowing for resection (at least partially) of in-vading gliomas in 16 patients.

In the dominant hemisphere, after resection of at leastpart of the glioma in the insular area, the deep boundary wasdetermined by the external capsule in which runs the arcu-ate fasciculus (eliciting paraphasia on stimulation). In onecase the depth of resection was limited by the lateral part ofthe putamen more anteriorly (inducing anarthria when stim-ulated; Fig 3). Gliomas located in the left basal ganglia werenever resected.

Low-Grade Gliomas in the Temporal Region. For anteriortemporal low-grade gliomas in four patients, the posteriorlimit of resection limited by the identification of the lan-guage pathways running from the cortical language sites tothe anterior brachium of the temporal part of the arcuatefasciculus (eliciting anomia or paraphasia when stimulat-ed). For low-grade gliomas of the midtemporal region (fourpatients), the anterior limit of resection was represented by



767

the language pathways constituting the anterior brachium ofthe arcuate fasciculus, and the posterior limit was indicatedby the language pathways constituting the posterior brachi-um of the arcuate fasciculus (Fig. 4). Moreover, above theventricle, the posterosuperior boundary of resection may berepresented by the pyramidal pathways within the internalcapsule. In two patients with posterior temporal low-gradegliomas, the depth of resection was limited by the language

pathways converging from the cortical sites to the posteriorbrachium of the temporal part of the arcuate fasciculus.

Low-Grade Gliomas of the Left ParietooccipitotemporalRegion. The depth of resection for gliomas of the left pari-etooccipitotemporal region was limited by the identificationof language pathways represented by the posterosuperiorloop of the arcuate fasciculus and, more superficially, by lo-coregional connections between cortical language sites.

H. Duffau, et al.


FIG. 1. Preoperative axial (upper left and center) and coronal (upper right) T1-weighted enhanced MR images demon-strating a right frontoprecentral low-grade glioma in a patient with no neurological deficit. Center Left: Intraoperativephotograph obtained before resection. The boundaries of the glioma were delineated using ultrasonography and markedby letters (A, B, C, and D). The cortical eloquent areas were detected with the aid of direct electrical stimulations as fol-lows: 10 to 15, primary somatosensory areas, located within the retrocentral gyrus; 1 to 6, primary motor areas of the supe-rior limb, located within the precentral gyrus; and 20, 21, 22, 24, and 25, primary motor areas of the face, located more lat-erally in the precentral gyrus. Center Right: Intraoperative photograph obtained after resection. The sensorimotor corticalsites were surgically preserved (for technical reasons, tag 21 was removed and tag 3 was moved medially). Furthermore,subcortical motor pathways were identified by performing intraoperative stimulations and represented the limits of the re-section depth. The convergence of the pyramidal fibers at the level of the corona radiata was marked by the following tags(from medial to lateral): 31, motor fibers of the inferior limb (coming from the paracentral gyrus, located more medially);35, motor fibers of the superior limb; and 32, motor fibers of the face. Postoperative axial (lower left and center) and coro-nal (lower right) T1-weighted enhanced MR images demonstrating that the boundaries of resection were represented firstby the primary motor cortex at the cortical level (lower left), second by the pyramidal pathways at the subcortical level(lower center, curved arrow pointing to the right corresponds to tag 32, straight arrow corresponds to tag 35, and curvedarrow pointing to the left corresponds to tag 31). Because the glioma had infiltrated functional structures, the resection wassubtotal (2 cm3), as demonstrated on MR images. Despite a transient postoperative motor deficit, the patient returned tonormal socioprofessional life (civil servant) 3 months postsurgery.

Low-Grade Gliomas of the Rolandic Area. After detectingprimary motor and/or somatosensory cortical sites, the sen-sorimotor pathways were followed closely, indicating thefunctional limits of glioma resection in the rolandic region.Moreover, in the dominant hemisphere in two patients, lan-guage pathways coming from the primary motor area of theface and running to the frontal horn of the lateral ventriclebehind the subcallosal fasciculus were identified by the in-duction of anarthria on stimulation.

Low-Grade Gliomas of the Parietoretrocentral Region. Theanterior limit of resection for gliomas of the parietoretro-central region was defined by the thalamocortical path-ways, with identification of their somatotopical organiza-tion from the lateral horn of the ventricle to correspondingsomatosensory cortical sites (for example, from medial tolateral directions, the sensory tracts of the inferior limb, su-perior limb, and face). In the dominant hemisphere in threepatients, the depth of the anterolateral limit was represented



769

FIG. 2. Preoperative axial (upper left and center) and sagittal (upper right) T1-weighted enhanced MR image showinga left frontoprecentral low-grade glioma invading the middle gyrus in a patient with no neurological deficit. Center Left:Intraoperative photograph obtained before resection. The boundaries of the glioma, delineated using ultrasonography, weremarked by letters (A, B, C, D, and E). Cortical eloquent areas were detected by direct electrical stimulations as follows:12 and 15, primary somatosensory areas, located within the retrocentral gyrus; 1 to 4, primary motor areas of the superiorlimb, located within the precentral gyrus; 10, 11, 13, 14, and 17, primary motor areas of the face and ventral premotor cor-tex, located more laterally in the precentral gyrus (inducing anarthria on stimulation); 20 and 21, the Broca area, elicitingspeech arrest when stimulated. Center Right: Intraoperative photograph obtained after resection. The sensorimotor andlanguage cortical sites were surgically preserved. Moreover, subcortical functional pathways were identified by intraoper-ative stimulations and represented the deep limits of resection. In addition to the pyramidal fibers of the limbs, languagepathways were also detected and marked by tags as follows: 28, fasciculus subcallosal medialis, coming from mesial struc-tures and inducing symptoms of transcortical motor aphasia during stimulations; 26, motor fibers of the face, coming fromthe ventral premotor cortex and eliciting anarthria when stimulated; 27, lateral language fibers, coming from the Broca areaand inducing speech arrest during stimulation. Postoperative axial (lower left and center) and sagittal (lower right) T1-weighted enhanced MR image showing that the boundaries of resection were represented first by the primary motor cor-tex at the cortical level (lower left), second by the functional pathways at the subcortical level (lower center, curved arrowcorresponds to tag 28, straight arrow corresponds to tag 26; lower right, arrows correspond to tag 27, connecting Brocaarea/insula and Broca area/motor cortex). Because of an infiltration of the deep functional structures by the glioma, the re-section was subtotal (1.5 cm3), as demonstrated on MR images. Despite a transient postoperative language deficit, the pa-tient returned to normal socioprofessional life after 2 months (secretary).

by deep language pathways, which on stimulation most of-ten elicited time anomia and/or paraphasia.

In summary, the surgical strategy in all cases was mod-ulated according to subcortical mapping results, with themost extensive resection possible performed in 103 pa-tients. Resection was never interrupted before the function-al fibers were encountered, but was stopped as soon as thesepathways were identified using subcortical stimulations.

During operations performed in patients in a state of gen-eral anesthesia, repetition of the cortical stimulations usedto check the integrity of pyramidal pathways elicited thesame motor response at the end of resection compared withresponses elicited before resection in all patients exceptfive; no movement was induced in three patients with a

glioma in the right frontotemporoinsular region and a ma-jor decrease in the motor response occurred in two patientswith a right frontoprecentral lesion despite the use of thesame electrical parameters (both had presented with a pre-operative motor deficit). For surgeries performed after theadministration of a local anesthetic, all patients except oneremained able to perform motor and language tasks follow-ing resection; speech became impossible in a patient witha left precentral low-grade glioma, who had presented withpreoperative language disturbances.

The median length of surgery was approximately 5hours. Patients were awake, but the effects of the local an-esthesia remained for a median 2 hours. With the use ofa local anesthetic, the patients tolerated the procedure well.Despite their fatigue at the end of the resection procedure,many patients commented that they were relieved to help inthe monitoring of their own neurological condition duringglioma removal, and that if another operation was neces-sary in the future they would agree to undergo surgery againwhile awake.

Clinical Results

There was no operative or postoperative mortality.

Immediate Postoperative Functional Results

Eighty-one patients experienced neurological worseningimmediately following the surgical procedure. Forty-one ofthese patients had a mild motor deficit (rated 1 on the stan-dardized motor scale) and/or slight language disturbances(essentially a mild reduction in spontaneous speech). Themedian length of their hospital stay was approximately 5 to7 days, and each patient underwent rehabilitation therapy athome. Twenty patients had a moderate motor deficit (rated2) and/or moderate language disturbances (that is, a reduc-tion in spontaneous speech, dysarthria, dysnomia, and im-pairment in verbal fluency, with no verbal comprehensiondisorder). The median duration of hospitalization in thissubgroup was between 10 and 14 days, and these patientsrequired a more intensive and prolonged rehabilitation ther-apy (3 months) at home. Twenty patients had a severe de-ficit (rated 3) and/or complete aphasia (most often due toa supplementary motor area syndrome or the aftermath ofa frontotemporoinsular glioma resection). A transfer to arehabilitation center for at least 1 month was necessary inthese cases.

Delayed Postoperative Functional Results

On examination at 3 months after surgery, eight patientshad a mild deficit (motor scale rating of 1 and/or slight re-tardation of spontaneous speech) and 89 patients had nor-mal neurological function. At this time, only a few patientshad returned to a normal professional life (� 20%). A to-tal of six patients maintained a severe deficit (motor scalerating of 3 in five cases and dysphasia in one case). At 9months postsurgery, 97 patients had returned to a normalsocioprofessional life. The KPS score was 80 in eight pa-tients, 90 in 44, and 100 in 45. During the neuropsycholog-ical examination, results on a Mini-Mental State Exami-nation were greater than 28 in all cases and preoperativeworking memory disorders slightly improved in 15% ofcases (as shown by a higher score in the reversal order task).

H. Duffau, et al.


FIG. 3. Upper Left: Preoperative axial T1-weighted enhancedMR image showing a left insular low-grade glioma in a patient withspeech disturbances. Upper Right: Intraoperative photograph ob-tained before resection. The boundaries of the glioma, delineatedusing ultrasonography and neuronavigation, were marked by letters(A, B, C, and D). Note that the surface of the tumor was visible afteropening the sylvian fissure (star). Cortical eloquent areas were de-tected on direct electrical stimulations as follows: 20, 21, and 23,Broca area, inducing speech arrest when stimulated; 22, namingsite, located within the superior temporal gyrus. There was no re-sponse during stimulations of the insular cortex. Lower Left:Intraoperative photograph obtained after resection. The languagecortical sites were surgically preserved. Furthermore, subcorticallanguage pathways were identified by intraoperative stimulationsthat elicited paraphasia at the level of the external capsule (tag 24),which represented the deep limits of resection. Pyramidal pathwayswere also detected more posteriorly in the depth. Lower Right:Postoperative axial T1-weighted enhanced MR image demonstrat-ing that the deep boundaries of resection were represented first bythe language structures anteriorly (that is, the external capsule andlentiform nucleus), second by the pyramidal pathways posteriorly(running in the posterior limb of the internal capsule). Glioma re-moval was total. The patient’s neurological status improved imme-diately after surgery, and he returned to a normal socioprofessionallife 2 months later (computer scientist).

The six patients with sequelae did become independent, butwere unable to work normally, having a KPS score of 70.

In summary, in comparison with their preoperative status(Table 2), 10 patients had improved neurological function-ing (from deficit and/or intracranial hypertension), and 19of 27 patients who had presented with preoperative chronicepilepsia had improved functioning (11 patients had EngelGrade I). Ninety-seven of 103 patients returned to a normallife. The neurological functioning in six patients was per-manently worsened—three with a right frontotemporoinsu-lar low-grade glioma due to a deep ischemia involving theinternal capsule (lenticulostriate artery damage, despite thepreservation of the pyramidal pathways) and another threewith a widely infiltrative frontocentral low-grade glioma,who had presented with a preoperative deficit.

Note that there was no difference in surgical methodolo-gy among the subgroups of patients with or without post-operative worsening of neurological functioning; that is, allresections were stopped based on cortical and subcorticalmapping findings.

Histological Results

Results of the histopathological examination revealed a

low-grade glioma (World Health Organization Grade II) inall cases.

Radiological Results

Thirty-one resections were defined as total (with no re-sidual signal abnormality on MR imaging), 51 as subtotal



771

FIG. 4. Preoperative axial (upper left), coronal (upper center), and sagittal (upper right) T1-weighted enhanced MR im-ages demonstrating a left midtemporal low-grade glioma in a patient without any neurological deficit. Center Left: In-traoperative photograph obtained before resection. The boundaries of the glioma, again delineated using ultrasonography,were marked by letters (A, B, and C). The cortical eloquent areas were detected by direct electrical stimulations as follows:10, Broca area, eliciting speech arrest when stimulated; and 20 to 23, naming sites, located within the temporal pole (21)and the superior temporal gyrus (20, 23, 22). Center Right: Intraoperative photograph obtained after resection. The cor-tical eloquent sites were surgically preserved. Furthermore, subcortical language pathways were identified by intraopera-tive stimulations and represented the deep limits of resection: anterior brachium of the arcuate fasciculus anteriorly (tag43), and posterior brachium of the arcuate fasciculus posteriorly (tags 40 and 41). Postoperative axial (lower left), coronal(lower center), and sagittal (lower right) T1-weighted enhanced MR image demonstrating a total resection, with retractionof the temporal horn of the ventricle. The patient had no postoperative deficit and returned to a normal socioprofessionallife after 2 months (surgeon).

TABLE 2Immediate and delayed postoperative neurological outcome

in comparison with the preoperative status

Delayed Postop OutcomePreop (�3 months)

Neurological Immediate Compared w/Abnormalities Postop Outcome Preop Status

(no. of patients) (no. of patients) (no. of patients)

ICH (3)deficit (8) slight deficit (41) worsened preexisting

moderate deficit (20) deficit (3)severe deficit (20) new severe deficit (3)

deficit & deficit improved (7),ICH (2) ICH improved (3)

(� 10 cm3 of signal abnormality [there was � 5 cm3 in 23cases]), and 21 as partial. There was no difference in sur-gical methodology among the subgroups of patients whohad undergone total or incomplete removal, again becauseall resections were limited based on cortical and subcorticalmapping findings.

Discussion

Although recent literature demonstrates a positive impactof surgery on the natural history of a low-grade glioma,8,11,65,

83,91,93,101,103,104,117 controversy persists, mainly because of therisk of inducing neurological sequelae, especially when theglioma is located in eloquent brain areas. Permanent deficitrates between 13 and 20% have been reported in severalstudies.3,18,40,116,129

Nonetheless, technical developments in functional map-ping methods in the last decade have allowed for improve-ments in surgical strategies and thus postoperative out-come. First, the use of preoperative noninvasive methods offunctional neuroimaging (PET scanning, fMR imaging, andmagnetoencephalography) allows one to estimate the loca-tion of cortical eloquent areas in relation to the glioma, thusenabling one to plan the best surgical approach and limits ofresection to avoid functional cortex.1,4,19,41,44,46,55,58,61,62,67,75,77,87,

89,96,110,119,125,128,136 Second, image-guided surgery allows theneurosurgeon to benefit from individual anatomical data aswell as the incorporation of preoperative functional neuro-imaging in the neuronavigational system.16,19,47,48,71,80,84,92,95,108,

113,122 Third, recording of evoked potentials during surgery iswidely described for lesions located within the central re-gion, to identify the rolandic sulcus to avoid sensorimotordeficits.2,5,18,45,47,48,54,80,82,86,87,111,114,135 Fourth, the use of electri-cal cortical stimulations in patients in whom either generalor local anesthesia has been induced is advocated to identi-fy eloquent areas intraoperatively15,25,26,37,38,41,46,55,58,60,68,69,71,72,79,

80,84,85,89,94,102,111,112,121,124,125,127,131,140 or, more rarely, preoperative-ly with the aid of subdural grids, although more often in epi-lepsy surgery57,78,118 than in tumor surgery.88

Despite improvements in postoperative functional resultsby using these mapping methods, the identification andpreservation of eloquent pathways within the white mat-ter remains problematic. This issue is particularly acute insurgery for low-grade gliomas, because this kind of tumormost often involves cortical and subcortical structures and,moreover, shows an infiltrative progression along the whitefibers.23,49 Thus, we report on the usefulness of intraopera-tive real-time direct electrical cortical and subcortical stim-ulations performed repetitively at every moment and eachsite of resection, to identify and to preserve eloquent cor-tical sites and subcortical pathways essential for sensori-motor and/or language functions. Indeed, in the six patientswho maintained definitive neurological worsening, threedeficits were due to a deep stroke (lenticulostriate arteriesdamaged during frontotemporoinsular low-grade glioma re-section) and three others were caused by a glioma invasionof the entire central region, which had induced a preop-erative paresis that was likely explained by a limit of theplasticity mechanisms often involved to compensate for gli-omas in eloquent areas.4,27,30,41,123,136 In other words, no post-operative sequela was due to the interruption of the white

fibers. Thus, this method should not be considered as an al-ternative to the mapping techniques mentioned earlier, butrather as a complementary method offering functional datathat cannot be provided using the other techniques.

Complement to Preoperative Neurofunctional Imaging

In addition to the fact that the sensitivity of these corti-cal mapping methods is still not optimal despite constantefforts for its improvement (sensitivity ranging from 82–100% for the identification of sensorimotor sites12,16,58,62,71,77,

106,107,110,113,121,122,137,139 and from 77–100% for the identificationof the language areas14,15,42,55,58,100,114,115,120,128), the actual map-ping of white fibers remains impossible for technical rea-sons. This point was emphasized in the work of Hirsch, etal.:58 “fMRI has little sensitivity for subcortical areas and,particularly, for white matter tracts. It is for this reason thatfunctional imaging is a useful guide in preoperative and in-traoperative planning, but does not replace the need forcareful intraoperative mapping in regions of cerebral elo-quence.” Data from our present study demonstrate furtherproof of this point of view, and we assert that the use of in-traoperative cortical and subcortical stimulations (comple-menting preoperative neurofunctional mapping) representsa valuable adjunct to avoid the occurrence of a definitivepostoperative deficit. Nonetheless, there are other recentdevelopments in MR imaging that can help delineate deepwhite matter tracts, specifically diffusion-tensor imaging. Acomparison between this new method and subcortical stim-ulation mapping should be studied, as suggested by Gossl,et al.50

Complement to Image-Guided Surgery

Despite the integration of functional data provided byneurofunctional imaging within systems of neuronaviga-tion,16,19,47,48,71,80,84,92,95,108,113,122 the problem of identifying sub-cortical pathways is still not solved because of an inabilityto map the white matter by using these methods, as detailedearlier. For this reason, some authors have recently suggest-ed incorporating data from the intraoperative image-guid-ed system with that provided by preoperative DW MR im-aging, a recently developed imaging method capable oftracking the white fibers.21,63,74,90 This technique has yet tobe validated, however, particularly concerning its reliabili-ty, accuracy, and sensitivity. Furthermore, its integration in-to the neuronavigational system still involves the problemof brain shift.56,109 Indeed, Coenen, et al.,20 recently reportedon a patient with a right temporoinsular glioma who had un-dergone surgery aided by the integration of spatial three-dimensional information on the pyramidal tracts, whichhad been acquired using anisotropic DW imaging, into acustomized system for frameless neuronavigation. Duringsurgery and tumor decompression, navigation became inac-curate because of brain shift and the patient suffered post-operative hemiplegia due to damage of the internal capsule,which was confirmed on control MR imaging. We sug-gest that the use of repetitive direct subcortical stimulationscould have allowed for the identification of the cortico-spinal tracts before their surgical disruption. Moreover, au-thors of a correlation study recently suggested that stimula-tions might represent a method of validating tractographyby using DW imaging (particularly in cases of small lesionswith minimal risk of intraoperative brain shift).59

H. Duffau, et al.


Complement to Intraoperative SSEPs

For the intraoperative identification of the sensorimotorregion, one can stimulate a peripheral nerve and record theindirect SSEP from the cortex.2,5,18,45,47,48,54,82,86,87,111,114,135 Note,however, that the value of motor cortex mapping by us-ing direct cortical stimulations has been previously demon-strated to be higher than by indirect cortical SSEP2 (accu-rate localization of the central sulcus was between 91 and94%).18,66,111,135 Moreover, phase-reversal recording demon-strates only the central sulcus itself and offers no directinformation on the particular distribution of motor functionon the exposed cerebral structures.2,18 Additionally, in casesof subcortical tumor removal, resection can be limited on-ly on the basis of indirect data, that is, when SSEPs beginto be altered. But this method is unable to provide directinformation about the exact location of the thalamocorti-cal ascending pathways or the pyramidal descending fibers.Consequently, there is a permanent double risk: to detectmodifications of SSEPs while the subcortical tracts are al-ready damaged or to have a higher sensitivity of SSEPsleading to the premature interruption of glioma removal,when the resection was not yet in contact with the function-al pathways. Again, the use of repetitive subcortical elec-trical stimulations may allow one to solve this problem bythe direct identification of sensorimotor fibers when theyare encountered.

Complements to Intraoperative DirectCortical Stimulations

Motor Evoked Potentials in Patients in a State of GeneralAnesthesia. Although this method was improved with mod-ified electrical parameters,130 repetitive intraoperative stim-ulations of the motor cortex rather than a single stimu-lation,68 and its combined use with SSEPs,18,69,72,140 severalproblems persist. First, when recording compound muscleaction potentials, only the monitored muscles can be con-trolled; that is, there is an inability to detect and possibly toavoid motor deficits in unmonitored muscles. For this rea-son, Cedzich, et al.,18 asserted that monitoring of “all mus-cle groups at risk” seemed necessary. In the case of exten-sive low-grade gliomas involving subcortical regions wherethe pyramidal tracts converge, however, we showed that allparts of the contralateral hemibody may produce motor re-sponses, making it difficult to predict or define all musclegroups at risk before resection. Second, as with SSEPs,motor evoked potentials give only indirect informationabout the location of pyramidal pathways, even with the useof repetitive stimulations, and do not allow for the directidentification of motor tracts when the resection interferes.In the same way, although authors of a recent study showedthat a reduction in amplitude greater than 80% and a pro-longed latency of more than 15% can be interpreted asintraoperative warning signs of potential mechanical dam-age to the motor system and lead to the end of resection,they emphasized that the occurrence of a complete or near-ly complete and sudden nonartifactual reduction in ampli-tude could be attributed to a lesion of the subcortical pyram-idal fibers, and that such irreversible changes in potentialcannot serve as a warning sign in such cases, because theyoccur only after damage.68 These limitations of the evokedpotential techniques may explain why 20% of sequelae,mainly due to resection of subcortical lesions in the imme-

diate vicinity of the pyramidal tract, have been reported byCedzich, et al.,18 who concluded that “an additional methodis necessary in patients requiring localization deep in thewhite matter and when tumors involving motor pathwaysare to be removed.” We argue that the use of repetitive intra-operative subcortical stimulations could offer the comple-mentary functional data not provided by evoked potentials,which does not exclude the simultaneous use of a multi-channel electromyographic recording, as previously sug-gested by Yingling, et al.138

Direct Cortical Sensorimotor and Language Mappings inPatients After Induction of General or Local Anesthesia. Al-though authors of a large number of series reported on theuse of direct electrical cortical stimulation in patients afterinduction of general or local anesthesia during surgery fortumors in eloquent brain areas,15,25,26,37,38,41,46,55,58,60,68,69,71,72,79,

80,84,85,89,94,102,111,112,121,124,125,127,131,140 very few mentioned an inter-est in also performing intraoperative repetitive subcorticalstimulations,39,133 except Berger and colleagues.6,7,9–11,51,52,81,

126,138 Consequently, prolonged and even permanent postop-erative neurological worsenings were reported due to thelack of accurate and repetitive white matter mapping. Ina recent study in which the resection of a low-grade glio-ma involved the supplementary motor area, Peraud, et al.,102

reported that some patients displayed severe paresis withslow recovery because of a lesion probably occurring “inthe depth of the white matter, when the most posterolateralpart of the tumor [wa]s removed.” In the comments to thispaper, Piepmeier confirmed that the “interruption of the de-scending motor fibers is a common cause of unanticipatedpermanent motor deficits.” Thus, Peraud, et al., concludedthat because of this risk of damage to the pyramidal tract atthe level of its deep infliction, “surgery for Grade II glio-mas in the superior frontal gyrus is more likely to result inpermanent morbidity when the resection is performed at adistance of less than 0.5 cm from the precentral gyrus orpositive stimulation points.” In recent publications summa-rizing their entire experience,25,94 the team of Danks and col-leagues, who accumulated a large series of patients withlow-grade gliomas,13,26,93 described their policy of resectionas being limited to within 0.5 cm of the primary sensorimo-tor cortex (Roux, et al.,113 had a limit of 1 cm). Note thatDanks, et al.,25 mentioned that they did not use stimulationmapping of subcortical white matter in their series and ac-cording to Nikas, et al.,94 “subcortical white matter stimu-lation mapping is occasionally performed.” Thus, even ifa complete resection rate of 57% was achieved (includingpilocytic astrocytoma, ganglioglioma, and ependymoma,that is, well-demarcated tumors, with only 28% of a realWorld Health Organization Grade II low-grade gliomas),this rate could be optimized if no margin was left around theeloquent areas, because many low-grade gliomas come intothe contact and even invade the primary sensorimotor sites.

On the basis of our experience, we were prone to stopresection immediately on contact with functional cortico-subcortical structures, a policy that automatically allowedus to improve the quality of low-grade glioma removal; thatis, 30% of all resections demonstrated no signal abnor-mality on repeated MR images, 27% resulted in less than 5cm3 of residual tumor, and 23% resulted in a tumor residuebetween 5 and 10 cm3 (totalling 80% of so-called complete[30%] and near-complete [50%] resections). This strategyseemed possible with a good reliability only because we



773

systematically used the repeated subcortical stimulationsduring the resection to follow the white fibers closely, fromthe surface to the depth.31,33–36,43 This procedure allowed usto anticipate a permanent deficit, except in six patients forreasons previously explained (lenticulostriate arteries dam-aged in three cases and diffuse glioma infiltration of theentire central region in the other three cases), but never dueto a surgical interruption of the subcortical pathways. Nev-ertheless, for reliability, this method necessitates regularlyrepeated subcortical stimulations. Indeed, in a recent seriesof patients with insular gliomas, Lang, et al.,73 used directsubcortical stimulation, but asserted that this technique “of-ten does not provide adequate warning of ensuing neuro-logical function, but merely reports the loss of neurologicalfunction. For example, interruption of motor fibers couldoccur between direct brain stimulation trials.” In their meth-odology, however, these authors stated that “subcorticalstimulation was used infrequently during tumor resection.”This can perfectly explain why a surgical injury to the py-ramidal tract may occur between two stimulation trials.Thus, online stimulations should be performed during theentire glioma removal procedure.

Finally, although authors of many studies have raised theissue of preserving (sensori)motor fibers, very few havestudied the language pathways except Berger and col-leagues.6,7,11,51,52,81,126 Indeed, most researchers perform elec-trical cortical mapping in patients harboring a tumor withinlanguage areas and the patients remain awake during resec-tion.55,124 In these cases, however, the patient performs func-tional tasks without the benefit of subcortical mapping.24,25,

37,85,93,102, 131,132 We think that the use of subcortical stimulationrepresents a valuable adjunct in patients in whom local an-esthesia has been induced. Indeed, the main problem of thisprocedure is its duration, followed by the fatigue of the pa-tient. Consequently, in the last stages of resection, patientsoften demonstrate a slowness in their language. At thistime, it is not easy to differentiate language disturbancesdue to fatigue or the fact that the resection has interferedwith language pathways. Thus, the surgeon must distin-guish between two risks: either to stop the resection prema-turely while speech disturbances are due to patient fatigue(and risk leaving residual tumor that could have been re-moved without inducing permanent aphasia), or to continueglioma removal while speech disorders represent a warningof injury to the language pathways (and risk causing a ma-jor language deficit during the ensuing resection). To solvethis problem, we advise the use of direct subcortical stim-ulation in a patient who successfully performs functionaltasks after several minutes of rest. If the stimulation doesnot disrupt language, the surgeon will know that it is possi-ble to complete the tumor resection without additional riskof sequelae. Conversely, if stimulation elicits the slightestspeech disturbance, an immediate interruption of the resec-tion is clearly mandatory.

Our experience is in accordance with those of the few re-searchers who have relied on the systematic use of the intra-operative subcortical stimulations in association with corti-cal mapping. Indeed, Vinas, et al.,133 reported that “duringresection of the structural lesions, intraoperative stimulationwas continued in the subcortical pathways, and five patientshad positive responses on areas not identified by the func-tional PET.” They then concluded that “intraoperative map-ping still remains the main means to avoid neurological

damage as it can be performed during the entire surgicalprocedure to avoid damage to cortex and pathways.” Thiswas a small series, however, including only 12 patients.More recently, Eisner, et al.,39 also showed that “electro-physiologically controlled surgery by bipolar electric stim-ulation of the cerebral cortex and descending fiber sys-tem is a very effective instrument for detecting cerebralfunctions in exact spatial distribution.” Again, this was aheterogeneous and small series. It was the team of Bergerand colleagues6,7,9,10,11,51,52,81,126,138 who honed the regular useof intraoperative cortical and subcortical electrical mappingin glioma surgery. Our experience with a homogeneousconsecutive series of 103 patients with low-grade gliomasseems clearly to confirm the value of such a procedure. Al-though Berger and colleagues reported that a margin of 7 to10 mm around the language areas resulted in significantlyfewer permanent postoperative linguistic deficits, we notedno higher rate of definitive language worsening despite a re-section coming in contact of the language sites (but a high-er rate of transient postoperative aphasia). This discrepancymay be partly explained in the work of Haglund, et al.,52

regarding the sphere of influence of electrocortical stim-ulation—all gliomas in their patients were located in thetemporal lobe, whereas many low-grade gliomas in patientsin our study invaded the frontal lobe. Indeed, results of a re-cent fMR imaging/electrostimulation correlation study105

demonstrated that in the temporal lobe, an electrostimula-tion radius of 10 mm must be assumed to achieve a near-ly 100% sensitivity in fMR imaging, whereas in the fron-tal lobe, a smaller stimulation radius of 5 mm can be usedto achieve 100% sensitivity. These findings are consistentwith reports that essential language sites are more concen-trated in the frontal rather than in the temporal lobe.98,99

Limitation of Intraoperative SubcorticalElectrical Mapping

Despite the contribution of subcortical stimulation, twokinds of limitation persist: functional and oncological. 1)Functional limitation: all resections in our patients werestopped when functional fibers detected on stimulationwere encountered. No postoperative sequela was due to theinterruption of the white fibers. Despite systematically ver-ifying the integrity of functional networks at the end of theglioma removal, however, neurological functioning in 80%of patients worsened in the immediate postoperative stage.This could be explained by the fact that secondary eloquentcorticosubcortical areas—namely those participating in, butnot essential to, function—were surgically removed (suchas the supplementary motor area or the insula, known toinduce a possible transient dysfunction with secondary re-covery when resected).31,36,43,70,73,85,140 Given that the surgicalmethodology was the same in all 103 patients, our data indi-cate that the subcortical stimulations allow one to identifyand then to preserve the pathways essential to a function,but cannot be used to predict the occurrence or the intensi-ty of a transitory postoperative deficit.

2) Oncological limitation: despite additional publishedarguments favoring the positive impact of the quality of re-section on the natural history of a low-grade glioma,8,11,65,83,91,

93,101,103,104,117 there is currently no absolute certainty regard-ing the long-term efficacy of such a therapeutic strategy.Consequently, although subcortical mapping may allow for

H. Duffau, et al.


greater tumor resection while sparing deep white mattertracts and therefore eloquent function, we have yet to provean impact on long-term survival. Thus, further series inwhich researchers study oncological results of low-gradeglioma surgery with a long follow up are mandatory to de-termine whether a less aggressive resection with minimiza-tion of the transient immediate postoperative deficit wouldbe better for the patient.

Conclusions

The 94% rate of favorable functional results in this seriessupports the systematic use of cortical and repetitive sub-cortical electrical stimulations as an adjunct to other meth-ods of functional mapping during resection of low-gradegliomas in eloquent areas. Such a strategy allows the neu-rosurgeon to perform a tumor resection according to func-tional boundaries and then to optimize the benefit (qualityof glioma removal)/risk (postoperative definitive deficit)ratio. The drawback of this procedure is the frequent oc-currence of immediate postoperative neurological wors-ening—80% in our series, which was in accordance withrecent data from the Mayo Clinic experience of 74% de-ficits.85 Although these deficits are secondarily regres-sive, they often necessitate rehabilitation therapy for 1 to 3months, and another 3 to 6 months to return to a normalsocioprofessional life. Consequently, it is mandatory to giveclear information to the patient and his or her family preop-eratively, which can be most specific by using preoperativeneurofunctional imaging.70 Also note that beyond function-al recovery, a patient’s delayed postoperative status may beimproved in comparison with his or her preoperative status(10% of patients in our series) and epilepsy better controlledin 80% of cases.32

Moreover, the use of intraoperative real-time direct sub-cortical stimulation can lead to a better understanding ofthe brain functional connectivity—especially complex andpoorly studied with regard to language34—and thus like-ly allow for future improvements in surgical preplanning.Data from both cortical and subcortical functional map-pings should be used and perhaps partly provided by DWMR imaging tractography if this method can be validatedusing subcortical stimulations.

References

1. Alberstone CD, Skirboll SL, Benzel EC, et al: Magnetic sourceimaging and brain surgery: presurgical and intraoperative plan-ning in 26 patients. J Neurosurg 92:79–90, 2000

2. Allison T, McCarthy G, Wood CC, et al: Human cortical poten-tials evoked by stimulation of the median nerve. II. Cytoarchitec-tonic areas generating short-latency activity. J Neurophysiol 62:694–710, 1989

3. Apuzzo MLJ: Brain Surgery: Complication Avoidance andManagement. New York: Churchill Livingstone, 1993, Vol 1, pp379–390

4. Atlas SW, Howard RS II, Maldjian J, et al: Functional magneticresonance imaging of regional brain activity in patients with intra-cerebral gliomas: findings and implications for clinical manage-ment. Neurosurgery 38:329–338, 1996

5. Babu KS, Chandy MJ: Reliability of somatosensory evoked po-tentials in intraoperative localization of the central sulcus inpatients with perirolandic mass lesions. Br J Neurosurg 11:411–417, 1997

6. Berger MS: Functional mapping-guided resection of low-gradegliomas. Clin Neurosurg 42:437–452, 1995

7. Berger MS: Lesions in functional (“eloquent”) cortex and subcor-tical white matter. Clin Neurosurg 41:444–463, 1994

8. Berger MS, Deliganis AV, Dobbins JD, et al: The effect of extentof resection on recurrence in patients with low grade cerebralhemisphere gliomas. Cancer 74:1784–1791, 1994

9. Berger MS, Ojemann GA: Intraoperative brain mapping tech-niques in neuro-oncology. Stereotact Funct Neurosurg 58:153–161, 1992

10. Berger MS, Ojemann GA, Lettich E: Neurophysiological moni-toring during astrocytoma surgery. Neurosurg Clin N Am 1:65–80, 1990

11. Berger MS, Rostomily RC: Low grade gliomas: functional map-ping resection strategies, extent of resection, and outcome. J Neu-rooncol 34:85–101, 1997

12. Bittar RG, Olivier A, Sadikot AF, et al: Localization of somatosen-sory function by using positron emission tomography scanning: acomparison with intraoperative cortical stimulation. J Neurosurg90:478–483, 1999

13. Black PM, Ronner SF: Cortical mapping for defining the limits oftumor resection. Neurosurgery 20:914–919, 1987

14. Bookheimer SY, Zeffiro TA, Blaxton T, et al: A direct compari-son of PET activation and electrocortical stimulation mapping forlanguage localization. Neurology 48:1056–1065, 1997

15. Brannen JH, Badie B, Moritz CH, et al: Reliability of functionalMR imaging with word-generation tasks for mapping Broca’s ar-ea. AJNR 22:1711–1718, 2001

16. Braun V, Dempf S, Tomczak R, et al: Functional cranial neuro-navigation. Direct integration of fMRI and PET data. J Neurora-diol 27:157–163, 2000

17. Cabantog AM, Bernstein M: Complications of first craniotomy forintra-axial brain tumour. Can J Neurol Sci 21:213–218, 1994

18. Cedzich C, Taniguchi M, Schäfer S, et al: Somatosensory evokedpotential phase reversal and direct motor cortex stimulation dur-ing surgery in and around the central region. Neurosurgery 38:962–970, 1996

19. Chabrerie A, Nabavi A, Ozlen F, et al: Three-dimensional recon-struction for cortical surgery: the Brigham and Women’s Hospitalexperience, in Loftus CM, Batjer HH (eds): Techniques in Neu-rosurgery. Philadelphia: Lippincott Williams & Wilkins, 2001,Vol 7, pp 61–69

20. Coenen VA, Krings T, Mayfrank L, et al: Three-dimensional visu-alization of the pyramidal tract in a neuronavigation system dur-ing brain tumor surgery: first experiences and technical note. Neu-rosurgery 49:86–93, 2001

21. Conturo TE, Lori NF, Cull TS, et al: Tracking neuronal fiber path-ways in the living human brain. Proc Natl Acad Sci USA 96:10422–10427, 1999

22. Cote R, Battista RN, Wolfson C, et al: The Canadian Neurolog-ical Scale: validation and reliability assessment. Neurology 39:638–643, 1989

23. Couldwell WT, Yong VW: Glioma invasion. J Neurosurg 95:379–380, 2001

24. Cukiert A, Gronich G, Marino R Jr: Surgical approaches to tumorsand epileptogenic zones close to Wernicke’s area. Arq Neuropsi-quiatr 53:587–591, 1995

25. Danks RA, Aglio LS, Gugino LD, et al: Craniotomy under lo-cal anesthesia and monitored conscious sedation for the resectionof tumors involving eloquent cortex. J Neurooncol 49:131–139,2000

26. Danks RA, Rogers M, Aglio LS, et al: Patient tolerance of crani-otomy performed with the patient under local anesthesia and mon-itored conscious sedation. Neurosurgery 42:28–36, 1998

27. Duffau H: Acute functional reorganisation of the human motorcortex during resection of central lesions: a study using intra-operative brain mapping. J Neurol Neurosurg Psychiatry 70:506–513, 2001

28. Duffau H: Intraoperative direct subcortical stimulation for identi-



775

fication of the internal capsule, combined with an image-guidedstereotactic system during surgery for basal ganglia lesions. SurgNeurol 53:250–254, 2000

29. Duffau H: Recovery from complete hemiplegia following re-section of a retrocentral metastasis: the prognostic value of intra-operative cortical stimulations. Case report. J Neurosurg 95:1050–1052, 2001

30. Duffau H, Bauchet L, Lehéricy S, et al: Functional compensa-tion of the left dominant insula for language. Neuroreport 12:2159–2163, 2001

31. Duffau H, Capelle L, Lopes M, et al: The insular lobe: physio-pathological and surgical considerations. Neurosurgery 47:801–811, 2000

32. Duffau H, Capelle L, Lopes M, et al: Medically intractable epilep-sy from insular low-grade gliomas: improvement after an extend-ed lesionectomy. Acta Neurochir 144:563–573, 2002

33. Duffau H, Capelle L, Sichez J, et al: Intra-operative direct electri-cal stimulations of the central nervous system: the Salpêtrière ex-perience with 60 patients. Acta Neurochir 141:1157–1167, 1999

34. Duffau H, Capelle L, Sichez N, et al: Intraoperative mapping ofthe subcortical language pathways using direct stimulations. Ananatomo-functional study. Brain 125:199–214, 2002

35. Duffau H, Denvil D, Capelle L: Absence of movement disordersafter surgical resection of glioma invading the right striatum. JNeurosurg 97:363–369, 2002

36. Duffau H, Lopes M, Denvil D, et al: Delayed onset of supple-mentary motor area syndrome after surgical resection of the me-sial frontal lobe: a time course study using intraoperative map-ping in an awake patient. Stereotact Funct Neurosurg 76:74–82,2001

37. Ebel H, Ebel M, Schillinger G, et al: Surgery of intrinsic cerebralneoplasms in eloquent areas under local anesthesia. Minim Inva-sive Neurosurg 43:192–196, 2000

38. Ebeling U, Schmid UD, Ying H, et al: Safe surgery of lesions nearthe motor cortex using intra-operative mapping techniques: a re-port on 50 patients. Acta Neurochir 119:23–28, 1992

39. Eisner W, Burtscher J, Bale R, et al: Use of neuronavigation andelectrophysiology in surgery of subcortically located lesions in thesensorimotor strip. J Neurol Neurosurg Psychiatry 72:378–381,2002

40. Fadul C, Wood J, Thaler H, et al: Morbidity and mortality of cra-niotomy for excision of supratentorial gliomas. Neurology 38:1374–1379, 1988

41. Fandino J, Kollias SS, Wieser HG, et al: Intraoperative validationof functional magnetic resonance imaging and cortical reorgani-zation patterns in patients with brain tumors involving the prima-ry motor cortex. J Neurosurg 91:238–250, 1999

42. FitzGerald DB, Cosgrove GR, Ronner S, et al: Location of lan-guage in the cortex: a comparison between functional MR imag-ing and electrocortical stimulation. AJNR 18:1529–1539, 1997

43. Fontaine D, Capelle L, Duffau H: Somatotopy of the supplemen-tary motor area: evidence by correlation of the extent of surgicalresection with the clinical patterns of deficit. Neurosurgery 50:297–305, 2002

44. Fried I, Nenov VI, Ojemann SG, et al: Functional MR and PETimaging of rolandic and visual cortices for neurosurgical plan-ning. J Neurosurg 83:854–861, 1995

45. Gallen CC, Schwartz BJ, Bucholz RD, et al: Presurgical localiza-tion of functional cortex using magnetic source imaging. J Neuro-surg 82:988–994, 1995

46. Gallen CC, Sobel DF, Waltz T, et al: Noninvasive presurgical neu-romagnetic mapping of somatosensory cortex. Neurosurgery 33:260–268, 1993

47. Ganslandt O, Fahlbusch R, Nimsky C, et al: Functional neuro-navigation with magnetoencephalography: outcome in 50 patientswith lesions around the motor cortex. J Neurosurg 91:73–79,1999

48. Ganslandt O, Steinmeier R, Kober H, et al: Magnetic source im-aging combined with image-guided frameless stereotaxy: a new

method in surgery around the motor strip. Neurosurgery 41:621–628, 1997

49. Giese A, Westphal M: Glioma invasion in the central nervous sys-tem. Neurosurgery 39:235–250, 1996

50. Gossl C, Fahrmeir L, Putz B, et al: Fiber tracking from DTI usinglinear state space models: detectability of the pyramidal tract.Neuroimage 16:378–388, 2002

51. Haglund MM, Berger M: Functional mapping of motor sensory,and language pathways during low-grade glioma removal. TechNeurosurg 2:141–149, 1996

52. Haglund MM, Berger MS, Shamseldin M, et al: Cortical localiza-tion of temporal lobe language sites in patients with gliomas. Neu-rosurgery 34:567–576, 1994

53. Harris AJ: Harris Test of Lateral Dominance. Manual of Direc-tions for Administration and Interpretation. New York: Psy-chological Corporation, 1947

54. Helmers SL: Evoked potentials for cortical mapping in childrenand adults, in Loftus CM, Batjer HH (eds): Techniques in Neu-rosurgery. Philadelphia: Lippincott Williams & Wilkins, 2001,Vol 7, pp 4–11

55. Herholz K, Reulen HJ, von Stockhausen HM, et al: Preoperativeactivation and intraoperative stimulation of language-related areasin patients with glioma. Neurosurgery 41:1253–1262, 1997

56. Hill DL, Maurer CR Jr, Maciunas RJ, et al: Measurement of intra-operative brain surface deformation under a craniotomy. Neuro-surgery 43:514–528, 1998

57. Hill DLG, Smith AD, Simmons A, et al: Sources of error in com-paring functional magnetic resonance imaging and invasive elec-trophysiological recordings. J Neurosurg 93:214–223, 2000

58. Hirsch J, Ruge MI, Kim KHS, et al: An integrated functional mag-netic resonance imaging procedure for preoperative mapping ofcortical areas associated with tactile, motor, language, and visualfunctions. Neurosurgery 47:711–722, 2000

59. Holodny AI, Schwartz TH, Ollenschleger M, et al: Tumor in-volvement of the corticospinal tract: diffusion magnetic resonancetractography with intraoperative correlation. Case illustration. JNeurosurg 95:1082, 2001

60. Horstmann GA, Talies S, Westermann B, et al: Microstereomet-rically guided cortical stimulation for the intraoperative identifica-tion of the central motor strip. Stereotact Funct Neurosurg 65:130–135, 1995

61. Hund M, Rezai AR, Kronberg E, et al: Magnetoencephalographicmapping: basis of a new functional risk profile in the selection ofpatients with cortical brain lesions. Neurosurgery 40:936–943,1997

62. Inoue T, Shimizu H, Nakasato N, et al: Accuracy and limitation offunctional magnetic resonance imaging for identification of thecentral sulcus: comparison with magnetoencephalography in pa-tients with brain tumors. Neuroimage 10:738–748, 1999

63. Inoue T, Shimizu H, Yoshimoto T: Imaging the pyramidal tractsin patients with brain tumors. Clin Neurol Neurosurg 101:4–10,1999

64. Karnofsky DA, Burchenal JH: The clinical evaluation of chemo-therapeutic agents in cancer, in MacLeod CM (ed): Evaluationof Chemotherapeutic Agents. New York: Columbia UniversityPress, 1949, pp 191–205

65. Keles GE, Lamborn KR, Berger MS: Low-grade hemisphericgliomas in adults: a critical review of extent of resection as a fac-tor influencing outcome. J Neurosurg 95:735–745, 2001

66. King RB, Schell GR: Cortical localization and monitoring duringcerebral operations. J Neurosurg 67:210–219, 1987

67. Kober H, Möller M, Nimsky C, et al: New approach to localizespeech relevant brain areas and hemispheric dominance usingspatially filtered magnetoencephalography. Hum Brain Mapp14:236–250, 2001

68. Kombos T, Suess O, Ciklatekerlio O, et al: Monitoring of intraop-erative motor evoked potentials to increase the safety of surgery inand around the motor cortex. J Neurosurg 95:608–614, 2001

69. Kombos T, Suess O, Funk T, et al: Intra-operative mapping of the

H. Duffau, et al.


motor cortex during surgery in and around the motor cortex. ActaNeurochir 142:263–268, 2000

70. Krainik A, Lehéricy S, Duffau H, et al: Role of the supplementarymotor area in motor deficit following medial frontal lobe surgery.Neurology 57:871–878, 2001

71. Krings T, Schreckenberger M, Rohde V, et al: Metabolic and elec-trophysiological validation of functional MRI. J Neurol Neuro-surg Psychiatry 71:762–771, 2001

72. Krombach GA, Spetzger U, Rohde V, et al: Intraoperative local-ization of functional regions in the sensorimotor cortex by neuro-navigation and cortical mapping. Comput Aided Surg 3:64–73,1998

73. Lang FF, Olansen NE, DeMonte F, et al: Surgical resection of in-trinsic insular tumors: complication avoidance. J Neurosurg 95:638–650, 2001

74. Le Bihan D, Mangin JF, Poupon C, et al: Diffusion tensor im-aging: concepts and applications. J Magn Reson Imaging 13:534–546, 2001

75. Lee CC, Ward HA, Sharbrough FW, et al: Assessment of func-tional MR imaging in neurosurgical planning. AJNR 20:1511–1519, 1999

76. Lehéricy S, Cohen L, Bazin B, et al: Functional MR evaluation oftemporal and frontal language dominance compared with the Wa-da test. Neurology 54:1625–1633, 2000

77. Lehéricy S, Duffau H, Cornu P, et al: Correspondence betweenfunctional magnetic resonance imaging somatotopy and individ-ual brain anatomy of the central region: comparison with intraop-erative stimulation in patients with brain tumors. J Neurosurg 92:589–598, 2000

78. Lüders H, Lesser RP, Dinner DS, et al: Localization of corticalfunction: new information from extraoperative monitoring of pa-tients with epilepsy. Epilepsia 29 (Suppl 2):S56–S65, 1988

79. Lurito JT, Lowe MJ, Sartorius C, et al: Comparison of fMRI andintraoperative direct cortical stimulation in localization of recep-tive language areas. J Comput Assist Tomogr 24:99–105, 2000

80. Makela JP, Kirveskari E, Seppa M, et al: Three-dimensional inte-gration of brain anatomy and function to facilitate intraoperativenavigation around the sensorimotor strip. Hum Brain Mapp 12:180–192, 2001

81. Matz PG, Cobbs C, Berger MS: Intraoperative cortical mapping asa guide to the surgical resection of gliomas. J Neurooncol 42:233–245, 1999

82. McCarthy G, Allison T, Spencer DD: Localization of the face areaof human sensorimotor cortex by intracranial recording of soma-tosensory evoked potentials. J Neurosurg 79:874–884, 1993

83. McCormack BM, Miller DC, Budzilovich GN, et al: Treatmentand survival of low-grade astrocytoma in adults—1977–1988.Neurosurgery 31:636–642, 1992

84. McDonald JD, Chong BW, Lewine JD, et al: Integration of pre-operative and intraoperative functional brain mapping in a frame-less stereotactic environment for lesions near eloquent cortex.Technical note. J Neurosurg 90:591–598, 1999

85. Meyer FB, Bates LM, Goerss SJ, et al: Awake craniotomy for ag-gressive resection of primary gliomas located in eloquent brain.Mayo Clin Proc 76:677–687, 2001

86. Mine S, Oka N, Yamaura A, et al: Presurgical functional localiza-tion of primary somatosensory cortex by dipole tracing method ofscalp-skull-brain head model applied to somatosensory evokedpotential. Electroencephalogr Clin Neurophysiol 108:226–233,1998

87. Morioka T, Shigeto H, Ishibashi H, et al: Magnetic source imag-ing of the sensory cortex on the surface anatomy MR scanning.Neurol Res 20:235–241, 1998

88. Morris HH III, Lüders H, Hahn JF, et al: Neurophysiological tech-niques as an aid to surgical treatment of primary brain tumors.Ann Neurol 19:559–567, 1986

89. Mueller WM, Yetkin FZ, Hammeke TA, et al: Functional mag-netic resonance imaging mapping of the motor cortex in patientswith cerebral tumors. Neurosurgery 39:515–521, 1996

90. Nakada T, Nakayama N, Fujii Y, et al: Clinical application ofthree-dimensional anisotropy contrast magnetic resonance axon-ography. Technical note. J Neurosurg 90:791–795, 1999

91. Nakamura M, Konishi N, Tsunoda S, et al: Analysis of prognos-tic and survival factors related to treatment of low-grade astrocy-toma in adults. Oncology 58:108–116, 2000

92. Nariai T, Senda M, Ishii K, et al: Three-dimensional imaging ofcortical structure, function and glioma for tumor resection. J Nu-cl Med 38:1563–1568, 1997

93. Nikas DC, Bello L, Zamani AA, et al: Neurosurgical considera-tions in supratentorial low-grade gliomas: experience with 175patients. Neurosurg Focus 4 (4):Article 4, 1998

94. Nikas DC, Danks RA, Black PM: Tumor surgery under local an-esthesia, in Loftus CM, Batjer HH (eds): Techniques in Neuro-surgery. Philadelphia: Lippincott Williams & Wilkins, 2001,Vol 7, pp 70–84

95. Nimsky C, Ganslandt O, Kober H, et al: Integration of func-tional magnetic resonance imaging supported by magnetoen-cephalography in functional neuronavigation. Neurosurgery 44:1249–1256, 1999

96. Nitschke MF, Melchert UH, Hahn C, et al: Preoperative func-tional magnetic resonance imaging (fMRI) of the motor systemin patients with tumors in the parietal lobe. Acta Neurochir 140:1223–1229, 1998

97. Ojemann G, Ojemann J, Lettich E, et al: Cortical language lo-calization in left, dominant hemisphere. An electrical stimula-tion mapping investigation in 117 patients. J Neurosurg 71:316–326, 1989

98. Ojemann GA: Functional mapping of cortical language areasin adults. Intraoperative approaches. Adv Neurol 63:155–163,1993

99. Ojemann JG, Ojemann GA, Lettich E: Cortical stimulation map-ping of language cortex by using a verb generation task: effectsof learning and comparison to mapping based on object naming.J Neurosurg 97:33–38, 2002

100. Papanicolaou AC, Simos PG, Breier JI, et al: Magnetoencephalo-graphic mapping of the language-specific cortex. J Neurosurg90:85–93, 1999

101. Peraud A, Ansari H, Bise K, et al: Clinical outcome of su-pratentorial astrocytoma WHO grade II. Acta Neurochir 140:1213–1222, 1998

102. Peraud A, Meschede M, Eisner W, et al: Surgical resection ofgrade II astrocytomas in the superior frontal gyrus. Neurosur-gery 50:966–977, 2002

103. Philippon JH, Clémenceau SH, Fauchon FH, et al: Supratentoriallow-grade astrocytomas in adults. Neurosurgery 32:554–559,1993

104. Piepmeier J, Christopher S, Spencer D, et al: Variations in thenatural history and survival of patients with supratentorial low-grade astrocytomas. Neurosurgery 38:872–879, 1996

105. Pouratian N, Bookheimer SY, Rex DE, et al: Utility of preoper-ative functional magnetic resonance imaging for identifying lan-guage cortices in patients with vascular malformations. J Neuro-surg 97:21–32, 2002

106. Puce A, Constable RT, Luby ML, et al: Functional magnetic res-onance imaging of sensory and motor cortex: comparison withelectrophysiological localization. J Neurosurg 83:262–270,1995

107. Pujol J, Conesa G, Deus J, et al: Clinical application of function-al magnetic resonance imaging in presurgical identification ofthe central sulcus. J Neurosurg 88:863–869, 1998

108. Rezai AR, Hund M, Kronberg E, et al: The interactive use ofmagnetoencephalography in stereotactic image-guided neurosur-gery. Neurosurgery 39:92–102, 1996

109. Roberts DW, Hartov A, Kennedy FE, et al: Intraoperative brainshift and deformation: a quantitative analysis of cortical displace-ment in 28 cases. Neurosurgery 43:749–760, 1998

110. Roberts TP, Ferrari P, Perry D, et al: Presurgical mapping withmagnetic source imaging: comparisons with intraoperative find-ings. Brain Tumor Pathol 17:57–64, 2000



777

111. Romstöck J, Fahlbusch R, Ganslandt O, et al: Localisation of thesensorimotor cortex during surgery for brain tumours: feasibilityand waveform patterns of somatosensory evoked potentials. JNeurol Neurosurg Psychiatry 72:221–229, 2002

112. Roux FE, Boulanouar K, Ranjeva JP, et al: Usefulness of motorfunctional MRI correlated to cortical mapping in Rolandic low-grade astrocytomas. Acta Neurochir 141:71–79, 1999

113. Roux FE, Ibarrola D, Tremoulet M, et al: Methodological andtechnical issues for integrating functional magnetic resonanceimaging data in a neuronavigational system. Neurosurgery 49:1145–1157, 2001

114. Ruge MI, Victor J, Hosain S, et al: Concordance between func-tional magnetic resonance imaging and intraoperative languagemapping. Stereotact Funct Neurosurg 72:95–102, 1999

115. Rutten GJ, Ramsey NF, van Rijen PC, et al: Development of afunctional magnetic resonance imaging protocol for intraopera-tive localization of critical temporoparietal language areas. AnnNeurol 51:350–360, 2002

116. Sayawa R, Hammoud M, Schoppa D, et al: Neurosurgical out-comes in a modern series of 400 craniotomies for treatment ofparenchymal tumors. Neurosurgery 42:1044–1056, 1998

117. Scerrati M, Roselli R, Iacoangeli M, et al: Prognostic factorsin low grade (WHO grade II) gliomas of the cerebral hemi-spheres: the role of surgery. J Neurol Neurosurg Psychiatry61:291–296, 1996

118. Schaffler L, Lüders HO, Beck GJ: Quantitative comparison oflanguage deficits produced by extraoperative electrical stimula-tion of Broca’s, Wernicke’s, and basal temporal language areas.Epilepsia 37:463–475, 1996

119. Schiffbauer H, Ferrari P, Rowley HA, et al: Functional activitywithin brain tumors: a magnetic source imaging study. Neuro-surgery 49:1313–1321, 2001

120. Schlosser M, Luby M, Spencer DD, et al: Comparative local-ization of auditory comprehension by using functional magneticresonance imaging and cortical stimulation. J Neurosurg 91:626–635, 1999

121. Schreckenberger M, Spetzger U, Sabri O, et al: Localisation ofmotor areas in brain tumour patients: a comparison of preopera-tive [18F]FDG-PET and intraoperative cortical electrostimula-tion. Eur J Nucl Med 28:1394–1403, 2001

122. Schulder M, Maldjian JA, Liu WC, et al: Functional image-guid-ed surgery of intracranial tumors located in or near the sensori-motor cortex. J Neurosurg 89:412–418, 1998

123. Seitz RJ, Huang Y, Knorr U, et al: Large-scale plasticity of thehuman motor cortex. Neuroreport 6:742–744, 1995

124. Signorelli F, Guyotat J, Isnard J, et al: The value of cortical stim-ulations applied to the surgery of malignant gliomas in languageareas. Neurol Sci 22:3–10, 2001

125. Simos PG, Papanicolaou AC, Berier JI, et al: Localization of lan-guage-specific cortex using magnetic source imaging and elec-trical stimulation mapping. J Neurosurg 91:787–796, 1999

126. Skirboll SS, Ojemann GA, Berger MS, et al: Functional cortexand subcortical white matter located within gliomas. Neurosur-gery 38:678–685, 1996

127. Suess O, Kombos T, Hoell T, et al: A new cortical electrode forneuronavigation-guided intraoperative neurophysiological moni-toring: technical note. Acta Neurochir 142:329–332, 2000

128. Szymanski MD, Perry DW, Gage NM, et al: Magnetic sourceimaging of late evoked field responses to vowels: toward an as-sessment of hemispheric dominance for language. J Neurosurg94:445–453, 2001

129. Tandon P, Mahapatra AK, Khosla A: Operations on gliomas in-volving speech centres. Acta Neurochir Suppl 56:67–71, 1993

130. Taniguchi M, Cedzich C, Schramm J: Modification of corticalstimulation for motor evoked potentials under general anesthesia:technical description. Neurosurgery 32:219–226, 1993

131. Taylor MD, Bernstein M: Awake craniotomy with brain map-ping as a routine surgical approach to treating patients with su-pratentorial intraaxial tumors: a prospective trial of 200 cases. JNeurosurg 90:35–41, 1999

132. Vanaclocha V, Saiz-Sapena N, Garcia-Casasola C: Surgicaltreatment of insular gliomas. Acta Neurochir 139:1126–1135,1997

133. Vinas FC, Zamorano L, Mueller RA, et al: [15O]-water PET andintraoperative brain mapping: a comparison in the localization ofeloquent cortex. Neurol Res 19:601–608, 1997

134. Vives KP, Piepmeier JM: Complications and expected outcomeof glioma surgery. J Neurooncol 42:289–302, 1999

135. Wood CC, Spencer DD, Allison T, et al: Localization of humansensorimotor cortex during surgery by cortical surface recordingof somatosensory evoked potentials. J Neurosurg 68:99–111,1988

136. Wunderlich G, Knorr U, Herzog H, et al: Precentral glioma loca-tion determines the displacement of cortical hand representation.Neurosurgery 42:18–27, 1998

137. Yetkin FZ, Mueller WM, Morris GL, et al: Functional MR acti-vation correlated with intraoperative cortical mapping. AJNR18:1311–1315, 1997

138. Yingling CD, Ojemann S, Dobson B, et al: Identification of mo-tor pathways during tumor surgery facilitated by multichannelelectromyographic recording. J Neurosurg 91:922–927, 1999

139. Yousry TA, Schmidt UD, Jassoy AG, et al: Topography of thecortical motor hand area: prospective study with functional MRimaging and direct motor mapping at surgery. Radiology 195:23–29, 1995

140. Zentner J, Hufnagel A, Pechstein U, et al: Functional results afterresective procedures involving the supplementary motor area. JNeurosurg 85:542–549, 1996

Manuscript received August 7, 2002.Accepted in final form November 25, 2002.Address reprint requests to: Hugues Duffau, M.D., Ph.D., Service

de Neurochirurgie, Hôpital de la Salpêtrière, 47-83 Boulevard de l’Hôpital, 75651 Paris, Cedex 13, France. email: [email protected].

H. Duffau, et al.


usefulness of intraoperative electrical subcortical mapping during … · 2007. 5. 3. · resection...

Documents