Surgery of Insular Nonenhancing Gliomas:

Volumetric Analysis of Tumoral Resection, Clinical

Outcome, and Survival in a Consecutive Series of
66 Cases
BACKGROUND: Despite intraoperative technical improvements, the insula remains
a challenging area for surgery because of its critical relationships with vascular and
neurophysiological functional structures.
OBJECTIVE: To retrospectively investigate the morbidity profile in insular nonenhancing
gliomas, with special emphasis on volumetric analysis of tumoral resection.
METHODS: From 2000 to 2010, 66 patients underwent surgery. All surgical procedures
were conducted under cortical-subcortical stimulation and neurophysiological monitor-
ing. Volumetric scan analysis was applied on T2-weighted magnetic resonance images
(MRIs) to establish preoperative and postoperative tumoral volume.
RESULTS: The median preoperative tumor volume was 108 cm
3
. The median extent of
resection was 80%. The median follow-up was 4.3 years. An immediate postoperative
worsening was detected in 33.4% of cases; a definitive worsening resulted in 6% of cases.
Patients with extent of resection of .90% had an estimated 5-year overall survival rate of
92%, whereas those with extent of resection between 70% and 90% had a 5-year overall
survival rate of 82% (P , .001). The difference between preoperative tumoral volumes on
T2-weighted MRI and on postcontrast T1-weighted MRI ([T2 2 T1] MRI volume) was
computed to evaluate the role of the diffusive tumoral growing pattern on overall survival.
Patients with preoperative volumetric difference , 30 cm
3
demonstrated a 5-year overall
survival rate of 92%, whereas those with a difference of . 30 cm
3
had a 5-year overall
survival rate of 57% (P = .02).
CONCLUSION: With intraoperative cortico-subcortical mapping and neurophysiologi-
cal monitoring, a major resection is possible with an acceptable risk and a significant
result in the follow-up.
KEY WORDS: Brain mapping, Direct electrical stimulation, Extent of resection, Functional outcome, Insular
gliomas surgery
Neurosurgery 70:10811094, 2012 DOI: 10.1227/NEU.0b013e31823f5be5 www.neurosurgery-online.com
I
nsular gliomas, accounting for up to 25% of
low-grade gliomas (LGGs) and up to 10% of
high-grade gliomas, have a clear propensity
for the insular lobe.
1
Their tendency to spread
along the intricate network of afferent and
efferent connections between the insula itself
and cortical structures
2
constitutes the major
difficulty in reaching a gross total resection. In
fact, the insular lobe represents the anatomo-
functional interface between the limbic system
and the neocortex.
2-4
Because of the functional
role and complex vascular anatomy of the insula,
insular gliomas have been considered unresect-
able for a long time.
5-12
Functionally, the insular
cortex has recently been defined as an associative
multimodal compensable
13
secondary area. More-
over, functional neuroimaging techniques have
revealed the insular cortex to be involved in
many different functions such as language
14
and
Miran Skrap, MD*
Massimo Mondani, MD*
Barbara Tomasino, PhD
Luca Weis, PhD
Riccardo Budai, MDk
Giada Pauletto, MD, PhDk
Roberto Eleopra, MDk
Luciano Fadiga, MD, PhD
Tamara Ius, MD*
Department of *Neurosurgery, and
kNeurology, Azienda Ospedaliero
Universitaria Santa Maria della Misericor-
dia, Udine, Italy; IRCCS E. Medea, Polo
Regionale del FVG, San Vito al Taglia-
mento, Italy; Department of RBCS,
Italian Institute of Technology, Genova,
Italy; Section of Human Physiology,
University of Ferrara, Ferrara, Italy
Correspondence:
Tamara Ius, MD,
Department of Neurosurgery,
Azienda OspedalieroUniversitaria
Santa Maria della Misericordia,
Piazzale Santa Maria della Misericordia
15, 33100, Udine, Italy.
E-mail: tamara.ius@gmail.com
Received, February 28, 2011.
Accepted, August 14, 2011.
Published Online, November 3, 2011.
Copyright 2011 by the
Congress of Neurological Surgeons
ABBREVIATIONS: CI, confidence interval; DTI,
diffusion tensor imaging; EOR, extent of resection;
IES, intraoperative electric stimulation; IFOF, in-
ferior fronto-occipital fasciculus; LGG, low-grade
glioma; MEP, motor evoked potential; OS, overall
survival; WHO, World Health Organization
RESEARCHHUMANCLINICAL STUDIES
TOPIC RESEARCHHUMANCLINICAL STUDIES
NEUROSURGERY VOLUME 70 | NUMBER 5 | MAY 2012 | 1081
Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
sensorimotor integration; cognitive-emotional processing
14-17
;
gustatory, auditory, and vestibular functions
3,13,18
; and neuro-
psychiatric disturbances.
19-21
Anatomically, the insula lies within
the depth of the sylvian fissure, covered by frontal, parietal, and
temporal opercula and overlying the deep basal nuclei.
12,22-25
It is
covered by the trunk of the middle cerebral artery and its
branches, and it has important relationships with the perforating
lateral lenticulostriate arteries.
7,12,25-29
Furthermore, the insula is
close to fibers: the pyramidal tract in the superior posterior part of
the insula, the inferior fronto-occipital fasciculus (IFOF) that
runs medially to the insula, and the arcuate fibers that runs more
superiorly and could be very close to insular tumors. This intri-
cate network has been visualized by the introduction of diffusion
tensor imaging (DTI), providing new functional-anatomic land-
marks in glioma surgery.
2,30,31
Recently, as a result of technical
developments and a better understanding of insular functional
anatomy, some publications have focused on the key role played
by surgery in insular glioma management,
5,8,9,11
highlighting
that an extensive resection performed at the time of the initial
diagnosis constitutes the major favorable prognostic factor to
improve patients overall survival (OS).
In the present retrospective study, we illustrate our experience
with insular nonenhancing glioma surgery, concentrating on the
functional aspects of this area with special emphasis on the role of
the extent of resection on OS, as evidenced by a volumetric analysis
performed on the tumoral mass.
METHODS
Patient Population
We retrospectively analyzed a series of 66 patients with insular
nonenhancing gliomas treated in 71 resections performed by the senior
author (M.S.). These patients underwent surgery in our institute between
January 2000 and July 2010. Preoperatively, all patients underwent
neuropsychological evaluation and functional magnetic resonance imag-
ing (MRI) studies to establish the dominant side and the locations of
language-related functions. In addition, language handedness dominance
was evaluated by means of the Edinburgh Inventory Questionnaire.
32
Forty-seven patients had a lesion involving the dominant hemisphere.
No patient had contrast enhancement on postcontrast T1-weighted MRI
sequences. We classified the lesions considering both the Yasargil classi-
fication system
12
and the Berger-Sanai insular glioma classification
system.
9
Histological type was determined according to the World
Health Organization (WHO) brain tumor classification.
33
This study
was approved by the local institution ethics committee on human
research.
Surgical Technique and Mapping
All surgical procedures were conducted under intraoperative cortical
and subcortical electric stimulation (IES), following the intraoperative
methodology previously described by Berger and colleagues
34-36
and with
the routine use of neurophysiological monitoring of motor evoked
potentials (MEPs) and somatosensory evoked potentials. In all cases, we
used a NeuroNavigation System (Stealth Station, Medtronic). The
selection of the anesthesiological protocol was based on the preoperative
evaluation of hemispheric dominance. Awake craniotomy (first surgery
in 43 cases and second surgery in 4 cases) was performed for lesions
involving the dominant hemisphere. We used a standard anesthesio-
logical protocol for the awake surgical procedure; all patients were awake
from the beginning of the procedure with a slight sedation during skull
opening and closing. No laryngeal mask was used. On the contrary,
patients with gliomas in the nondominant hemisphere (first surgery in
23 cases and second surgery in 1 case; total of 24) underwent surgery
under general anesthesia with cortical and subcortical IES and neuro-
physiological monitoring to detect the corticospinal pathways. Electro-
encephalographic and electrocorticographic recordings were always used
to monitor the occurrence of afterdischarge phenomena, both electric
and clinical. Somatosensory evoked potential and MEP recordings were
also performed to continuously monitor motor and somatosensory
potentials (Eclipse Neurovascular Workstation-Axon; Video Polygraphic
Station System Plus-Micromed). Patients were positioned supine with
the head fixed in a Mayfield frame and turned at an angle of 30 toward
the floor. Immediately after dura opening and before cortical mapping,
an 8-electrode strip was placed across the central sulcus and two
4-electrode strips were placed over the frontal and temporal lobes
(1 each) to record the electrocortical activity. Before the resection was
started, with the help of the navigator and the bipolar IES, we put some
methylene bluecolored small markers in the depth along the medial
border of the temporal and frontal extension of the lesion to have a better
anatomic orientation during the most advanced phases of surgery, when
brain shifting occurs.
At the beginning, we used a transsylvian approach, but for several years
now, on the basis of data provided by cortical mapping, we have preferred
to use transcortical transfrontal and/or transtemporal approaches. This
technique allowed us to obtain more operative space, which made it easier
to dissect the medial cerebral artery and its branches subpially. We prefer-
red to do this at the end of the procedure. By increasing the space around
the vessels, we decreased the risk of evoking pain while manipulating the
middle cerebral artery in an early phase of surgery. To reduce the risk of
ischemia, we now prefer to leave a minimal part of the tumor around the
lenticulostriate arteries in case they are encased.
Performing cortical and subcortical brain mapping, we rarely needed
more than 4 mA of current intensity for the cortex; we usually started
with 6 or 8 mA for subcortical simulation. Proceeding to the depth, we
alternated resection of thin tumoral layers with subcortical stimulation
(to detect the pyramidal tract) particularly in the posterior part of
the lesion and subcortical pathways of language (arcuate fasciculus and
IFOF) for tumors harboring on the dominant hemisphere. The medial
border of the resection was anatomically represented by the lenticulo-
striate arteries
7,8,11,12
and functionally by the detection of internal
capsule and the pyramidal pathways.
5
In the present study, we analyzed a patient population operated on in
the last 10 years, during which time the intraoperative technical protocol
was changed. Therefore, 2 consecutive periods were identified to evaluate
the impact of each protocol on clinical outcome and extent of resection
(EOR):
Series 1 (from January 2000 to December 2004): 25 patients were
operated on with the aid of the cortico-subcortical IES, neurophys-
iological monitoring, and intraoperative use of a NeuroNavigation
System. Fifteen patients had a lesion in the nondominant hemisphere
and were treated under general anesthesia. In the remaining 10
patients, the lesion was located in the dominant hemisphere; con-
sequently, awake surgery was performed.
Series 2 (from January 2005 to July 2010): 41 patients were operated
on with the aid of the cortico-subcortical IES, neurophysiological
SKRAP ET AL
1082 | VOLUME 70 | NUMBER 5 | MAY 2012 www.neurosurgery-online.com
Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
monitoring, and intraoperative use of a NeuroNavigation System (as
done for series 1). In this second group, we added the overlapping of
functional MRI/DTI data on the T1/T2 3-dimensional MRIs to
better define the surgical plan. Eight patients had a lesion in the
nondominant hemisphere and were treated under general anesthesia.
In the remaining 33 patients, the lesion was in the dominant
hemisphere; consequently, awake surgery was performed (Figure 1).
Only in the more recent cases (12 patients) has an enriched testing for
language functions been performed continuously during the entire
duration of the surgical procedure, providing complementary information
to that provided by IES. The following tasks were used and were con-
tinuously repeated throughout the resection: counting, object picture
naming, action picture naming, word comprehension and repetition,
sentence comprehension, spontaneous speech, and digit span test.
Patient Outcome Measurements
Neurological examinations were performed preoperatively and 1 week,
3 months, and 6 months after surgery. Postoperative neurological deficits
were graded as mild (minimal, unnoticeable deficit), moderate (deficit
interfering with functions but potentially reversible), or severe (deficit that
significantly disables functions). Patients with no clinical improvement
at the 6-month follow-up examination were considered to have a per-
manent deficit. In addition to neurological morbidity, the OS was
evaluated in relation to the EOR and histopathological results.
Volumetric Analysis
Volumetric analysis was applied to establish the extent of tumor
resection. MRIs in DICOM (digital imaging and communications in
medicine) format were used to compute volumetric analyses of both pre-
operative and postoperative tumoral volume by using axial T2-weighted
MRIs. For the postoperative volume reconstructions, we used images
acquired 4 months after surgery. All preoperative and postoperative
tumoral segmentations were performed manually with the OSIRIX
software tool.
37
The EOR was calculated as follows, as previously
described by Smith et al
38
: (preoperative tumor volume 2 postoperative
tumor volume)/preoperative tumor volume.
Statistical Analysis
Commercially available software was used for statistical analysis (SAS
version 9.2; SAS Institute Inc, Cary, North Carolina). Univariate and
multivariate analyses were performed to study the influence of different
variables on patient survival. The cases of second surgery were excluded
from statistical analysis. Survival probabilities were estimated with the
Kaplan-Meier method. The EOR was classified as , 70%, 70% to 90%,
and . 90% to ensure consistency between this study and previous
studies that focused on the impact of glioma resections in terms of
volume.
9,38,39
The statistical significance of survival differences among
subgroups of subjects was assessed through the log-rank test. The
difference between preoperative tumoral volumes on T2- weighted MRI
and on postcontrast T1-weighted MRI ([T2 2 T1] MRI volume) was
computed to evaluate the role of a diffusive tumoral pattern of insular
nonenhancing gliomas on OS (Figure 2).
A multivariate Cox proportional hazards model was used to define
whether the EOR (treated as a continuous variable) predicted OS, after
adjustment for the effects of age, Karnofsky Performance Status,
preoperative tumoral volume, histology, and difference between pre-
operative tumoral volumes on T2- weighted MRI and on postcontrast
T1-weighted MRI ([T2 2 T1] MRI volume). Results are presented as
hazard ratios (HRs) and 95% confidence intervals (95% CIs). The
Student t test was used to calculate P values for comparisons of means.
All P values were obtained from 2-tailed tests (long-rank test, t test,
x
2
test), with statistical significance defined as P , .05
RESULTS
Patient Population
Patients demographic, preoperative clinical, and radiological
data are presented in Table 1.
The present study included 40 male and 26 female patients.
According to the classification proposed by Yasargil et al,
12
2
gliomas involved the sole insular lobe (type 3A), 8 gliomas
involved both the insula and at least 1 adjacent operculum (type
3B), and 46 gliomas involved either or both other paralimbic-
fronto-orbital, temporopolar areas (type 5A, 37 cases) and/or
parts of the limbic system (type 5B, 9 cases). Type 5 gliomas
(74%), occupying all zones identified by the Berger-Sanai Insular
Glioma Classification System,
9
can also be classified as giant.
Presenting symptoms included seizures in 64 cases and intra-
cranial hypertension in 2 cases. Pharmacologically resistant
epilepsy was present in 15 patients (23%). Preoperative neurolog-
ical examination was normal in 62 cases. None of the patients of
this series had discovered tumors incidentally. None of the patients
underwent preoperative chemotherapy or radiotherapy. Sixty-six
patients underwent a primary craniotomy (43 awake procedures
and 23 procedures with general anesthesia). The median time
between the first and the second surgical procedures was 39.8
months (range, 18-77 months). In all cases, preoperative MRIs
showed a lesion that was hypointense on a T1-weighted MRI
sequence without gadolinium-enhanced contrast and hyperintense
on a T2-weighted MRI sequence. The mean preoperative tumoral
volume, computed on T2-weighted MRI, was 108 6 47.9 cm
3
.
Postoperative Course and Histological Results
The clinical, histological, radiological, and follow-up data are
summarized in Table 2. In the immediate postoperative phase, a
worsening of neurological status was observed in 22 patients
(33.4%) as follows: Motor deficits developed in 11 cases (16.7%;
hemiplegia in 2 patients and moderate hemiparesis in 9 patients),
and speech disorders occurred in 11 patients (16.7%; articulatory
disorders in 1 patient, phonemic paraphasia without comprehen-
sion deficit in 8 patients, speech disorders with comprehension
deficit in 2 patients). At the 3-month follow-up examination, the
neurological conditions of all but 2 patients improved and
returned to the initial level or better. In these 2 patients, per-
manent hemiplegia (3%) was associated with permanent speech
disorders (3%). All permanent deficits occurred in the series
1 patients (compared with those belonging to series 2, x
2
= 4.5;
P = .03). Moreover, the 3 patients with preoperative neurological
deficits completely recovered 3 months after surgery. In 10 of
16 patients with postoperative motor deficits, we noticed a sudden
intraoperative reversible reduction in MEP recordings, whereas
INSULAR NONENHANCING GLIOMA SURGERY
NEUROSURGERY VOLUME 70 | NUMBER 5 | MAY 2012 | 1083
Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
there was no change in 4 patients. Irreversible MEPs loss occurred
in 2 patients who developed permanent motor deficit. Intra-
operative clinical seizures occurred in 3 patients, whereas electric
seizures were recorded in 8 patients. Six of 47 patients operated
on under local anesthesia developed intraoperative neurovegeta-
tive syndrome characterized by bradycardia, sickness, and
epigastric aura. Furthermore, there was no intraoperative or peri-
operative death related to surgical procedures. Histological diag-
noses were as follows: 2 astrocytomas with gemistocytic foci
(WHO II), 34 fibrillar astrocytomas (WHO II), 10 oligoastro-
cytomas (WHO II), 7 oligodendrogliomas (WHO II), 9 ana-
plastic astrocytomas (WHO III), 2 anaplastic oligoastrocytomas
(WHO III), and 2 anaplastic oligodendrogliomas (WHO III).
Adjunctive postoperative treatment was administrated in 42
cases. Immediate postoperative radiotherapy was performed in 13
patients with WHO III gliomas, whereas the other 29 patients
received adjunctive treatment because of radiological signs of
tumoral progression: chemotherapy alone in 4 patients, radio-
therapy in 2 patients, and chemotherapy combined with
radiotherapy in 23 patients.
Extent of Resection
The median tumoral residual volume, computed on post-
operative T2-weighted MRI, was 13 cm
3
(range, 0-112 cm
3
).
Removal of . 90% of the preoperative tumoral volume was
achieved in 33% of cases (22 patients). Surgical removal between
70% and 90% of the preoperative tumoral volume was achieved
in a further 45% of cases (30 patients). Partial resection (, 70%
of the preoperative tumoral volume) was performed in 22% of
cases (14 patients). Seven patients (11%) had an EOR between
0% and 59%. Seventeen (22%) were within the 60% to 79%
EOR range, whereas 20 patients (30%) were between 80% and
89%, and 22 patients (33%) had $ 90% EOR. The correlation
between preoperative tumoral volume and postoperative residual
volume was not significant (R
2
linear = 0.235). In particular, the
median extent of tumoral volume resection was 80% (range,
28%-100%). In 42 patients (65% of cases), the EOR was .
85%. A further volumetric analysis was performed by separating
patients operated on with the 2 different intraoperative protocols
(series 1 vs series 2). The series 1 patients demonstrated a mean
FIGURE 1. A, functional MRI linguistic map (P , .05, familywise error cor-
rected at the cluster level) in a left giant frontotemporoinsular nonenhancing glioma.
The functional analysis highlights the cortical areas activated by counting (a), object
naming (b), and verb generation tasks (c). The eloquent cortical regions, displaced by
the tumor, were easily detected with cortical mapping during the surgical procedure.
More critical is the detection of subcortical functional pathways. B, 3-dimensional
diffusion tensor imaging (DTI) fiber tracking reconstruction shows the corticospinal
tract (CST) (blue regions of interest [ROIs]), arcuate fasciculus (AF; orange ROIs),
(IFOF; green ROIs); ACHIEVA 3T, Philips, Netherlands). C, The tumor mass
displaces the AF upward, the IFOF medially downward, and the corticospinal tract
medially and posteriorly. Functional MRI/DTI data were overlapped on T2-
weighted MRIs and loaded into the neuronavigation system, allowing a better
evaluation of the preoperative surgical planning.
SKRAP ET AL
1084 | VOLUME 70 | NUMBER 5 | MAY 2012 www.neurosurgery-online.com
Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
EOR of 77% (range, 54% to 97%), whereas the series 2 patients
had a mean EOR of 83% (range, 28% to 100%; test statistic t for
mean percent EOR comparison [df = 64] = 1.8%; P = .001).
Finally, a last volumetric analysis was computed by using the
preoperative difference (T2 2 T1) MRI volume. The study
population was divided into 2 subgroups (subgroups A and B).
Our results evidenced that in cases with preoperative (T2 2 T1)
MRI volume ,30 cm
3
(subgroup A = 39 cases), the mean EOR
was 88.5%, whereas when this difference was . 30 cm
3
(subgroup B = 28 cases), the mean EOR was 68.4%. Subgroup
A demonstrated a mean EOR of 88.5%, whereas subgroup B had
a mean EOR of 68.4% (t test statistics for mean percent EOR
comparison [df = 64] = 6.8%; P = .000). Consequently, our data
pointed out that the mean EOR was , 70% when the
preoperative difference of preoperative tumoral volume com-
puted on T2 and T1 MRI was . 30 cm
3
.
Overall Survival
Overall, there were 14 deaths (15.4%) and the median follow-
up in the surviving patients was 4.3 years (range 3-127 months).
Death occurred in 6 patients (9%) with WHO II gliomas, and
8 (12%) with WHO III gliomas. A Kaplan-Meier estimate was
used to calculate survival rates. The 5-years OS was 80%. Patients
with WHO II gliomas, compared with those with WHO III
gliomas, demonstrated a significantly improved OS (HR, 0.075;
95% CI, 0.091-0.0819; P = .002). Patients with a histological
diagnosis of oligodendroglioma and oligoastrocytoma had an
estimated 5-year (60-month) OS rate of 92%, whereas those with
a diagnosis of astrocytoma had a 5-year OS rate of 72%. Finally,
patients with a histological diagnosis of WHO III gliomas had
a 5-year OS rate of 58% (Figure 3). For graphic visualization
purposes, survival was stratified according to 3 categories: EOR
, 70%, EOR 70% to 90%, and EOR . 90%. Using this
stratification scheme in a Kaplan-Meier curve, we observed
a stepwise improvement in OS associated with an increasing
EOR among patients with LGGs. Patients with $ 90% EOR
had an estimated 5-year OS rate of 92%; those with EOR
FIGURE 2. A case of a right frontotemporoinsular fibrillary astro-
cytoma. A, preoperative axial and coronal T2-weighted MRI slices of
right frontotemporoinsular low-grade glioma. B, preoperative
tumoral volume computed on postcontrast T1-weighted MRI was
122 cm
3
(axial slices). C, overlap on preoperative T2-weighted MRI
sequences of the tumoral region of interest defined on the postcontrast
T1-weighted MRI (red) and the T2-weighted images (green). The
preoperative tumoral volume computed on T2-weighted MRI se-
quence was 130 cm
3
(axial slices). The tumoral volume was computed
with OSIRIX software. D, volumetric analysis of postoperative
tumoral residue on postcontrast axial T1-weighted MRI evidenced
a tumoral residual volume of 4 cm
3
(axial slices). E, volumetric
analysis of postoperative tumoral residue computed on T2-weighted
MRI showed a tumoral residual volume of 6 cm
3
(axial slices). The
extent of the tumoral volume resection, computed on a T2-weighted
MRI sequences, was 95.4%.
INSULAR NONENHANCING GLIOMA SURGERY
NEUROSURGERY VOLUME 70 | NUMBER 5 | MAY 2012 | 1085
Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
between 70% and 90% had a 5-year OS rate of 82%; and those
with EOR , 70% had a 5-year OS rate of 57% (HR, 0.885;
95% CI, 0.827-0.947; P = .001; Figure 4A). Moreover, patients
with postoperative tumoral volume , 20 cm
3
, compared with
those with postoperative tumoral volume . 20 cm
3
, demon-
strated a significantly improved OS (P = .03). We identified
4 subgroups relative to residual volume: ,10 cm
3
, 10 to 20 cm
3
,
20 to 35 cm
3
, and . 35 cm
3
. The 5-year OS rates were 90%,
87%, 65%, and 28%, respectively (HR, 0.944; 95% CI, 0.900-
0.989; P = .02; Figure 4B). Finally, the difference between
tumoral volume computed on T2-weighted images and post-
contrast T1-weighted MRI was assessed (Figure 5). Patients with
preoperative (T2 2T1) volume ,30 cm
3
demonstrated a 5-year
OS rate of 92%, whereas those with preoperative (T2 2 T1)
volume .30 cm
3
had a 5-year OS rate of 57% (HR, 1.049; 95%
CI, 1.008-1.092; P = .02).
Multivariate Cox regression analysis (variables: age, Karnofsky
Performance Status, preoperative tumoral value, histological type,
preoperative (T2 2 T1) MRI volume, and EOR) revealed
preoperative tumoral volume, histological type, preoperative
(T2 2 T1) MRI volume, and EOR as independent predictors
of OS. After controlling for age, Karnofsky Performance Status,
preoperative tumoral value, histological type, and preoperative
(T2 2 T1) MRI volume, the EOR remained the strongest
independent significant predictor of OS (HR, 0.919; 95% CI,
0.871-0.970; P = .002; Table 3).
DISCUSSION
The incidence of insular gliomas represents 25% of all LGGs.
1
The infiltrative growing pattern in functional areas, particularly
for tumors in the dominant side, and the close relationship with
TABLE 1. Summary of Preoperative Clinical and Radiological
Features in 66 Patients With Insular Nonenhancing Gliomas
Parameter
Age at diagnosis, y
Median 40
Range 19-68
Sex, n
M 40
F 26
Glioma site, n
Right insula 22
Left insula 44
Preoperative tumoral volume computed
on T2-weighted images, cm
3
Median 108
Range 6-250
Symptoms at presentation, n
Seizures of recent onset 64
Generalized 44
Partial 20
Intracranial hypertension 2
Pharmacologically resistant epilepsy 15
Preoperative Karnofsky Performance Status score
Median 95
Range 80-100
Preoperative neurological examination, n
Normal 62
Mild language disorders 3
Mild hemiparesis 1
Tumoral location according to the Yasargil
classification,
12
n
Type 3A 2
Type 3B 8
Type 5A 37
Type 5B 9
TABLE 2. Summary of Postoperative Clinical, Radiological Results,
and Histological Data
a
Parameter
Immediate postoperative clinical findings, n (%)
Normal 44 (66.6)
Neurological deficits 22 (33.4)
Motor deficits 11 (16.7)
Speech disorders 11 (16.7)
Clinical outcome at 6 mo, n (%)
Normal 62 (94)
Motor deficit 2 (3)
Speech disorders 2 (3)
WHO tumor grade, n
Astrocytomas with gemistocytic foci (WHO II) 2
Fibrillar astrocytomas (WHO II) 34
Oligoastrocytomas (WHO II) 10
Oligodendrogliomas (WHO II) 7
Anaplastic astrocytomas (WHO III) 9
Anaplastic oligoastrocytomas (WHO III) 2
Anaplastic oligodendrogliomas (WHO III) 2
Extent of resection
b
Median residual volume, cm
3
13
Range, cm
3
0-112
EOR . 90%, n 22
EOR 70%-90%, n 30
EOR , 70%, n 14
Additional treatment, n
None 24
Second surgical procedure 5
Chemotherapy 4
Radiotherapy 15
Chemotherapy plus radiotherapy 23
Clinical follow-up
c
Mean follow-up, mo 52 (4.3 y)
Range, mo 3-127
Deaths, n (%) 14 (21)
WHO II, n (%) 6 (9)
WHO III, n (%) 8 (12)
a
EOR, extent of resection; WHO, World Health Organization.
b
Determined on the basis of preoperative and postoperative T2-weighted MRIs,
following the methodological procedure described by Smith et al.
38
c
Since the first operation.
SKRAP ET AL
1086 | VOLUME 70 | NUMBER 5 | MAY 2012 www.neurosurgery-online.com
Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
the vascular structures make surgery of low grade insular tumors
a challenge.
5,7-9,11,12,27,29,40,41
Furthermore, the spreading of the tumor along the complex
network of afferent and efferent fibers, involving cortical structures
and distinct subcortical pathways (arcuate fasciculus, uncinate
fasciculus, IFOF) underlying the insular lobe,
42
limits the
achievement of the oncological goal of total tumoral removal.
Despite the frequent involvement of the insula by gliomas,
1
very few cases have been reported in the literature
(Table 4).
5,7-9,11,12,40,41
Considering that these gliomas are
fated to a systematic anaplastic transformation over time,
42
more recent studies report that more extensive resections
constitute a major favorable prognostic factor to improve patients
OS.
5,9
In addition, computerized techniques to assess the tumoral
volume have allowed evaluation of a statistical correlation between
the role of EOR and OS.
38,39
The present study represents one of
the largest reported experiences on nonenhancing insular glioma
surgery. The aim of the present study is to provide further
statistical evidence that a more aggressive resection correlates to
a significant improvement in OS compared with a simple
debulking procedure as a result of to several technical intra-
operative developments.
Volumetric and Survival Analyses
Fewarticles on EORand follow-up after insular gliomas surgery
have been published. Despite the lack of Class I evidence, in the
last years, retrospective studies have confirmed the key role played
by surgery onOS andquality of life.
5,7-9,11,12,39,43-46
Simon et al
11
and Duffau
5
were the first to assess the Kaplan-Meier estimation
of survival in patients who underwent surgery for insular gliomas.
In particular, the first work reported a 5-year OS rate of 68% and
58% for LGGs and anaplastic gliomas, respectively, whereas the
second one showed a 5-year OS rate of 72% since the first surgery
for LGGs. Keles et al
43
published a review on the role of EOR in
LGGs, identifying 30 prospective studies. All but 5 were excluded
from further comparison because of significant methodological
limitations. Most notably, none of the above-mentioned studies
included a volumetric analysis, and 4 relied solely on the
surgeons intraoperative impressions of EOR. To overcome these
problems, a methodological procedure to volumetrically quantify
the EOR was introduced.
37,47
Besides that, in a subsequent series
focused on the EOR value on survival outcome in patients with
insular gliomas, Sanai et al
9
analyzed the EOR value on OS of
45 insular LGGs, demonstrating a 5-year OS rate of 100% when
EOR was .90% and 84% for an EOR ,90%. In line with this
recent article, as mentioned above, we described a volumetric
statistical analysis focused on insular nonenhancing gliomas,
FIGURE 3. Kaplan-Meier curves revealing the overall survival in patients with
insular nonenhancing glioma based on histological findings. Patients with insular
oligodendroglioma have a better prognosis. LGG, low-grade glioma; WHO,
World Health Organization.
FIGURE 4. Kaplan-Meier curves revealing the overall survival in patients with
insular nonenhancing glioma stratified by extent of resection (A) and by the
postoperative tumoral residual volume (B). A more extensive resection provides
a survival advantage.
INSULAR NONENHANCING GLIOMA SURGERY
NEUROSURGERY VOLUME 70 | NUMBER 5 | MAY 2012 | 1087
Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
confirming that a more aggressive resection correlates to a sig-
nificant improvement in OS compared with a simple debulking
procedure.
38,39
Overall, the present volumetric analysis evi-
denced a 5-year OS rate of 92% for patients with EOR . 90%
compared with an OS rate of 57% for those with EOR , 70%
(P , .001). Our results are in line with the growing evidence
reported in the literature, according to which the more extensive
resection at time of initial diagnosis may be the most favorable
prognostic factor for OS. Our findings provide further objec-
tive evidence in favor of the positive prognostic role played
by EOR. Moreover, the Cox proportional hazards model applied
in our patient population to investigate the effects of different
variables on OS showed that EOR represents the strongest
predictor factor for OS. The multivariate analysis in our study
population confirmed that EOR is an independent prognostic
factor for OS in patients with nonenhancing glioma, as previously
reported by Smith et al
38
in LGG surgery and by Stummer et al
46
in high-grade glioma surgery. Furthermore, our analysis showed
that patients with insular oligodendrogliomas have a better
prognosis with respect to those with fibrillary astrocytomas or
with WHO grade III gliomas (5-year OS rate of 92% vs 72% and
58%, respectively), confirming a different natural history of
the tumor and a different survival in patients with low-grade
oligodendrogliomas compared with those with low-grade astrocy-
tomas.
48
Finally, our findings indicate that patients with pre-
operative tumoral volume (T2 2 T1) , 30 cm
3
have a better
prognosis than those with preoperative tumoral volume (T2 2T1)
. 30 cm
3
(P = .02). The recently developed biomathematical
models of glioma growing patterns have highlighted that tumoral
growth results from 2 underlying mechanisms: proliferation and
diffusion. The first mechanism generates the bulky tumor (with
a regular shape, comparable in both postcontrast T1-weighted
MRI and T2-weighted preoperative MRI sequences), whereas the
second causes its diffusion along the white matter (having a complex
shape with digitations more visible on T2-weighted images).
Therefore, when proliferation is the major diffusivity phenomenon,
it does not affect the tumor shape (resulting in a regular shape,
comparable in both preoperative postcontrast T1-weighted MRI
and T2-weighted preoperative MRI sequences), which is grossly
bulky; in contrast, when the diffusive pattern is predominant, the
tumor will result in a complex shape with digitations along white
matter (resulting in a complex shape more visible on T2-weighted
MRI).
42
Moreover, Mandonnet et al
42
argued about the need to
complete the description of tumoral location, identifying the
status of subcortical pathways. Lang et al
8
asserted that lesions
with diffuse margins on T2-weighted MRI are less amenable to
radical resection with respect to those with sharp margins. Con-
sequently, in our study, we have tried to suggest an MRI pre-
operative predictive analytic index based on volumetric analysis to
discriminate the proliferative growing mechanism from the
infiltrative diffusive spreading mechanism along the white matter
and to assess preoperatively the degree of resection. Our volu-
metric analysis showed that when the diffusive mechanism
prevails over the proliferative mechanism, the difference in
preoperative tumoral volume computed on T2 and T1 MRI
progressively increases. Our data pointed out that when the pre-
operative difference of preoperative tumoral volume computed
on T2 and T1 images is . 30 cm
3
, the mean EOR is , 70%
(Figure 6). Our results showed that in cases with preoperative
FIGURE 5. Kaplan-Meier curves revealing the overall survival (OS) of patients
with insular nonenhancing gliomas based on preoperative volumetric analysis of
differences between tumoral mass computed on T2-weighted MRIs and postcontrast
T1-weighted MRIs. Patients with preoperative (T2 2T1) MRI volume ,30 cm
3
demonstrated a better 5-year OS compared with those with preoperative (T2 2T1)
MRI volume .30 cm
3
(P = .018). A major volumetric difference between T2- and
postcontrast T1-weighted MRI sequences suggests a greater propensity of the tumor to
have a diffuse growing pattern and consequently to be less resectable.
TABLE 3. Prognostic Factors and Overall Survival: Multivariate
Analysis Data (Cox Proportional Hazards Model) in 66 Patients
With Insular Nonenhancing Gliomas
a
Factor
OS
HR 95% CI P
Age 0.946 0.889-2.542 1.01 (NS)
Preoperative Karnofsky
Performance Status
0.917 0.707-1.188 .51 (NS)
Preoperative tumoral volume
on T2-weighted images
0.018 1.001-1.036 .04
Histological type (WHO II vs
WHO III)
0.015 0.037-0.705 .02
Difference of preoperative
tumoral volume (T2-weighted
volume 2 T1-weighted volume)
(, 30 vs . 30 cm
3
)
0.087 0.008-0.936 .04
EOR 0.919 0.871-0.970 .002
a
CI, confidence interval;. EOR, extent of resection; HR, hazard ratio; WHO, World
Health Organization.
SKRAP ET AL
1088 | VOLUME 70 | NUMBER 5 | MAY 2012 www.neurosurgery-online.com
Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
TABLE 4. Insular Glioma Studies Selected for Review
a
Author Cases, n
Histological
Data
New
Transient
Deficits
Permanent
Deficits Percent of Resection
Volumetric
Analysis
Second
Surgery,
n
Improvement of
Preoperative
Seizures
Yasargil
et al
12
177 77 Malignant
tumors
NA 5% NA No NA 84%
Zentner
et al
41
30 2 WHO I,
13 WHO II,
6 WHO III,
9 WHO IV
63% 10% 17% (5 Cases) total
removal, 70% (21
cases) EOR . 80%,
13% (4 cases) EOR ,
80% (postcontrast T1-
weighted sequences)
No NA 89%
Vanaclocha
et al
40
23 5 WHO I,
11 WHO II,
3 WHO III,
4 WHO IV
22% 9% 86.9% Complete
resection, 13.1%
subtotal resection
(postcontrast T1-
weighted sequences)
No 5 NA
Lang et al
8
22 1 PNET,
11 WHO II,
5 WHO III,
5 WHO IV
36% 9% 42% of Cases
EOR . 90%, 51% of
cases EOR . 70%, 7%
of cases EOR , 70%
(postcontrast T1-
weighted sequences)
Yes NA NA
Hentschel
and
Lang
7
36 22 WHO II,
7 WHO III,
4 WHO IV,
3 others
11% Transient
motor deficits,
30% transient
speech deficits
8% Permanent
motor deficits,
, 3%
permanent
speech deficits
NA NA NA NA
Moshel
et al
27
38 28 WHO II,
6 WHO III,
4 WHO IV
16% 8% 55% (21 Cases) gross
total removal, 17% (7
cases) near-gross
total removal, 28%
(10 cases) subtotal
removal (T2 weighted
sequences)
No NA NA
Simon
et al
11
101
Operations
in 94
patients
6 WHO I,
30 WHO II,
44 WHO
III, 21
WHO IV
NA Permanent
hemiparesis
9%
42% of Cases
EOR . 90%, 51% of
cases EOR . 70%, 7%
of cases EOR , 70%
(postcontrast T1-
weighted sequences)
No 7 76% of Patients
with preoperative
pharmacologically
resistant seizures
Duffau
5
51 51 WHO II 59% 4% 77% of Resections were
total or subtotal in
FLAIR imaging
sequences
Yes 9 78% of Patients
with preoperative
pharmacologically
resistant seizures
Sanai
et al
9
115
Operations
in 104
patients
45 WHO II,
70 WHO IV
14.4% 6% Median EOR for LGG =
82% (range, 31%-
100%) (T2- or FLAIR-
weighted sequences);
median EOR for
HGG=81% (range,
47%-100%)
(postcontrast T1-
weighted sequences)
Yes 11 NA
a
EOR, extent of resection; FLAIR, fluid attenuated inversion recovery; HGG, high-grade gliomas; LGG, low-grade gliomas; NA, not applicable (absence of data); PNET, primitive
neuroectodermal tumor; WHO, World Health Organization.
INSULAR NONENHANCING GLIOMA SURGERY
NEUROSURGERY VOLUME 70 | NUMBER 5 | MAY 2012 | 1089
Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
(T2 2 T1) MRI volume , 30 cm
3
(subgroup A = 39 cases), the
mean EOR was 88.5%, whereas when this difference was . 30
cm
3
(subgroup B = 28 cases), the mean EOR was 68.4%
(Subgroup A patients demonstrated a mean EOR of 88.5%,
whereas subgroup B patients had a mean EOR of 68.4%; t test
for mean percent EOR comparison [df = 64] = 6.8; P = .000).
Thus, this index could represent a prognostic preoperative
value of EOR itself. Moreover, after controlling for the role of
preoperative (T2 2T1) MRI tumoral volume, we found that the
EOR remained the strongest predictive factor on OS. However,
after Cox analysis, the value of preoperative (T2 2 T1) MRI
volume had a significant preoperative prognostic value (P = .04)
on OS, although weaker than EOR (P = .002). As far as the
methodological procedure is concerned, in designing the present
study, we followed the methodology proposed by Smith et al
38
for the volumetric assessment of preoperative and postoperative
FIGURE 6. Preoperative and postoperative volumetric analyses of a frontotemporoinsular glioma (Yasargil type 5 tumors) computed with OSIRIX software. A, preoperative
T2-weighted MRIs showing a left frontotemporoinsular oligodendroglioma (axial and coronal slices). B, preoperative tumoral volume computed on postcontrast T1-weighted
MRI was 98 cm
3
(axial slices). C, overlap on preoperative T2-weighted MRI sequence of the tumoral region of interest defined on the postcontrast T1-weighted images (red) and
the T2-weighted images (green). The preoperative tumoral volume computed on T2-weighted MRI sequence was 112 cm
3
(axial slices). D, volumetric analysis of postoperative
tumoral residue on postcontrast axial T1-weighted MRIs evidenced a tumoral residual volume of 3 cm
3
(axial slices). E, volumetric analysis of postoperative tumoral residue
computed on T2-weighted MRIs showed a tumoral residual volume of 6 cm
3
(axial slices). The extent of the tumoral volume resection, computed on T2-weighted MRI sequence,
was 94.7%.
SKRAP ET AL
1090 | VOLUME 70 | NUMBER 5 | MAY 2012 www.neurosurgery-online.com
Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
tumoral volumes. In all cases, to compute the residual volume,
we used images acquired 4 months after surgery rather than the
immediately postoperative images. To date, different postoperative
MRI timing and sequences have been used in recently published
studies of EORevaluation. Standards to compute the postoperative
tumoral residue in gliomas with no or minimal enhancement have
not been established yet. Early postoperative MRI may over-
estimate residual tumor on T2-weighted MRIs because of post-
operative resection-induced changes.
49
Therefore, we decided to
compute the residual volume of the 4-month scan. Patients with
WHO grade II gliomas (53 cases) did not receive any immediate
postoperative treatments. Thus, in patients with WHO grade II
gliomas, the residual volume was not positively influenced by
postoperative treatments. Moreover, it is well know that WHO II
gliomas are slow-growing tumors, progressing at about 4 mm/y.
50
Consequently, it is extremely rare that the residual tumor has
experienced regrowth 4 months after surgery.
51
A limitation of
this study is that the 13 patients with WHO II gliomas were
treated postoperatively with radiotherapy. We recognize that in
these cases the postoperative radiotherapy may have interfered
with residual tumoral volume analysis. However, radiotherapy
may induce structural changes such as vasogenic edema,
demyelization, and gliosis in the white matter beyond the
boundaries of the area of irradiation. Such changes result in
homogeneous hyperintensity,
49,52
which might cause an over-
estimation of the postoperative tumoral volume computed on
T2-weighted images rather than an underestimation of EOR
computed 4 months after surgery. The future developments of
diffusion-weighted imaging and molecular MRI could detect
postoperative resection-induced phenomena from residual tumor
in early postoperative examinations, overcoming the doubts
arising from the interpretation of hyperintensity on T2-weighted
images.
52,53
Neurological Deficits and Functional Outcome
The insular lobe represents a complex functional
relay,
3-5,13,14,16,18,19,21,24,25,54-58
but no single clinical deficit
can be attributed unequivocally to the insular cortex alone,
considering the usual infiltration of surrounding structures.
40
It is
likely that this associative area is compensable.
5,13,54
Seizures are
present at the onset in . 80% of patients. Transient postoper-
ative deficits in insular glioma surgery are frequent, as well as their
secondary improvement, clearly demonstrating that this structure
seems not to be essential.
54
An immediate postsurgical deficit is
reported in the literature with incidence rates of 63%,
40
22%,
59
36%,
8
16%,
27
and 59%.
5
Hentschel and Lang
7
reported 11% of
transient motor deficits and 30% of transient speech deficits,
whereas Sanai et al
9
recorded transient postoperative deficits in
only 14.4% of cases. Our results showed an immediate worsening
of patients neurological status in 22 cases (33.4%), supporting
the idea of a functional reshape of the multimodal insular lobe.
13
As far as the vascular damage of lateral lenticulostriate arteries is
concerned, it represents the main cause of permanent neurological
deficit, and all reports emphasize the preservation of the lentic-
ulostriate arteries that supply the internal capsule.
5,7,8,11,12,22,24-28,40
The devastating sequelae in case of damage explains why Yasargil
et al,
12
Duffau,
5
and Simon et al
11
have suggested leaving a layer
of tumoral tissue along these vessels. In the last decade, the rate
of permanent deficit has decreased: 10% in the series of Zentner
et al
41
; 9% in the series of Lang et al,
8
Vanaclocha et al,
40
and
Simon et al
11
; and 8% in the series reported by Moshel et al.
27
Moreover, Sanai et al
9
noted postoperative permanent deficits in
6% of cases; Duffau,
5
in 4% of cases. In the present series, the
rate of definitive worsening was 6%.
Surgical Considerations
Recent advantages in microsurgical and brain mapping techni-
ques with neurophysiological monitoring have allowed an increase
of surgical indications for insular gliomas. In the present series, we
routinely used intraoperative transcranial MEPmonitoring, which
constitutes a tool in guiding the EOR and preventing or minimiz-
ing direct injury to the posterior limb of the internal capsule and
the superior limit toward the corona radiate.
7,8,11,27
Changes in
MEP amplitude represent a warning sign, and a sudden loss or
a significant reduction of MEP amplitude probably indicates that
an injury to perforating vessels has already occurred. In our
experience, MEPs are useful to predict a direct trauma of the
fibers, but they do not guarantee prevention of vascular damage.
Finally, we think that the use of intraoperative electrocorticog-
raphy is essential to record afterdischarge phenomena and electric
and clinical seizures, particularly if they are short-lasting focal
seizures that may interfere with responses induced by mapping
with the loss of the patients collaboration. The preoperative
surgical planning, especially the DTI fiber tracking, represents a
useful tool to analyze preoperatively the 3-dimensional relation-
ships between the tumor and the subcortical pathways.
30
Interestingly, our results demonstrated that the intraoperative use
of a functional guided navigation system in the second series is
related to a statistically significant increase in EOR (Series 1
patients demonstrated a mean EOR of 77%, whereas series 2
patients had a mean EOR of 83%). In addition, in the most
recent cases, we recorded specific speech disorders by continu-
ously performing neuropsychological testing during the entire
surgical procedure. We have found this kind of testing to be more
reliable than the sole subcortical stimulation of tracts inherent to
language function like IFOF or arcuate fibers.
CONCLUSION
A greater EOR has been proposed to potentiate the efficacy of
adjuvant therapy and to achieve a longer OS. Insular gliomas sur-
gery remains a challenge because of the vascular structures and the
functions of this area, particularly within the dominant hemi-
sphere. As a result of the preoperative functional techniques
(functional MRI and DTI), intraoperative monitoring with cor-
tical and subcortical mapping, and neurophysiological monitor-
ing, achieving a major resection for insular nonenhancing gliomas
INSULAR NONENHANCING GLIOMA SURGERY
NEUROSURGERY VOLUME 70 | NUMBER 5 | MAY 2012 | 1091
Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
is possible with an acceptable risk of permanent neurological
deficits.
Our preliminary data suggest that the routine use of continuous
intraoperative neuropsychological tasks may constitute an intra-
operative tool complementary to subcortical IES to monitor the
language functions/processes. Future developments in intraoper-
ative brainmapping with electrophysiological arrays systems might
improve our knowledge of the tumor-induced pathophysiological
mechanisms and potential plastic phenomena.
Disclosure
The authors have no personal financial or institutional interest in any of the
drugs, materials, or devices described in this article.
REFERENCES
1. Duffau H, Capelle L. Preferential brain locations of low-grade gliomas. Cancer.
2004;100(12):2622-2626.
2. Kalani MY, Kalani MA, Gwinn R, Keogh B, Tse VC. Embryological development
of the human insula and its implications for the spread and resection of insular
gliomas. Neurosurg Focus. 2009;27(2):E2.
3. Augustine JR. Circuitry and functional aspects of the insular lobe in primates
including humans. Brain Res Brain Res Rev. 1996;22(3):229-244.
4. Mesulam MM, Mufson EJ. The insula of Reil in man and monkey: architectonics,
connectivity, and function. In: Peters A, Jones EG, eds. Cerebral Cortex. New
York, NY: Plenum Press; 1985;4:179-226.
5. Duffau H. A personal consecutive series of surgically treated 51 cases of insular
WHO grade II glioma: advances and limitations. J Neurosurg. 2009;110(4):
696-708.
6. Ebeling U, Kothbauer K. Circumscribed low grade astrocytomas in the dominant
opercular and insular region: a pilot study. Acta Neurochir (Wien). 1995;132(1-3):
66-74.
7. Hentschel SJ, Lang FF. Surgical resection of intrinsic insular tumors. Neurosurgery.
2005;57(1 suppl):176-183.
8. Lang FF, Olansen NE, DeMonte F, et al. Surgical resection of intrinsic insular
tumors: complication avoidance. J Neurosurg. 2001;95(4):638-650.
9. Sanai N, Polley MY, Berger MS. Insular glioma resection: assessment of patient
morbidity, survival, and tumor progression. J Neurosurg. 2010;112(1):1-9.
10. Schatz CR, Kreth FW, Faist M, Warnke PC, Volk B, Ostertag CB. Interstitial
125-iodine radiosurgery of low-grade gliomas of the insula of Reil. Acta Neurochir
(Wien). 1994;130(1-4):80-89.
11. Simon M, Neuloh G, von Lehe M, Meyer B, Schramm J. Insular gliomas: the case
for surgical management. J Neurosurg. 2009;110(4):685-695.
12. Yasargil MG, von Ammon K, Cavazos E, Doczi T, Reeves JD, Roth P. Tumours of
the limbic and paralimbic systems. Acta Neurochir (Wien). 1992;118(1-2):40-52.
13. Duffau H, Taillandier L, Gatignol P, Capelle L. The insular lobe and brain
plasticity: lessons from tumor surgery. Clin Neurol Neurosurg. 2006;108(6):
543-548.
14. Ackermann H, Riecker A. The contribution of the insula to motor aspects of
speech production: a review and a hypothesis. Brain Lang. 2004;89(2):320-328.
15. Craig AD. How do you feelnow? The anterior insula and human awareness. Nat
Rev Neurosci. 2009;10(1):59-70.
16. Jones CL, Ward J, Critchley HD. The neuropsychological impact of insular cortex
lesions. J Neurol Neurosurg Psychiatry. 2010;81(6):611-618.
17. Williamson JW, McColl R, Mathews D. Evidence for central command activation
of the human insular cortex during exercise. J Appl Physiol. 2003;94(5):
1726-1734.
18. Bamiou DE, Musiek FE, Luxon LM. The insula (Island of Reil) and its role in
auditory processing. Brain Res Brain Res Rev. 2003;42(2):143-154.
19. Crespo-Facorro B, Kim J, Andreasen NC, OLeary DS, Bockholt HJ, Magnotta V.
Insular cortex abnormalities in schizophrenia: a structural magnetic resonance
imaging study of first-episode patients. Schizophr Res. 2000;46(1):35-43.
20. Naqvi NH, Bechara A. The hidden island of addiction: the insula. Trends Neurosci.
2009;32(1):56-67.
21. Paulus MP, Stein MB. An insular view of anxiety. Biol Psychiatry. 2006;60(4):
383-387.
22. Guenot M, Isnard J, Sindou M. Surgical anatomy of the insula. Adv Tech Stand
Neurosurg. 2004;29:265-288.
23. Reil JC. Untersuchungen uber den Bau des grossen Gehirns im Menchen: Vierte
Forsetzung VIII. Arch Physiol. 1809;9:136-146.
24. Tanriover N, Rhoton AL Jr, Kawashima M, Ulm AJ, Yasuda A. Microsurgical
anatomy of the insula and the sylvian fissure. J Neurosurg. 2004;100(5):
891-922.
25. Ture U, Yasargil DC, Al-Mefty O, Yasargil MG. Topographic anatomy of the
insular region. J Neurosurg. 1999;90(4):720-733.
26. Gibo H, Carver CC, Rhoton AL Jr, Lenkey C, Mitchell RJ. Microsurgical
anatomy of the middle cerebral artery. J Neurosurg. 1981;54(2):151-169.
27. Moshel YA, Marcus JD, Parker EC, Kelly PJ. Resection of insular gliomas: the
importance of lenticulostriate artery position. J Neurosurg. 2008;109(5):825-834.
28. Ture U, Yasargil MG, Al-Mefty O, Yasargil DC. Arteries of the insula. J Neurosurg.
2000;92(4):676-687.
29. Varnavas GG, Grand W. The insular cortex: morphological and vascular anatomic
characteristics. Neurosurgery. 1999;44(1):127-136.
30. Berman JI, Berger MS, Chung SW, Nagarajan SS, Henry RG. Accuracy of
diffusion tensor magnetic resonance imaging tractography assessed using intra-
operative subcortical stimulation mapping and magnetic source imaging.
J Neurosurg. 2007;107(3):488-494.
31. Duffau H, Thiebaut de SM, Mandonnet E. White matter functional connectivity
as an additional landmark for dominant temporal lobectomy. J Neurol Neurosurg
Psychiatry. 2008;79(5):492-495.
32. Oldfield R. The assessment and analysis of handedness: the Edinburgh inventory.
Neuropsychologia. 1971;9(1):97-113.
33. Louis DN, Ohgaki H, Wiestler OD, et al. The 2007 WHO classification of
tumours of the central nervous system. Acta Neuropathol. 2007;114(2):97-109.
34. Berger MS, Ojemann GA. Intraoperative brain mapping techniques in neuro-
oncology. Stereotact Funct Neurosurg. 1992;58(1-4):153-161.
35. Berger MS, Deliganis AV, Dobbins J, Keles GE. The effect of extent of resection
on recurrence in patients with low grade cerebral hemisphere gliomas. Cancer.
1994;74(6):1784-1791.
36. Ojemann G, Ojemann J, Lettich E, Berger M. Cortical language localization in
left, dominant hemisphere: an electrical stimulation mapping investigation in 117
patients. J Neurosurg. 1989;71(3):316-326.
37. Rosset A, Spadola L, Ratib O. OsiriX: an open-source software for navigating in
multidimensional DICOM images. J Digit Imaging. 2004;17(3):205-216.
38. Smith JS, Chang EF, Lamborn KR, et al. Role of extent of resection in the long-
term outcome of low-grade hemispheric gliomas. J Clin Oncol. 2008;26(8):
1338-1345.
39. Sanai N, Berger MS. Glioma extent of resection and its impact on patient
outcome. Neurosurgery. 2008;62(4):753-764.
40. Vanaclocha V, Saiz-Sapena N, Garcia-Casasola C. Surgical treatment of insular
gliomas. Acta Neurochir (Wien). 1997;139(12):1126-1134.
41. Zentner J, Meyer B, Stangl A, Schramm J. Intrinsic tumors of the insula:
a prospective surgical study of 30 patients. J Neurosurg. 1996;85(2):263-271.
42. Mandonnet E, Capelle L, Duffau H. Extension of paralimbic low grade gliomas:
toward an anatomical classification based on white matter invasion patterns.
J Neurooncol. 2006;78(2):179-185.
43. Keles GE, Chang EF, Lamborn KR, et al. Volumetric extent of resection and
residual contrast enhancement on initial surgery as predictors of outcome in adult
patients with hemispheric anaplastic astrocytoma. J Neurosurg. 2006;105(1):
34-40.
44. McGirt MJ, Chaichana KL, Attenello FJ, et al. Extent of surgical resection is
independently associated with survival in patients with hemispheric infiltrating
low-grade gliomas. Neurosurgery. 2008;63(4):700-707.
45. McGirt MJ, Chaichana KL, Gathinji M, et al. Independent association of extent of
resection with survival in patients with malignant brain astrocytoma. J Neurosurg.
2009;110(1):156-162.
46. Stummer W, Reulen HJ, Meinel T, et al. Extent of resection and survival in
glioblastoma multiforme: identification of and adjustment for bias. Neurosurgery.
2008;62(3):564-576.
47. Lacroix M, Abi-Said D, Fourney DR, et al. A multivariate analysis of 416 patients
with glioblastoma multiforme: prognosis, extent of resection, and survival.
J Neurosurg. 2001;95(2):190-198.
SKRAP ET AL
1092 | VOLUME 70 | NUMBER 5 | MAY 2012 www.neurosurgery-online.com
Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
48. Shaw EG, Scheithauer BW, OFallon JR. Supratentorial gliomas: a comparative
study by grade and histologic type. J Neurooncol. 1997;31(3):273-278.
49. Belhawi SM, Hoefnagels FW, Baaijen JC, et al. Early postoperative MRI
overestimates residual tumour after resection of gliomas with no or minimal
enhancement. Eur Radiol. 2011;21(7):1526-1534.
50. Mandonnet E, Delattre JY, Tanguy ML, et al. Continuous growth of mean tumor
diameter in a subset of grade II gliomas. Ann Neurol. 2003;53(4):524-528.
51. Mandonnet E, Jbabdi S, Taillandier L, et al. Preoperative estimation of residual
volume for WHO grade II glioma resected with intraoperative functional
mapping. Neuro Oncol. 2007;9(1):63-69.
52. Hein PA, Eskey CJ, Dunn JF, Hug EB. Diffusion-weighted imaging in the follow-
up of treated high-grade gliomas: tumor recurrence versus radiation injury. AJNR
Am J Neuroradiol. 2004;25(2):201-209.
53. Zhou J, Tryggestad E, Wen Z, et al. Differentiation between glioma and radiation
necrosis using molecular magnetic resonance imaging of endogenous proteins and
peptides. Nat Med. 2011;17(1):130-134.
54. Duffau H, Capelle L, Lopes M, Faillot T, Sichez JP, Fohanno D. The insular
lobe: physiopathological and surgical considerations. Neurosurgery. 2000;47(4):
801-810.
55. Oppenheimer SM, Gelb A, Girvin JP, Hachinski VC. Cardiovascular effects of
human insular cortex stimulation. Neurology. 1992;42(9):1727-1732.
56. Ostrowsky K, Isnard J, Ryvlin P, Guenot M, Fischer C, Mauguiere F. Functional
mapping of the insular cortex: clinical implication in temporal lobe epilepsy.
Epilepsia. 2000;41(6):681-686.
57. Schreckenberger M, Siessmeier T, Viertmann A, et al. The unpleasantness of tonic
pain is encoded by the insular cortex. Neurology. 2005;64(7):1175-1183.
58. Shelley BP, Trimble MR. The insular lobe of Reil: its anatamico-functional,
behavioural and neuropsychiatric attributes in humans: a review. World J Biol
Psychiatry. 2004;5(4):176-200.
59. Duffau H, Capelle L, Sichez N, et al. Intraoperative mapping of the subcortical
language pathways using direct stimulations: an anatomo-functional study. Brain.
2002;125(1):199-214.
Acknowledgment
We thank Francesca Valent, MD, MSPH (Istituto di Igiene ed Epidemiologia
Clinica Azienda Ospedaliero-Universitaria di Udine Via Colugna, 50, 33100
Udine, Italy), for valuable advice in the statistical analysis.
COMMENTS
T
he authors have described their careful approach to 66 patients un-
dergoing 71 resections with an emphasis on using neurophysiological
monitoring and functional imaging to extend resections and to determine
the effect of extent of resection on nonenhancing insular gliomas with
a medianfollow-up of 4.3 years. The authors divided their series into a first
and second half, showing that the major complications were early in the
series (25 patients) and the extent of resection improved inthe latter half of
the series (41 patients). Finally, in the last 12 patients, continuous neu-
ropsychological monitoring was performed. The surgical approach
changed during the series from a transsylvian approach to a transcortical
approach, which the authors believed assisted them with exposure and
extent of resection and the clear need for lateral lenticulostriate artery
preservation. The main finding regardless of the technology used was
that a better extent of resection (. 90%) and postoperative tumor vol-
umes (, 20 cm
3
) led to longer survivals. The authors are to be con-
gratulated for their careful analysis and application of currently available
technology both extraoperatively and intraoperatively to improve patient
outcomes and results in these difficult-to-treat diffuse tumors.
Michael M. Haglund
Durham, North Carolina
I
n this article, the authors provide important data confirming what
previous surgical series have shown, namely that radical resections of
insular tumors can be performed safely and that patient survival is en-
hanced by more aggressive surgical resections. This article is now one of
several that describe resection of insular tumors. Historically, credit must
be given to Professor Yasargil for recognizing the origin of tumors in the
insular and for opening neurosurgeons eyes to the feasibility of resecting
tumors in this area.
1
In the past 10 years, several surgeons have reported
that radical resection for insular tumors can be performed with low
morbidity, particularly if the surgeon understands the sulcal and vascular
anatomy of the insula.
2-4
More recent series have shown that radical
resection is also important because it is likely to extend patient sur-
vival.
5-7
In this context, the authors now add their series of 66 non-
enhancing gliomas to this expanding body of literature on neurosurgical
management of insular tumors. Most important, they show that more
aggressive resections are associated with improved survival.
Although their data on extent of resection (EOR) are clearly valuable,
the authors also emphasize that awake craniotomy and intraoperative
testing improved both EOR and outcome. Indeed, my experience with
insular tumors supports this concept. Awake craniotomy not only informs
the surgeon of possible problems but also, potentially more important,
gives the surgeon confidence that the patient is not having a problem and
that the surgeon can continue resection in this highly eloquent area. The
assurance given to the surgeon by knowing that the patient is doing well
results in more radical tumor removals.
Another important piece of information fromthis series is the high rate
(20%) of World Health Organization (WHO) grade III tumors in this
series of 66 nonenhancing tumors. This finding emphasizes that clinicians
must not assume that patients with nonenhancing tumors have low-grade
(WHO grade II) gliomas. Instead, they should encourage early in-
tervention to make a proper diagnosis.
Interestingly, it appears that 2 methodological schools are developing
for insular tumors. One school relies on the transsylvian approach in
which the sylvian fissure is split widely and normal operculum is not
removed to access the insula. The other school uses a transcortical
approach in which normal but functionally silent opercula is removed
to access the tumor. Interestingly, the authors of this article began as
proponents of the transsylvian approach but switched to the transcortical
method. Clearly, there is no correct way to resect insular tumors. Critical
to success, however, is that the surgeon understands the detailed
topographic and vascular anatomy of the insula and its surrounding
structures and becomes familiar with the nuances of the approach he/she
will use. My preference is the transsylvian approach, which I believe
provides early identification and dissection of the middle cerebral artery
branches before tumor resection, early demarcation of the peri-insular
sulci to define the borders of the tumor, and preservation of the over-
lying, often normal operculum.
3,4
Indeed, even when the tumor extends
into the temporal or frontal lobe, I prefer to spilt the sylvian fissure first
and to define the insular vessels and borders before resecting tumor in the
opercula. Admittedly, the transsylvian approach can add challenges
because of the sylvian veins, and more work needs to be done on the
sylvian veins and their patterns of flow.
Finally, it is worth mentioning that increasingly there are reports of
series that include large numbers of insular tumors. This trend is com-
mendable in that it indicates that neurosurgeons no longer viewthe insular
as a surgical no mans land. As a word of caution, however, it is im-
portant to distinguish between tumors that are purely insular gliomas
with a main tumor mass primarily in the insular and gliomas that arise
mostly in the frontal or temporal lobes and have only a minority of the
INSULAR NONENHANCING GLIOMA SURGERY
NEUROSURGERY VOLUME 70 | NUMBER 5 | MAY 2012 | 1093
Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
tumor mass infiltrating into the insular (eg, see Figure 2 vs 6). This
distinction is critical when evaluating surgical morbidity and the impact
of EOR on outcome. Clearly, resecting 90% of a temporal tumor that
has some extension into the insular may mean that only a small part (or
potentially none) of the insular portion of the tumor is resected. In
contrast, resecting 90% of a purely insular tumor means that dissection
of the entire insula was required. As clinical investigators, we must be
careful to ensure that we do not oversell our results because of inclusion
of cases that are not truly insular tumors. As we become more proficient
at removing insular tumors, we must continue to assess the risks and
benefits of our treatments in an unbiased fashion.
Frederick F. Lang
Houston, Texas
1. Yasargil MG, von Ammon K, Cavazos E, Doczi T, Reeves JD, Roth P. Tumours of
the limbic and paralimbic systems. Acta Neurochir (Wien). 1992;118(1-2):40-52.
2. Duffau H, Capelle L, Lopes M, Faillot T, Sichez JP, Fohanno D. The insular lobe:
physiopathological and surgical considerations. Neurosurgery. 2000;47(4):801-811.
3. Lang FF, Olansen NE, DeMonte F, et al. Surgical resection of intrinsic insular
tumors: complication avoidance. J Neurosurg. 2001;95(4):638-650.
4. Hentschel SJ, Lang FF. Surgical resection of intrinsic insular tumors. Neurosurgery.
2005;57(1 suppl):176-183.
5. Duffau H. A personal consecutive series of surgically treated 51 cases of insular
WHO Grade II glioma: advances and limitations. J Neurosurg. 2009;110(4):
696-708.
6. Sanai N, Polley MY, Berger MS. Insular glioma resection: assessment of patient
morbidity, survival, and tumor progression. J Neurosurg. 2010;112(1):1-9.
7. Simon M, Neuloh G, von Lehe M, Meyer B, Schramm J. Insular gliomas: the case
for surgical management. J Neurosurg. 2009;110(4):685-695.
T
he authors have retrospectively evaluated oncological and functional
results in a series of 66 patients with a diffuse nonenhancing glioma
involving the insular lobe, with special emphasis on volumetric analysis of
the tumor. They observed 6%permanent worsening with a median extent
of resection (EOR) of 80% (median follow-up, 4.3 years). A signification
correlation was demonstrated between EORand 5-year overall survival. In
addition, on preoperative imaging, the difference between tumoral vol-
umes measured on T2- and T1- weighted magnetic resonance imaging
(MRI) was also correlated with 5-year overall survival. The authors con-
cluded that, thanks to intraoperative cortico-subcortical mapping and
neurophysiological monitoring, an extensive resection was possible with
an acceptable risk and a significant impact on 5-year overall survival.
The authors are to be congratulated for their excellent results in this
complex surgery. Beyond the fact that they confirmed what was previously
reported by several authors,
1-3
ie, that surgery for insular glioma can be
performed with a low risk of permanent deficit despite a major resection
and with a significant impact on the natural history of the gliomas, they
provided additional interesting information.
First, the authors changed their surgical technique between the first and
second periods, moving from a transsylvian approach to a transcortical
approach with continuous neurocognitive monitoring under local anes-
thesia in the last 12 patients. Interestingly, in the latter half of their series,
the rate of major complications decreased while the extent of resection
improved. These findings are in agreement with the literature, showing
a higher rate of definitive neurological worsening around 8% to 10% in
the series with a transsylvian approach (see elsewhere
2
for a review),
whereas this rate is only around 3.9% to 6% in the series using
a transopercular approach,
2,3
as confirmed in the present article. Indeed,
the use of a subpial dissection may limit the risk of vascular injury by
avoiding a direct dissection of the middle cerebral artery branches and by
preventing a possible spasm. Indeed, it is important to underline that
long perforator arteries that supply the corona radiata arise from the M2
segment of the middle cerebral artery. As a consequence, a deep stroke
may also be due to damage of the vessels in the sylvian fissure, in addition
to injury of the lenticulostriate arteries.
Therefore, the removal of the operculum in the dominant hemisphere
implies that surgery be performed in awake patients, with online func-
tional feedback throughout the resection, to preserve the quality of life.
Furthermore, it is possible to detect the subcortical pathways subserving
both movement and language function at the end of the procedure to
continue the resection until eloquent structures have been encountered
while minimizing the risk of permanent deficit, as previously reported.
4
Moreover, the authors demonstrated that it was crucial to perform in all
cases objective volumetric analysis of gliomas before and after surgery,
on both T1- and T2-weighted MRI, because of the significant impact
of volume and growing pattern on median survival in nonenhancing
gliomas. As a consequence, it seems currently unacceptable not to
evaluate the extent of resection on at least T2-/fluid-attenuated inversion-
recoveryweighted MRI in the series on nonenhancing gliomas, in con-
trast to the classic literature that for a long time was based on the sole
subjectivity of the neurosurgeons.
Finally, the authors confirmed that around 20% of nonenhancing glio-
mas had already at least some foci of malignant transformation, supporting
early surgery in diffuse low-grade gliomas (even in insular gliomas), rather
than the traditional wait and see approach, which should definitely be
avoided, in accordance with the recent European guidelines.
5
In this state
of mind, resection of (insular) diffuse low-grade gliomas incidentally
discovered in asymptomatic patients may be considered.
6
Hugues Duffau
Montpellier, France
1. Lang FF, Olansen NE, DeMonte F, et al. Surgical resection of intrinsic insular
tumors: complications avoidance. J Neurosurg. 2001;95(4):638-650.
2. Duffau H. A personal consecutive series of surgically treated 51 cases of insular
WHO grade II glioma: advances and limitations. J Neurosurg. 2009;110(4):
696-708.
3. Sanai N, Polley MY, Berger MS. Insular glioma resection: assessment of patient
morbidity, survival and tumor progression. J Neurosurg. 2010;112(1):1-9.
4. Duffau H, Moritz-Gasser S, Gatignol P. Functional outcome after language
mapping for insular World Health Organization grade II gliomas in the dominant
hemisphere: experience with 24 patients. Neurosurg Focus. 2009;27(2):E7.
5. Soffietti R, Baumert BG, Bello L, et al. Guidelines on management of low-grade
gliomas: report of an EFNS-EANO Task Force. Eur J Neurol. 2010;17(9):1124-1133.
6. Duffau H. Awake surgery for incidental WHO grade II gliomas involving eloquent
areas [published online ahead of print December 6, 2011]. Acta Neurochir. doi:10.
1007/s00701-011-1216-x.
SKRAP ET AL
1094 | VOLUME 70 | NUMBER 5 | MAY 2012 www.neurosurgery-online.com
Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.