surgical resection gliomas
DESCRIPTION
Surgical resection of gliomas WHO grade II and IIIlocated in the opercular regionTRANSCRIPT
Acta Neurochir (Wien) (2004) 146: 9–18
DOI 10.1007/s00701-003-0165-4
Clinical ArticleSurgical resection of gliomas WHO grade II and IIIlocated in the opercular region
A. Peraud1, J. Ilmberger2, and H.-J. Reulen1
1 Department of Neurosurgery, Ludwig-Maximilians-University, Munich, Germany2 Department of Physical Medicine and Rehabilitation, Ludwig-Maximilians-University, Munich, Germany
Published online December 11, 2003
# Springer-Verlag 2003
Summary
Background. Surgery in the opercular region especially in the domi-
nant hemisphere impose a major challenge for the neurosurgeon due to
the close vicinity to functional important motor and speech areas. The
purpose of the present study is to analyse on a homogenous patient
group pre- and postoperative functional deficits with regard to different
speech qualities (e.g. aphasia, apraxia), and to correlate these data with
MR and intraoperative monitoring results.
Method. Fourteen patients with suspected low grade astrocytomas in
the opercular region consecutively treated by surgery were eligible for
this study (histology revealed 3 WHO grade III tumours). Degree and
duration of postoperative deficits were retrospectively evaluated accord-
ing to tumour location and boundaries on MR, intraoperative neuro-
monitoring results and extent of tumour resection.
Findings. Postoperatively, 8 patients showed speech or language dis-
turbances, in 4 patients combined with motor deficits mainly of the con-
tralateral upper extremity. Fifty percent of the neuropsychologically tested
patients exhibited speech apraxia while the other 50% had a true aphasic
syndrome. Recovery of the latter deficits was in general faster and more
complete. The severity and duration of postoperative deficits was in good
correlation with the distance of the resection margin to the next positive
stimulation point(s), and a distance of more than 0.5 cm proved to avoid
major impairments. The distribution of functional important stimulation
points in relation to the tumour extension was not predictable, and –
unexpectedly – up to 50% of these sites were found overlaying the tumour.
Interpretation. Surgery for WHO grade II and III gliomas in the
opercular region can result in speech apraxia or an aphasic syndrome
with or without concomitant motor deficits. Intraoperative cortical elec-
trical stimulation is essential in resecting tumours in the opercular region
to avoid permanent morbidity.
Keywords: Glioma; WHO grade II and III; opercular region; func-
tional outcome.
Introduction
Patients with tumours in the opercular region, parti-
cularly in the dominant hemisphere, represent a signifi-
cant challenge with regard to surgical management. The
opercular region is composed of portions of the inferior
frontal gyrus and the inferior parietal lobe surrounding
the insular region. The inferior frontal gyrus with the
Broca area in the pars triangularis is a language relevant
area on the dominant side. Posterior to the pars triangu-
laris lays the pars opercularis which often merges with
the inferior aspect of the precentral gyrus, the primary
motor area for face and hand muscles. Adjacent follows
the inferior portion of the postcentral gyrus, the primary
sensory cortex. Directly behind lays the secondary sen-
sory cortex in the inferior parietal lobule which is essen-
tial for tactile discrimination and pain perception. The
perisylvian cortex also contains the Wernicke area, the
cortex traditionally connected to speech decoding. Most
posteriorly is the angular gyrus responsible for calcula-
tion and reading on the dominant side. In the depth of
the insular region is the primary auditory cortex and
adjacent to this are the basal ganglia and the internal
capsule.
Clinical studies on the surgical outcome for tumours in
the opercular region in contrast to the insular region are
rare [4, 3]. Therefore the data of 14 cases with astrocyto-
mas in the opercular region operated at the Neurosurgical
Department of the Ludwig-Maximilians-University of
Munich between 1994 and 2001 were retrospectively ana-
lysed. All patients were intraoperatively monitored for
motor and=or language function under local or general-
ized anaesthesia. The decision whether gross total or
subtotal tumour resection can be achieved was based
upon the results of intraoperative motor and language
mapping, and the findings and conclusions will be
reported in the present study.
Patients and methods
Patient evaluation
Out of 430 patients operated on four low grade astrocytoma in the
years 1994 until 2001 at the Neurosurgical Department of the Ludwig-
Maximilians-University, 14 tumours were located in the opercular
region. Although all showed typical MR appearance of a low grade
astrocytoma WHO grade II with a well circumscribed hypointensity
on T1-weighted images and without contrast enhancement, three proved
later after histological assessment to have anaplastic features. Medical
records and MR scans of these patients were reviewed and extracted for
the following items: age, sex, symptoms, histology, tumour location,
tumour extension and delineation on T1- and T2-weighted MR images,
extent of tumour resection, results from neuropsychological monitoring
of motor and speech function, postoperative neurological status, follow-
up data and outcome, including the Karnofsky performance status
(KPS). Histology was reevaluated by a Tumour Reference Centre (O. D.
Wiestler, Bonn; P. Kleihues, Z€uurich=Lyon).
Neuroradiology
MR images of all patients were analysed for tumour extensions on
preoperative axial, sagittal and coronal sections of T1-weighted images.
In the series presented here the tumour is located in the opercular part of the
supramarginal, the postcentral, the precentral and the subcentral gyrus=
diagonal sulcus. Vice versa, tumours involving the medial and inferior
insular region, the frontal and superior precentral or parietal lobes were
excluded. Tumour extensions on T1- and T2-weighted images were com-
pared, and inhomogeneities on T2-weighted images and contrast enhance-
ment were registered. In order to compare the individual tumour
extensions, the T1-weighted images were enlarged to a standard size and
the tumour contours transferred to a schematic brain section underlayed by
a 5 mm square grid (Fig. 1a). As described previously, T1-weighted MR
images define the volume of the solid tumour whereas T2-weighted MR
images rather show the infiltrated brain parenchyma [11].
Surgical procedure and cortical electrical stimulation
All patients underwent open surgery for tumour removal. For local
anaesthesia the head of the patient was positioned in a comfortably
moulded vacuum pillow. The skin was infiltrated with 60–80 ml of
0.25% Bupivacain. During the craniotomy the patients received an
ultra-short-acting intravenous analgosedation with Remifentanyl=
Propofol. Following cortical electrical stimulation and tumour resection,
Fig. 1. Diagrams of individual tumour extensions based on axial, coronal and sagittal T1-weighted MR images, enlarged to a standard size and
transferred to a schematic brain section (see methods). An example of a MR image in sagittal and coronal view is given
10 A. Peraud et al.
the latter medication could be continued for the rest of the surgical
procedure [7]. An 8 year old girl with a left sided tumour as well as 2
patients with the tumour in the non-dominant hemisphere underwent
surgery in general anaesthesia.
Following opening of the dura mapping of the motor and language
relevant areas was performed by cortical stimulation according to the
techniques described previously [5, 18]. The interelectrode distance of
the bipolar forceps was 4 mm. Stimulation was done with rectangular
pulses of 0.2 ms duration and a repetition rate of 50=s at intensities of 5
to 12 mA. The stimulation was undertaken in a systematic manner every
eight to ten millimetres starting from anterior to posterior. All sites were
tested at least twice, marked with a sterile numbered label and photo-
graphed for documentation. From the photos the positive and negative
stimulation points were later on transferred to the schematic brain sec-
tion, underlayed by a 5 mm square grid. After resection was completed
the distance from the posterior resection line to the next positive motor
point was measured in mm. Mapping of speech and language functions
was accomplished by a confrontation naming task and the Token Test.
Visual stimuli in form of line drawings of common objects were pre-
sented to the patient together with a carrier phrase (‘‘This is a . . .’’); the
patient’s task was to read aloud the carrier phrase and to add the article
and the name of the object within the stimulation period of 4 seconds.
Ojemann and Ward [16] established this technique, which allows a
distinction between speech and naming errors. A speech arrest was
registered whenever there was no utterance, while slowing of speech
or a dysarthric disturbance was registered as speech disturbance. An
aphasic arrest was recorded when the carrier phrase was correctly uttered
but the name of the object could not be produced, an aphasic disturbance
was defined as an aphasic misnaming (semantic paraphasia, confusion of
expressions). The methods of intraoperative language and speech mon-
itoring as well as the postoperative neuropsychological testing were
standardized over the study period in order to obtain comparable data.
Tumour resection was tailored according to the results of motor and
language mapping, and was performed under microsurgical conditions
with the use of an operating microscope. If no positive areas were found
over the tumour, the resection was carried out along the visible tumour
border or along the confines of a gyrus. If several positive areas were
mapped over the tumour, the resection remained subtotal or only an
extensive biopsy was undertaken.
Neuropsychological testing
Neuropsychological evaluation was performed in those patients with
tumours located in or near language areas. The patients were tested 1 to
3 days preoperatively, seen 1 to 8 days postoperatively, and 8 patients
were reevaluated up to 18 months postoperatively. Neuropsychological
testing included a standard German aphasia test (Aachener Aphasie Test
[10]); this aphasia battery (AB) consists of subtests for naming, repeti-
tion, receptive and written language and the Token Test. Also included in
the neuropsychological evaluation was a test of verbal learning and
memory (VM, M€uunchner Ged€aachtnistest, a German research version
of the California Verbal Learning Test [2]), in which word lists have
to be learned. Free and cued recall as well as recognition scores were
gained as performance indices. All patients with language or motor
deficits received intensive speech and physiotherapy.
Results
Clinical data and preoperative
neurological deficits
The demographic data of the 14 patients are presented
in Table 1. In 10 patients the tumour was on the left
(dominant) side, in 4 patients the tumour was on the
right side and two of them were left handed with the
dominant hemisphere on the right side. Seizures were
the most common initial symptoms (13=14¼ 92.96%),
generalized in 8 and focal in 5 patients. Speech distur-
bances were present preoperatively in 5 patients, but
they occurred in 3 only after seizures. Three patients
showed contralateral muscle weakness, one patient sen-
sory deficits, one patient double vision, and 3 patients
complained of headache.
MR-anatomy and analysis
The precise topography of the opercular region may
differ considerably. According to the publications of
Ebeling and Steinmetz four different types for each the
anterior suprasylvian and the parietal opercular topogra-
phy can be decerned [20, 6]. This anatomy should be
considered, when studying the location of a tumour in
these regions. The superimposed extensions of the indi-
vidual tumours of the present study in sagittal and cor-
onal planes are shown in Fig. 1. All tumours involved
the inferior portion of the pre- and=or postcentral gyrus;
8 tumours showed a more frontal extension involving the
pars opercularis respectively the pars triangularis of the
frontal lobe, while 3 tumours showed a more parietal
Table 1. Demographic data of 14 patients with tumours in the opercular
region
Variables Total
Age (years) 31.1 � 10.1 (range 8–50)
Sex (m=f) 6=8
Histologic grade
II 11
III 3
Histologic subtype
fibrillary 10
oligoastrocytic 4
Symptoms (duration, months) 14.8 � 22.7 (range 1–82)
Epileptic seizures 13
Focal neurological deficits 5
(paresis, language deficits)
Signs of increased intracranial pressure 3
Tumour location
right (dominant in 2 pts.) 4
left 10
Surgery
under local anaesthesia 11
under generalized anaesthesia 3
GTR 7
STR 7
Gross total resection GTR; subtotal resection STR.
Surgical resection of gliomas WHO grade II and III located in the opercular region 11
extension. All tumours respected medially the basal
ganglia, the internal capsule and the fan of the motor
strip. Focal gadolinium enhancement was present in one
case.
Surgery, histology and adjuvant therapy
Open surgery with the aim of gross total resection was
performed in all cases. Gross total tumour removal
according to postoperative MR scans was achieved in
7 cases (50%). Based on the results of the intraoperative
language and motor monitoring tumour resection had to
remain subtotal in 5 cases and in two patients only an
extended open biopsy was possible since language and
motor function was widely distributed over the tumour.
Postoperative complications were recorded in 2 patients
(14.3%): one suffered an epileptic seizure, one devel-
oped meningitis. In addition, two patients showed an
allergic reaction against antiepileptic medication with
rash. No patient died.
Histological subclassification revealed in 10 tumours
a fibrillary and in 4 a mixed oligoastrocytic subtype.
Three tumours showed anaplastic features and were
classified as WHO grade III although the MR scans were
rather typical for a low grade astrocytoma WHO grade
II. On postoperative supportive therapy the three patients
with WHO grade III tumours received conventional
radiotherapy with a mean dose of 60.8� 2.4 Gy (range
59–64 Gy) and one patient an intracavitary radioimmu-
notherapy with radioactive labelled antibodies against
tenascin. In two patients radiotherapy was applied for
a recurrent tumour. Chemotherapy was used when
dedifferentiation occurred.
Cortical electrical stimulation
For intraoperative electrophysiological monitoring 12
to 25 sites were usually stimulated at least twice. It was
always possible to identify the precentral gyrus, which
was often either displaced or infiltrated by the tumour.
Stimulation of language relevant areas led to speech or
aphasic disturbance respectively arrest. In 7 out of 14
patients (50%), one or several positive motor and=or
language sites were found over the tumour.
Figure 2 illustrates three examples of tumours. In
Fig. 2a the location of the next positive motor or
speech areas are in more than 0.5 cm distance to the
tumour border, and there were no positive stimulation
sites over the tumour. The tumour was removed
completely without any motor or speech deficits post-
operatively.
Fig. 2. Three examples of opercular tumours. 2a. Tumour border with
a distance of more than 0.5 cm to the next positive motor or speech
points, rendering complete resection possible. No postoperative
neurological deficit occurred. 2b. Tumour involving the central region
with displacement of speech and motor relevant areas and some
positive stimulation points in a distance of less than 0.5 cm to the
tumour border. Surgery was aggravated by motor deficits of lip and
tongue during manipulation along the superior border which recovered.
Postoperatively the patient presented a slight clumsiness of speech,
which resolved within a few days. 2c. Positive stimulation points were
found almost all over the tumour and only an extended biopsy was
possible to prevent neurological deficits. No response; Motor
response; Speech disturbance; Aphasic disturbance; Speech
arrest; Aphasic arrest
12 A. Peraud et al.
Tab
le2
.C
linic
al
da
ta
Pat
.n
o.
Ag
e=se
xT
um
ou
rlo
cati
on
R=L
han
ded
nes
sR=L
Pre
oper
ativ
ed
efici
ts
neu
ropsy
cholo
gic
alre
sult
s
Po
sto
per
ativ
esy
mp
tom
san
dd
efici
tsn
euro
psy
cho
log
ical
resu
lts
Day
1–
81
–5
mon
ths
6–
18
mo
nth
s
1.
29=m
L=R
2G
rand
Mau
xw
ord
fin
din
gd
iffi
cult
ies
wo
rdfi
nd
ing
dif
ficu
ltie
s–
AB
,V
M:
no
defi
cits
AB
:n
ofo
rmal
test
ing
po
ssib
le:
sever
eap
has
ia
no
tte
sted
AB
,V
M:
no
defi
cits
VM
:n
ot
test
ed
2.
30=m
L=R
Jack
son
epil
epsy
rig
ht
leg
–o
ne
Jack
son
seiz
ure
un
der
stre
ss
AB
,V
M:
no
defi
cits
AB
:n
od
efici
tsA
B,
VM
:n
od
efici
tsn
ot
test
ed
VM
:im
pai
red
3.
41=f
R=R
inte
rmit
ten
tw
eak
nes
sle
ftar
m,
dy
sart
hri
a
par
esis
left
arm
,
spee
chap
rax
ia
spee
chap
rax
iasp
eech
apra
xia
un
der
stre
ss
4.
32=m
L=R
foca
lep
ilep
syfo
cal
epil
epsy
,n
od
efici
tsfo
cal
epil
epsy
,n
od
efici
tsfo
cal
epil
epsy
,n
od
efici
ts
AB
,V
M:
no
defi
cits
no
tte
sted
no
tte
sted
no
tte
sted
5.
8=f
L=R
Jack
son
epil
epsy
,G
ran
dm
al,
spee
chap
rax
ia
mil
dp
ares
iso
fle
ftth
um
b,
spee
chap
rax
ia
foca
lep
ilep
sy,
no
spee
chd
efici
tsfo
cal
epil
epsy
6.
36=f
L=R
foca
lan
dg
ener
aliz
edse
izu
res
––
–
AB
,V
M:
no
defi
cits
no
tte
sted
AB
,V
M:
no
defi
cits
AB
,V
M:
no
defi
cits
7.
41=f
L=R
gen
eral
ized
seiz
ure
sw
ith
po
stic
tal
aph
asia
hem
ipar
esis
rig
ht
side
pro
no
un
ced
on
arm
mil
dp
ares
iso
fri
gh
tar
m,
no
sen
sory
dis
crim
inat
ion
for
tem
per
ature
,sp
eech
apra
xia
un
der
stre
ss
no
sen
sory
dis
crim
inat
ion
for
tem
per
atu
re
AB
,V
M:
no
defi
cits
AB
:re
pet
itio
nm
ildly
imp
aire
d,
no
tte
sted
AB
,V
M:
no
defi
cits
VM
:n
od
efici
ts
8.
28=m
R=R
Gra
nd
mal
mil
dp
ares
iso
fle
fth
and
–se
izu
res
un
der
stre
ss
9.
28=f
L=R
foca
lse
izu
re,
mil
dh
emip
ares
is
rig
ht
side,
dy
sart
hri
csp
eech
,w
ord
fin
din
gd
iffi
cult
ies
pro
nou
nce
dh
emip
ares
is,
dy
sart
hri
csp
eech
resi
du
alh
emip
ares
isar
m
mo
reth
anle
g,
dy
sart
hri
c
spee
ch
resi
du
alh
emip
ares
is,
dy
sart
hri
csp
eech
un
der
stre
ss
AB
:re
pet
itio
nim
pai
red
AB
:re
pet
itio
n,
read
ing
,
nam
ing
imp
aire
d:
aph
asia
no
tte
sted
no
tte
sted
VM
:im
pai
red
VM
:n
od
efici
t
10
.2
2=f
L=R
gra
nd
mal
wit
hp
ost
icta
l
aph
asic
dis
turb
ance
mil
dp
ares
isri
gh
th
and
,
spee
chd
istu
rban
ce
epil
epti
cse
izu
re,
no
oth
erd
efici
t
AB
,V
M:
no
defi
cits
AB
:re
pet
itio
n,
nam
ing
imp
aire
d,
no
tte
sted
AB
,V
M:
no
defi
cits
VM
:n
od
efici
t
(co
nti
nued)
Surgical resection of gliomas WHO grade II and III located in the opercular region 13
Tab
le2
(co
nti
nued
)
Pat
.n
o.
Ag
e=se
xT
um
ou
rlo
cati
on
R=L
han
ded
nes
sR=L
Pre
oper
ativ
ed
efici
ts
neu
ropsy
cholo
gic
alre
sult
s
Po
sto
per
ativ
esy
mp
tom
san
dd
efici
tsn
euro
psy
cho
log
ical
resu
lts
Day
1–
81
–5
mo
nth
s6
–1
8m
on
ths
11
.3
0=f
R=L
foca
lan
dg
ener
aliz
edse
izu
res
––
–
AB
,V
M:
no
defi
cits
AB
:n
od
efici
tsn
ot
test
edn
ot
test
ed
VM
:n
ot
test
ed
12
.5
0=f
L=R
foca
lJa
ckso
nse
izu
res
spee
chap
rax
ia,
hem
ipar
esis
pro
no
un
ced
rig
ht
arm
,ep
ilep
tic
seiz
ure
s
mil
dsp
eech
apra
xia
mil
dsp
eech
apra
xia
AB
,V
M:
no
defi
cits
AB
:re
pet
itio
nm
ildly
dis
turb
ed,
AB
:re
pet
itio
nm
ildly
dis
turb
ed,
no
tte
sted
VM
:n
od
efici
tsV
M:
no
defi
cits
13
.3
8=m
R=L
foca
lse
izu
res
of
mo
uth
and
face
wit
hp
ost
icta
lap
has
ia,
faci
alp
ares
is
fin
em
oto
rd
efici
tle
fth
and
,
faci
alpar
esis
,ap
has
ia
hy
per
pat
hia
of
left
side,
aph
asia
un
der
stre
ssh
yp
erp
ath
iao
fle
ftsi
de
AB
:n
od
efici
ts,
VM
:>
2
SD
bel
ow
MV
AB
:T
ok
enT
est,
rep
etit
ion
and
wri
tten
lan
gu
age
mil
dly
imp
aire
d:
aph
asia
no
tte
sted
AB
,V
M:
no
defi
cits
VM
:n
od
efici
ts
14
.2
3=m
L=R
foca
lan
dg
ener
aliz
edse
izu
res
mil
dp
ares
isri
gh
th
and
,se
izu
res
foca
lse
izu
res,
hem
ihy
pae
sth
esia
foca
lse
izu
res,
hem
ihy
pae
sth
esia
AB
,V
M:
no
defi
cits
AB
,V
M:
no
defi
cits
no
tte
sted
no
tte
sted
SD
Sta
nd
ard
dev
iati
on
;M
Vm
ean
val
ue;
AB
aph
asia
bat
tery
;V
Mver
bal
mem
ory
.
14 A. Peraud et al.
In Figure 2b the tumour involved the inferior part of
the precentral and the postcentral gyrus, the speech rele-
vant areas were anterior to the tumour and the motor
areas were displaced to the superior tumour border. Sev-
eral of the latter sites were within a distance of <0.5 cm
to the tumour. During tumour removal along the superior
border weakness of lip and tongue movements occurred.
The surgeon waited 10 minutes until function recovered
and then carefully removed the tumour rest. There was a
slight clumsiness of speech, which resolved within a few
days.
Figure 2c demonstrates a tumour with many positive
motor and speech points over the surface of the tumour.
In this case only an enlarged biopsy was possible. No
postoperative deficits were evident.
Postoperative clinical course
The occurrence of postoperative motor and language
deficits is shown in Table 2. An impairment of speech or
language function immediately after surgery was present
in 8 patients, combined with motor deficits in 7 patients.
Of these 6 were neuropsychologically tested, and half of
them had problems with oral repetition of words and
sentences. Written language was not disturbed, and their
impairment was interpreted as apraxia of speech. In the
others several aspects of language processing were
impaired, indicating a true aphasic syndrome. The symp-
toms improved gradually over the first 5 months, leaving
2 patients with impaired language function. 4 patients
reported of speech problems under stress only. Two
patients still had motor deficits. At the latest follow-up
(6 to 18 months postoperatively) 3 patients had mild
problems of speech under stress. One patient still
showed a contralateral paresis.
Severity and duration of postoperative deficits were
correlated with the distance of the resection margin to
the next positive stimulation point(s). In 3 cases one of
several positive stimulation sites for language function
was located within the tumour margin and was resected
or undermined. Two of these patients exhibited already
intraoperatively and all 3 postoperatively impairment in
speech and the recovery lasted up to 11 months (Fig. 2c).
If the resection was carried out at a distance of <0.5 cm
to the next positive stimulation point, postoperative
speech deficits occurred in 5 of 6 patients and resolved
within 1–5 months. An example for such a case is illu-
strated in Fig. 2b. Resection at a distance >0.5 cm to the
next positive stimulation site never resulted in intra- or
postoperative deficits (5 patients). A typical example is
outlined in Fig. 2a.
A summary of the positive stimulation sites for motor
and language function of all patients is illustrated in
Fig. 3. Sites where aphasic arrest or disturbance were
elicited are widely distributed, but are in general located
anterior to the precentral sulcus. Speech arrest, with or
without motor involvement, was predominantly found at
the precentral gyrus.
Discussion
Function related to the opercular region
Knowledge of the complex functions of the opercular
cortex in the dominant hemisphere has derived from elec-
trical stimulation studies in patients undergoing opera-
tions for the treatment of medically intractable epilepsy
[14, 15, 12], and from functional imaging studies. Corti-
cal stimulation revealed small areas where orofacial
movements could be elicited, other areas were involved
in naming, reading, phoneme identification, sequencing in
sound production and short-term verbal memory perfor-
mance. There was also common cortex for language and
motor function. In the last years, a large number of func-
tional imaging studies have shown the involvement of
frontal-parietal networks in motor behaviour, language
processing and working memory [1]. Obviously, language
arises in part in predominant motor cortex, and these
cortical areas are located in the perisylvian region. A
recent clinical report describes a transient Foix-
Chavany-Marie syndrome, after resection of a low grade
insulo-opercular glioma in the non-dominant hemisphere,
Fig. 3. Summary of positive stimulation sites for motor and language
function of all patients. Stimulation points where aphasic arrest or
disturbance were elicited are generally distributed anterior to the
precentral sulcus, while those with speech arrest (with or without motor
involvement) are mainly found at the precentral gyrus and sulcus
Surgical resection of gliomas WHO grade II and III located in the opercular region 15
which resolved within 3 months. This syndrome consisted
of bilateral dysfunction of the face, pharynx, tongue, and
muscles of mastication, and the patient was unable to
speak [3]. The authors related these deficits to injury of
the corticosubcortical facial motor structures, and strong-
ly recommended intraoperative functional mapping and
awake surgery in order to identify these pathways.
In the present study with tumours involving the opercu-
lar region, the motor cortex respectively motor relevant
areas were rather constant and could be easily defined by
stimulation. This is helpful for orientation. In about 50%
of the cases, motor cortex was displaced by the tumour in
a superior or anterior direction, while in the other 50%
motor function was found in the cortex overlying the
tumour. Consequently, this cortex may still bear function.
In contrast to the motor cortex, the distribution of speech
relevant stimulation points was extremely variable. Only
the retrospective analysis (Fig. 3) demonstrated that areas
at which speech was arrested or disturbed (speech apraxia)
were located predominantly in the motor cortex and may
represent the above described common motor-language
cortex. Such areas may be responsible for the final cortical
motor pathway of speech. Areas where aphasic arrest
resulted from stimulation (the carrier phrase was uttered,
but naming was disturbed) were in general located more
anteriorly, and this corresponds more or less to the Broca
area (Fig. 3). Due to the relative large extension of the
two areas, any individual prediction of functional relevant
areas was impossible and the intraoperative localisation
was only possible with cortical stimulation.
Surgical conclusions
Despite the retrospective character of our study and
the possible bias in treatment strategies, our findings
may lead to some essential conclusions:
a) This spatial separation of two cortical areas with dif-
ferent language function may explain the two most
important postoperative speech deficits, which are
either apraxic or aphasic in nature. The pure aphasia
syndrome is related to areas with aphasic arrest, and
this deficit recovered fast in the present cases (Broca
area) (Fig. 3). Areas with speech arrest are mainly
located in the motor cortex and seem to share common
brain mechanisms, e.g. motor and language function.
Disturbances of such areas cause the so-called speech
apraxia syndrome. The recovery of such deficits was
more prolonged in the cases presented here.
b) Since such functional important sites are densely and
not predictably distributed over the opercular cortex,
surgery in this region becomes extremely delicate.
Initially, we assumed, that in analogy to astrocytomas
WHO grade II in the temporal lobe or supplementary
motor area [17], functional important sites would be
found outside the tumour, displaced during tumour
growth, and could thus be identified and spared. This
was the case for instance in patients represented in
Fig. 2a and b. However in approximately 50% of all
patients functional important sites were detected in
the cortex overlying the tumour. This is in contrast to
low grade astrocytomas of the superior frontal gyrus
or the temporal lobe, where positive stimulation sites
were found only in 4.3% and 5% of the patients over
the tumour, respectively [17, 18]. In addition, we
initially suggested that such sites are probably no
longer involved in the language network, but the def-
icits following resection or surgical undermining of
one or several sites in 3 patients demonstrated that
these sites still bear important language functions.
With these clinical experiences we became con-
vinced that such areas must be protected or even that
resection has to be abandoned if mapping showed
important function over the tumour. Consequently,
only 7 out of 14 patients were amenable for complete
tumour removal.
c) This study supports previous reports [8, 17], and
demonstrates that a safety margin of at least 0.5 cm,
according to Haglund and coworkers even >1 cm,
from the next positive motor or language site is
necessary to avoid persistent functionally disabling
deficits. This simple paradigm will be of great use
for neurosurgeons who perform surgery within this or
other functional important cortical regions.
d) A critical discussion of the technique of cortical elec-
trical stimulation, although presently accepted as the
method of choice for speech mapping, seems appro-
priate. Bipolar stimulation, even at low intensities,
occasionally may induce a focal seizure in the oro-
facial or hand=arm muscles of short duration. Local
application of cold Ringer’s lactate on the cortex was
recommended to rapidly terminate them [19].
Recently, Neuloh and Schramm advocated the use
of monopolar stimulation to reduce this risk [13].
e) An important question is whether in the future func-
tional MR may replace intraoperative cortical stimu-
lation. By using neuronavigation, the preoperative
MR-data including localisation of specific functions
can be superimposed to the intraoperative situation.
However, the unsolved problem of brainshift carries a
non-negligible risk of mislocation. A comparison of
16 A. Peraud et al.
pre- and intraoperative language areas with the same
naming task, by PET technique and cortical stimula-
tion respectively, showed only a correlation of 84%
of the tested sites. The positive PET site was always
less delineated than the positive stimulation site [9].
At present, these facts make clear that intraoperative
stimulation techniques are still most accurate and can
not be replaced by any other tool.
On the other side, preoperative functional MR may help
to clarify the question, whether functional important
areas are overlying the tumour or are definitively outside
the tumour volume. If this is the case, an alternative
mode of treatment may be considered, and open tumour
resection may be reserved for those cases where func-
tional important areas lay clearly 0.5 cm or more outside
the tumour border.
References
1. Cabeza R, Nyberg L (2000) Imaging cognition II: An empirical
review of 275 PET and fMRI studies. J Cogn Neurosci 12: 1–47
2. Delis DC, Kramer JH, Kaplan E, Ober BA (eds) (1987) The
California verbal learning test. Psychology Corporation, New York
3. Duffau H, Karachi C, Gatignol P, Capelle L (2003) Transient
Foix-Chavany-Marie syndrome after surgical resection of a right
insulo-opercular low-grade glioma: case report. Neurosurgery 53:
426–431
4. Ebeling U, Kothbauer K (1995) Circumscribed low grade astro-
cytomas in the dominant opercular and insular region: a pilot study.
Acta Neurochir (Wien) 132: 66–74
5. Ebeling U, Reulen HJ (1995) Space-occupying lesions of the
sensori-motor region. In: Symon L et al (eds) Advances and
technical standards in neurosurgery. Springer, Wien New York,
pp 138–176
6. Ebeling U, Steinmetz H, Huang Y, Kahn T (1989) Topography and
identification of the inferior precentral sulcus in MR imaging. AJR
Am J Roentgenol 153: 1051–1056
7. Eisner W, Reulen HJ, Ilmberger J, Swozil U, Bise K (1999)
Intraoperative mapping of eloquent brain areas. Front Radiat Ther
Oncol 33: 28–36
8. Haglund MM, Berger MS, Shamseldin M, Lettich E, Ojemann GA
(1994) Cortical localization of temporal lobe language sites in
patients with gliomas. Neurosurgery 34: 567–576
9. Herholz K, Reulen HJ, von Stockhausen HM, Thiel A, Ilmberger J,
Kessler J, Eisner W, Yousry TA, Heiss WD (1997) Preoperative
activation and intraoperative stimulation of language-related areas
in patients with glioma. Neurosurgery 41: 1253–1260
10. Huber W, Poeck K, Willmes K (1984) The Aachen Aphasia Test.
Adv Neurol 42: 291–303
11. Kelly PJ, Daumas-Duport C, Kispert DB, Kall BA, Scheithauer
BW, Illig JJ (1987) Imaging-based stereotaxic serial biopsies in
untreated intracranial glial neoplasms. J Neurosurg 66: 865–874
12. Lesser RP, Lueders H, Dinner DS, Hahn J, Cohen L (1984) The
location of speech and writing functions in the frontal language
area. Brain 107: 275–291
13. Neuloh G, Schramm J (2002) Intraoperative neurophysiological
mapping and monitoring. In: Deletis V, Schils JL (eds) Neurophy-
siology in neurosurgery. Academic Press, San Diego, pp 339–402
14. Ojemann GA (1982) Interrelationships in the localization of
language, memory, and motor mechanisms in human cortex and
thalamus. In: Thompson RA et al (eds) New perspectives in cerebral
localization. Raven Press, New York, pp 157–175
15. Ojemann GA, Mateer C (1979) Human language cortex: Localiza-
tion of memory, syntax, and sequential motor-phoneme identifica-
tion systems. Science 205: 1401–1403
16. Ojemann GA, Ward AA (1971) Speech representation in ventro-
lateral thalamus. Brain 94: 669–680
17. Peraud A, Meschede M, Eisner W, Ilmberger J, Reulen HJ (2002)
Surgical resection of grade II astrocytomas in the superior frontal
gyrus. Neurosurgery 50: 966–977
18. Reulen HJ, Schmid UD, Ilmberger J, Eisner W, Bise K
(1997) Tumor surgery of the speech cortex in local anesthesia.
Neuropsychological and neurophysiological monitoring during
operations in the dominant hemisphere. Nervenarzt 68:
813–824
19. Sartorius CJ, Berger MS (1998) Rapid termination of intraoperative
stimulation-evoked seizures with application of cold Ringer’s
lactate to the cortex. Technical note. J Neurosurg 88: 349–351
20. Steinmetz H, Ebeling U, Huang Y, Kahn T (1990) Sulcus topo-
graphy of the parietal opercular region: an anatomic and MR study.
Brain Lang 38: 515–533
Comments
Microsurgical removal of tumours located in the opercular region
is certainly a major neurosurgical challenge, mainly because of the
high risk of speech apraxia, aphasic syndromes, motor impairments.
This investigation is based on 14 consecutive patients, operated upon
for opercular gliomas (11 astrocytomas grade II, 3 with anaplastic
features). The relevance of postoperative deficits was analysed accord-
ing to three major parameters: a) peculiar location of the tumour
b) intraoperative neuromonitoring pictures; c) extent of surgical
resection.
The study confirmed the extreme anatomical variability typical of this
region, as formerly described by Ebeling and Steinmetz (1): within such
a relatively small group of patients four different anatomical patterns
were recognized. Furthermore, results seem to stress the fundamental
role of cortical electric stimulation to avoid permanent surgical morbid-
ity. Indeed the severity and duration of postoperative sequelae was
consistently related to the distance between the resection margin and
the next positive stimulation point(s), with a safety cut-off distance of
more than 0.5 cm. Finally the distribution of critical stimulation points
on the tumour boundaries was unpredictable, with up to 50% of these
sites overlying the tumour.
M. Gerosa
This article gives important information on the reliability of
intraoperative neuromonitoring as an aid to surgery on eloquent cortex.
The authors objectively presented their data and experience for dominant
side opercular region tumours and speculated on the failures in terms of
intraoperative stimulation data and postoperative language deficits.
The conclusions are sound and reflect the insight of a highly experi-
enced team. The only point to criticize is the difficulty in reading the
article. One has to correlate all anatomical, electrophysiological and
surgical facts together with the figures, which require quite a concentra-
tion. Still, I believe, the authors did their best by carrying the structured
information and descriptions to the method and results section and
keeping the discussion concise.
N. Akalan
This is a retrospective analysis of a cohort of patients with gliomas of
the opercular region with respect to intraoperative and postoperative
Surgical resection of gliomas WHO grade II and III located in the opercular region 17
functional deficits and their correlation with tumour location and extent
of resection. The scientific approach is valid and its minor shortcomings
are adequately discussed. The ‘Discussion’ briefly reviews the anatomy
of the opercular cortex and outlines the implications of this study’s
findings on surgical technique.
Although similar studies have been published in the past, the
article stands out due to its systematic approach and the excellent
documentation of tumour location, intraoperative stimulation and func-
tional consequences of stimulation and surgical resection.
J. M. Baehring
Correspondence: Dr. Aurelia Peraud, Klinikum Grosshadern, Department
of Neurosurgery, Marchioninistrasse 15, 81377 Munich, Germany. e-mails:
[email protected], [email protected]
18 A. Peraud et al.: Surgical resection of gliomas WHO grade II and III located in the opercular region