- Email: [email protected]
Volumetric Analysis of Extent of Resection, Survival, and Surgical Outcomes for Insular Gliomas
Accepted Manuscript Volumetric analysis of extent of resection, survival, and surgical outcomes, for insular gliomas Chikezie I. Eseonu, MD, Karim ReFaey, MD, Oscar Garcia, MPH, Gugan Raghuraman, Alfredo Quinones-Hinojosa, MD PII:
S1878-8750(17)30496-5
DOI:
10.1016/j.wneu.2017.04.002
Reference:
WNEU 5518
To appear in:
World Neurosurgery
Received Date: 24 January 2017 Revised Date:
30 March 2017
Accepted Date: 1 April 2017
Please cite this article as: Eseonu CI, ReFaey K, Garcia O, Raghuraman G, Quinones-Hinojosa A, Volumetric analysis of extent of resection, survival, and surgical outcomes, for insular gliomas, World Neurosurgery (2017), doi: 10.1016/j.wneu.2017.04.002. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Volumetric analysis of extent of resection, survival, and surgical outcomes, for insular gliomas
1 2 3 4 5
Authors: Chikezie I. Eseonu, MDa, Karim ReFaey, MDa,b, Oscar Garcia, MPHa,b, Gugan Raghuramana, Alfredo Quinones-Hinojosa, MDa,b
6
9 10
a
Department of Neurological Surgery and Oncology Outcomes Lab. Johns Hopkins University, Baltimore, Maryland b Neurological Surgery Department, Mayo Clinic, Jacksonville, Florida
11 12 13 14
Conflict of Interest: There is no conflict of interests to report or financial disclosures related to this manuscript
15
Running Title: Insular glioma extent of resection
17 18
M AN U
16
SC
8
RI PT
7
Keywords: Insula, glioma, low grade, high grade, insular tumors, survival
19
Corresponding author:
21
Alfredo Quiñones-Hinojosa, M.D.
22
Brain Tumor Stem Cell Laboratory
23
Department of Neurologic Surgery
24
Mayo Clinic, Florida
25
4500 San Pablo Rd. S
26
Jacksonville, FL 32224
27
Phone: 410-502-2906 / fax: 410-502-5559 / email: [email protected]
30 31 32 33 34 35 36 37 38 39 40 41 42 43
EP
29
AC C
28
TE D
20
ACCEPTED MANUSCRIPT Insular glioma extent of resection
1
2
Abstract
2 3 4 5 6
Background Insular gliomas are challenging tumors to surgically resect due to the anatomy surrounding them. This study evaluates the role of extent of resection(EOR) and molecular markers on surgical outcome and survival for insular gliomas.
8 9 10 11 12
RI PT
7
Methods Seventy-four patients who had undergone an initial resection for an insular glioma by the same surgeon from 2006 to 2016 were analyzed. Low(grade II) and high(grade III/IV) grade gliomas were analyzed for the prognostic role of volumetric EOR and molecular markers on patient survival outcomes.
15 16 17 18 19 20 21 22 23 24
Results The cohort includes 25 low grade gliomas (LGGs) patients(33.8%), and 49 high grade glioma(HGGs) patients(66.2%). The median EOR was 91.7%(range 10-100%). New permanent postoperative deficits were found in 2.7% of patients. LGG patients with a ≥90% EOR had a 5year survival rate of 100% and patients with a <90% EOR had 5-year survival of 80%. HGG patients with a ≥90% EOR had a 2-year survival rate of 83.7%, and patients with a <90% EOR had 2-year survival of 43.8%. For LGGs, the EOR was predictive of OS(p=0.017), progression free survival(PFS, p=0.039), and malignant progression free survival(MPFS, p=0.014), while the 1p/19q co-deletion was predictive for PFS(p=0.014). For HGGs, the EOR was predictive of OS(p=0.020) and PFS(p=0.024). Preoperative tumor volume was a factor that most significantly affected the EOR for insular gliomas(R2=0.053,p=0.048).
M AN U
14
SC
13
25
29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
TE D
28
Conclusion Extensive resections of insular gliomas can be achieved with low morbidity and can improve OS and PFS. In this series of low-grade gliomas, EOR was associated with longer MPFS, and the 1p/19q co-deletion was predictive of PFS.
EP
27
AC C
26
ACCEPTED MANUSCRIPT Insular glioma extent of resection
1
3
Introduction Insular gliomas are the most common intrinsic tumor of the insular cortex and make up
3
around 25% of all low-grade gliomas and 10% of high grade gliomas.1-3 Tumors in this cortical
4
area are surrounded by eloquent brain that controls motor and language function. In addition, the
5
vascular supply for the descending motor pathway also passes through the insular region.4-6
6
Patients with insular tumors can often present with devastating intractable epilepsy or motor
7
dysphasia making it essential to properly treat these lesions.7-10
SC
RI PT
2
Due to the deep location in the Sylvian fissure and complex anatomy that surrounds the
9
area, insular tumors are a challenging tumor to resect.11,12 The standard of care for insular
10
gliomas includes a maximal safe resection followed by chemoradiation therapy.13 Surgery in the
11
insular region may require opening the Sylvian fissure for insular exposure which puts vital
12
functional cortical areas, motor and language subcortical fibers, and vascular structures at risk in
13
this region.12,14 There is a high risk of neurological deficits following resection of insular tumors
14
given the proximity of the insula to vital vascular and neural structures, and severe speech and
15
motor disorders can result in the postoperative period.15
TE D
M AN U
8
Given the high risk of surgery, some authors have reported that insular tumors are
17
inoperable, opting for stereotactic biopsies for diagnosis and then chemotherapy and or
18
radiotherapy as an alternative treatment.2,16-18 However, other authors have reported favorable
19
reports with insular tumor resections based on specialized microsurgical techniques and an
20
understanding of the insular anatomy.13,17,19-22 Despite the high risks of surgery in the insular
21
region, there are limited studies that evaluate the role of the extent of resection (EOR), analyzed
22
volumetrically, on patient outcomes and survival for insular gliomas.23,24 Additionally, with an
23
increasing understanding about glioma genetics and their role in prognosis for low grade gliomas
AC C
EP
16
ACCEPTED MANUSCRIPT Insular glioma extent of resection
4
1
(isocitrate dehydrogenase 1/2 [IDH1/2] mutation and 1p/19q co-deletion), no previous
2
volumetric analysis for insular gliomas have accounted for these variables when evaluating the
3
role of EOR on patient survival outcome.25 In this study, we present a cohort of insular gliomas operated by a single surgeon at a
5
single center and evaluate the role of EOR and molecular markers on patient outcome and
6
survival.
RI PT
4
SC
7
8
10
Methods
M AN U
9
Patient Selection
Seventy-four consecutive patients with insular gliomas resected for the first time by a
12
single surgeon between January 2006 and July 2016 were prospectively collected and
13
retrospectively reviewed. Patients were ≥ 18 years old. Pathology assessment of the tumor was
14
based on WHO guidelines. Patients with gliomatosis cerebri were not included in this study.
15
Clinical data was collected from the patient chart review.
16
18
Surgical approach
All insular tumors were approached via a transcortical approach. The location and side of
AC C
17
EP
TE D
11
19
the hemisphere of the tumor determined whether the tumor was resected under general
20
anesthesia (45 cases) or by awake craniotomy (29 cases). Our method for awake anesthesia has
21
been previously described.26 After the craniotomy was performed and dura opened, cortical and
22
subcortical sensorimotor mapping was conducted. Once functional regions were identified,
23
corticectomies above and below the sylvian fissure in noneloquent regions were made allowing
ACCEPTED MANUSCRIPT Insular glioma extent of resection
5
for tumor resection around the sylvian vessels. Cortical mapping of the medial border of the
2
tumor was also conducted if needed to identify the internal capsule. The goal of all surgeries in
3
this cohort was for a safe, 100% total resection, however, tumor invasion into areas deemed
4
eloquent for motor, sensory, or language areas by intraoperative cortical/subcortical mapping, or
5
extensive involvement of the middle cerebral artery within the tumor, limited the extent of
6
resection.
RI PT
1
8
SC
7
Patient Outcome Data
Patients were examined preoperatively, immediately postoperatively, and during their
10
follow up visits by the senior surgeon (AQH). Neurological deficits were determined by
11
identifying new or worsening deficits related to motor, sensory, or language function. Outpatient
12
examinations were conducted at 1 month, 6 months, and then annually following surgery.
13
Transient deficits were considered new or worsening neurological deficits following surgery that
14
resolved prior to the 6-month follow up clinic visit. Permanent deficits were neurological deficits
15
that persisted beyond the 6-month follow up clinic visit. The length of hospitalization was
16
calculated based on the duration of hospital stay from the day of surgery until the day of
17
discharge.
TE D
EP
Survival data was also analyzed in the form of overall survival (OS), progression free
AC C
18
M AN U
9
19
survival (PFS), and malignant progression free survival (MPFS). Overall survival was the time
20
between initial surgery and death. PFS was the time from initial surgery to unequivocal
21
demonstration of tumor re-growth on follow up imaging. MPFS was defined as the cases where
22
grade II gliomas, determined from pathology from the initial surgery, were found to transform
23
into a higher grade lesion, via histopathology from a subsequent resection/biopsy, or
ACCEPTED MANUSCRIPT Insular glioma extent of resection
6
1
enhancement from follow up imaging. Patients with no progression of their tumor were censored
2
as of their last documented imaging date.
3
Tumor Volumetric Analysis
RI PT
4
The preoperative tumor volume was determined by using the T1-weighted with
6
gadolinium or T2-weighted and fluid-attenuated inversion recovery (FLAIR) MRI (1.5 to 3 mm
7
axial cuts) depending on the tumor type. The OsiriX software (Pixmeo) was used to quantify
8
tumor volume by two clinicians with neuroscience and radiology training who were blinded to
9
the patient information, as we have previously described.26-29 If the inter-observer volume
10
difference was large, then a third clinician, also blinded to the cohorts, would evaluate the
11
volumes, and the mean volume was determined from the two similarly calculated volumes. The
12
postoperative tumor volume was calculated using the MRI images obtained 48 hours following
13
surgery, in the same manner as the preoperative imaging. The extent of resection was calculated
14
with the equation (preoperative - postoperative tumor volume)/preoperative tumor volume.
15
Statistical Analysis
EP
16
TE D
M AN U
SC
5
Univariate statistics were conducted to produce descriptive statistics that were reported as
18
number of patient subjects and percent for categorical variables, median and range for
19
continuous nonparametric variables, and mean and standard deviations for continuous parametric
20
variables. Kaplan-Meier curves were used to generate OS, PFS, and MPFS for all patients. The
21
Kaplan-Meier curves plotted EOR and OS or PFS, using the ≥ 90, 70-89, and <70% EOR groups
22
to maintain consistency with previous insular glioma studies.24,30 For the MPFS, the EOR for the
23
Kaplan-Meier curve was divided into ≥ 90 and < 90% as this has been a common grouping when
AC C
17
ACCEPTED MANUSCRIPT Insular glioma extent of resection
7
assessing the impact of EOR.13 Differences were assessed for statistical significance using the
2
log-rank test. A univariate Cox proportional hazards models were used to determine predictors of
3
survival curves. For survival risk factors, variables that had a p-value <0.15 on the univariate
4
analysis were entered into the multivariate analysis. The multivariate Cox proportional hazards
5
models required a p < 0.05 for significance to be achieved. EOR was treated as a continuous
6
variable in the Cox proportional hazards models to determine if it predicted OS, PFS, MPFS. A
7
linear regression was used to evaluate variables associated with EOR. A biostatistician (O.G)
8
contributed to the statistical analysis. STATA 14 was used for all statistical analyses. The
9
university institutional review board (IRB) approved this study.
M AN U
SC
RI PT
1
10
11
Result
12
Patient Characteristics
The patient cohort included 42 males and 32 females, with the median age of 54 years
14
(range 18-80 years). Tumors were in the left hemisphere in 58.1% of the patients, and awake
15
craniotomy in 39.2% of the cases. Patients most frequently presented with preoperative seizures
16
(35.1%). Grade II gliomas were found in 33.8% of patients, Grade III for 21.6% of patients, and
17
Grade IV in 44.6% of patients. Fifty-three patients had postoperative chemotherapy (71.6%) and
18
33 patients (44.6%) had postoperative radiation.
20
EP
AC C
19
TE D
13
Tumor Characteristics
21
The median preoperative tumor volume was 43.6 cm3 and the median postoperative
22
tumor volume was 3.7 cm3. In this series, according to the Berger-Sanai zone classification for
23
insular tumors, there were 21 Zone I tumors (28.4%), 6 Zone II (8.1%), 7 Zone III (9.5%), 2
ACCEPTED MANUSCRIPT Insular glioma extent of resection
8
Zone IV (2.7%), 7 Zone I+II (9.5%), 11 Zone I+IV (14.9%), 8 Zone II+ III (10.8%), 2 Zone
2
III+IV (2.7%) and 10 giant insular tumors that involved all zones (13.5%).13 Following the first
3
surgery, 19 patients (25.7%) had a EOR of less than 70%, 15 patients (20.3%) had a EOR
4
between 70% and 89%, and 40 patients (54%) had an EOR of 90% or greater. For the patients
5
that had less than 100% total resection (61 patients), 85.2% of the resections were limited by
6
tumor invasion into eloquent cortical/subcortical regions determined by intraoperative mapping,
7
while 14.8% were limited by extensive tumor involvement with the middle cerebral artery or
8
feeding branches of the MCA. For the low grade gliomas, 19 patients (76%) had the IDH1
9
mutation, and 18 patients (72%) had a 1p/19q codeletion (Table 1).
M AN U
SC
RI PT
1
10
11
Postoperative outcome
At the six-month follow up, postoperative seizures were seen in 16% of patients.
13
Transient motor deficits were seen in 8.1% of patients, while permanent deficits were seen in
14
2.7% of patients. From the motor deficit patients, ischemic injury was identified in one patient
15
with a permanent deficit based on postoperative diffusion-weighted imaging. New sensory and
16
language deficits were found in 4.1% and 6.8% of patients, respectively. No ischemic injury was
17
identified on postoperative imaging in any patient with postoperative sensory or language
18
deficits. The median length of hospitalization was 4.5 days (range 1-22 days) (Table 2).
20
EP
AC C
19
TE D
12
Survival Analysis
21
Overall Survival: At the last follow up there were a total of 16 (21.6%) patients that died
22
from their disease. The median follow up time of 4.4 years in surviving patients. Death occurred
ACCEPTED MANUSCRIPT Insular glioma extent of resection
9
1
in 3 (4.1%) Grade II gliomas, 6 (8.1%) Grade III gliomas, and 14 (18.9%) Grade IV gliomas
2
(Figure 1). Low Grade Gliomas Overall Survival:
4
For the Low grade (grade II) gliomas (LGGs), a univariate Cox proportional analysis
5
showed EOR, IDH1 mutation, and 1p19q deletion as significant predictors for OS and PFS
6
(Table 3). A multivariate Cox proportional analysis showed the EOR, treated as a continuous
7
variable, had a significant association with OS, while controlling for IDH1 mutation and
8
presence of the 1p19q codeletion, (HR 0.882, 95% CI 0.783-0.994, p= 0.039) (Table 4). Patients
9
with a ≥ 90% EOR had a 5-year survival rate of 100% and patients with a < 90% EOR had 5
10
year survival of 80%. For graphic display, survival was categorized into three groups ≥90%, 70-
11
89%, and <70%. A stepwise improvement in OS was associated with increasing EOR among our
12
LGG cohort (p= 0.041, Figure 2).
M AN U
SC
RI PT
3
High Grade Gliomas Overall Survival:
14
For patients with high grade gliomas (HGGs), Grade III and Grade IV gliomas, a Cox
15
proportional hazards analysis shows a predictive association between EOR and OS (HR 0.974,
16
95% CI 0.953-0.996, p= 0.020, Table 5). Patients with a ≥ 90% EOR had a 2-year survival rate
17
of 83.7%, and patients with a < 90% EOR had 2 year survival of 43.8%. The stratification of
18
EOR into ≥90%, 70-89%, and <70% showed significant improvement in OS with the Kaplan
19
Meier curves (p=0.0165, Figure 2).
EP
AC C
20
TE D
13
Progression free survival: Forty-nine patients (66.2%) demonstrated radiographic
21
evidence of glioma progression. Eight (10.8%) Grade II gliomas, 8 (10.4%) Grade III gliomas,
22
and 33 (44.6%) Grade IV gliomas (Figure 3).
23
Low Grade Gliomas Progression Free Survival:
ACCEPTED MANUSCRIPT Insular glioma extent of resection
10
For the LGGs, a univariate analysis identified EOR, IDH1 mutation, and 1p/19q co-
2
deletion as significant variables for PFS (Table 3). A multivariate Cox proportional analysis
3
showed the EOR, treated as a continuous variable, (HR 0.949, 95% CI 0.909-0.991, p = 0.017)
4
and the 1p19q co-deletion (HR 0.029, 95% CI 0.002-0.491, p = 0.014) were significant
5
predictors for PFS, while controlling for IDH1 mutation (Table 4). Patients with a ≥ 90% EOR
6
had a 5-year PFS rate of 100%, and patients with a < 90% EOR had 5 year PFS of 41.7% . When
7
stratifying the EOR into three groups ≥90%, 70-89%, and <70%, a stepwise improvement in PFS
8
was associated with increasing EOR among our LGG cohort (p=0.006, Figure 4).
SC
High Grade Gliomas Progression Free Survival:
M AN U
9
RI PT
1
For patients with HGGs, a Cox proportional hazards analysis shows a predictive
11
association between EOR and PFS (HR 0.971, 95% CI 0.946-0.996, p = 0.024, Table 5).
12
Patients with a ≥ 90% EOR had a 2-year PFS rate of 51.4%, and patients with a < 90% EOR had
13
2 year PFS of 33.3%. The stratification of EOR into ≥90%, 70-89%, and <70% showed
14
significant improvement in PFS with the Kaplan Meier curves (p=0.017, Figure 4).
TE D
10
Malignant PFS: Malignant transformation was seen in 6 (8.1%) patients with Grade II
16
gliomas. There was pathological evidence of malignant transformation in 5 of the patients and
17
radiographic evidence of transformation in 1 patient. The Cox proportional hazards analysis
18
showed that malignant progression was predicted by EOR (HR 0.958, 95% CI 0.926 -0.991, p =
19
0.014). The 5-year MPFS rate was 100% for EOR ≥90% and 57.7% for EOR <90% (Figure 5).
20
The IDH1 mutation (p=0.548) or the 1p/19q co-deletion (p=0.261) were not found to be
21
significant predictors of MPFS.
AC C
EP
15
22
23
Extent of resection predictors
ACCEPTED MANUSCRIPT Insular glioma extent of resection
11
1
A linear regression was used to evaluate factors that influenced the extent of resection for
2
insular gliomas (Table 6). Preoperative tumor volume was found to be a significant predictor of
3
extent of resection (R2=0.053, p= 0.048).
5
RI PT
4
Discussion
Insular gliomas have been shown to be distinct from other gliomas, where the
7
microenvironment of the insula gives rise to gliomas with unique growth patterns and anatomical
8
locations than other gliomas.8,13 With these intrinsic differences, it is important to understand the
9
factors that affect the prognosis for insular gliomas. This study provides the first volumetric
10
analysis of insular glioma resection that also analyzes the genetic characteristics of the low grade
11
gliomas in order to evaluate survival outcome. Our study finds that EOR is a significant predictor
12
of OS and PFS for high and low grade insular gliomas. For low grade gliomas, EOR significantly
13
predicts MPFS, and EOR and the 1p/19q co-deletion are significant predictors for OS and PFS.
14
We also find that the extent of resection is associated with preoperative tumor volume.
M AN U
TE D
16
Improving the Extent of resection
EP
15
SC
6
Given the eloquent anatomy in proximity to the insula, achieving an extensive resection
18
in this area can be difficult. Yasargil et al.31 presented an early study that showed the possibility
19
of removal of insular tumors with good neurological outcomes and substantial resections done in
20
most cases, although the exact extent of resection was not reported. Subsequent studies followed
21
that grouped extent of resections into categories of 'gross total' and 'subtotal' resection in an
22
effort to evaluate the efficacy of surgery in the insular region, although the results varied.
23
Vanaclocha et al.32 reported 23 patients with insular tumors where 20 (87%) had gross total
AC C
17
ACCEPTED MANUSCRIPT Insular glioma extent of resection
12
resections and 70% of the lesions were in the left hemisphere. In contrast, Zentner et al.19
2
reported gross total resection in 17% of 30 patients with insular tumors, subtotal in 70%, and
3
partial in 13%. The introduction of volumetric analysis for tumor resection allowed for more
4
precise evaluations of surgical outcome and survival. For insular gliomas extents of resection of
5
≥ 90% has been reported at a rate of 23-42%.23,24 Our study found similar results with 40% of
6
our cohort having a EOR ≥ 90%. For the high grade gliomas, we found a significantly improved
7
overall survival at an EOR ≥ 81%, while there was an significantly better PFS for at an EOR ≥
8
71%. Our low grade glioma cohort was not large enough to analyze a similar threshold EOR for
9
survival.
M AN U
SC
RI PT
1
In order to achieve a high EOR for insular tumors while maintaining low morbidity,
11
additional tools are need for surgery. Several instruments aid in improving the extent of resection
12
while minimizing the damage to surrounding structures. Preoperative MRI or CT anigiography
13
has been used to evaluate the relationship of lenticulostriate arteries to the insular tumor, to aid in
14
reducing the risk of violating the arteries during the tumor resection.2,33 Intraoperative functional
15
monitoring for dominant side lesions for transcortical approaches has also been used.13 Duffau et
16
al.2,17 used cortical stimulation during awake craniotomy to identify eloquent regions for his
17
transcortical approach for insular tumors and reported 16% total resections, with 61% subtotal,
18
and 23% partial. Additional assistive devices include fluorescence-guided resection,
19
neurophysiological monitoring, tractography, neuro-navigation, and ultrasound.34 Lang et al.33
20
utilized ultrasound to help guide their insular tumor resection and found that from 22 patients
21
45% had gross total resections, 27% had subtotal resections and 27% had partial resections. For
22
our cohort, we used preoperative MRI and neuro-navigation for operative planning,
AC C
EP
TE D
10
ACCEPTED MANUSCRIPT Insular glioma extent of resection
13
1
neurophysiological monitoring, and direct cortical mapping with awake craniotomy for dominant
2
hemispheric lesions that involved eloquent regions. The surgical approach that is chosen for an insular glioma can also affect the extent of
4
resection. Many approaches exist to resect insular tumors, including the transsylvian,
5
transcortical, and both approaches combined, but there is no consensus on which approach is
6
considered the safest or provides the maximal extent of resection.31,33,35-37 The transsylvian
7
approach can be associated with injury to the arteries and veins of the sylvian fissure which can
8
cause ischemia in the insular region. In addition, retraction of the opercular region during this
9
approach can also lead to neurological damage, such as dysphasia and short term memory
10
deficits.38,39 With the transcortical approach, motor function can be at risk, or for dominant
11
hemisphere lesions near Broca's or Wernicke's area, language can also be affected with this
12
approach. In a cadaveric study, Benet et al.5 showed that for approaches to the insula, the
13
transcortical approach can provide better exposure and surgical freedom for most insular tumors
14
compared to the transsylvian approach. The transcortical approach can avoid iatrogenic spasm of
15
branches of the MCA.13 Duffau et al.2 proposed a multi-staged approach for tumors that involved
16
the lateral lenticulostriate arteries or posterior segment of the insula, claiming that the initial
17
incomplete resection would allow for functional remapping of the insula so that the second
18
surgery would be safer.
SC
M AN U
TE D
EP
AC C
19
RI PT
3
Once the surgical approach has gained access to the insula, caution must be taken when
20
resecting at the insular tumor boundaries. Retraction of the frontal operculum (dominant side), to
21
obtain superficial tumor access, can lead to dysphasia with dysnomia, or damage to the
22
horizontal fibers of the arcuate fasciculus can cause conduction aphasia, impaired perception, or
23
short-term memory loss.39 Dissecting the medial aspect of the tumor can lead to damage to the
ACCEPTED MANUSCRIPT Insular glioma extent of resection
14
external capsule, leading to semantic paraphasias, or to the lateral lenticulostriate or long
2
perforators of M2 arteries, causing hemiparesis or hemiplegia.39 Dissecting the superior aspect of
3
the insular tumor may cause damage to the medial and lateral fibers of the corona radiata,
4
causing sensorimotor impairment to the lower and upper limbs.21,33 Our general practice is to use
5
the transcortical approach and direct cortical and subcortical stimulation mapping to avoid
6
eloquent regions.
RI PT
1
8
SC
7
Surgical Morbidity
The benefits of an extensive resection must be weighed against the risks of morbidities.
10
With insular glioma resections, injury to nearby cortical and subcortical structures can lead to
11
neurological deficits. Transient and permanent postoperative deficits following insular surgery
12
has ranged in the literature from 14-59% (transient) to 0 to 20% (permanent).2,13,24,32,33
13
Hemiparesis is a common deficit seen postoperatively and occurs in 2-37% of patients.2,19,23,31,32
14
Our study found that 10.8% of patients had new motor deficits with 2.7% being permanent
15
deficits. Motor deficits following insular surgery have been associated with worse
16
prognosis.40,41Maintaining good motor function postoperatively is important to help patients
17
maintain better performance status and help with overall recovery.
19
20
TE D
EP
AC C
18
M AN U
9
Factors for Survival
Few studies have produced the survival and prognostic assessments with extent of
21
resection for insular gliomas.13,24 Simon et al.24found that the 5-year overall and progression free
22
survival was 68 and 58% for grade II gliomas and 83 and 80% for anaplastic astrocytomas, as
23
well as determined that prognostic factors for OS and PFS were age, tumor pathology, Yasargil
ACCEPTED MANUSCRIPT Insular glioma extent of resection
15
type 5 tumor, and EOR. Sanai et al.13 incorporated volumetric analysis of tumor resection and
2
found that the 5-year overall survival was 100% for grade II gliomas for EOR >90%, and 84%
3
for EOR < 90%, while the 2 year overall survival for high grade gliomas was 91% in for EOR
4
>90% and 75% for EOR <90%. EOR was also a significant predictor of OS and PFS in the Sanai
5
et al.23 study. Both the Simon and Sanai studies presented informative results regarding insular
6
gliomas, however, neither study evaluated the genetic makeup of the tumors which has been
7
shown to also influence prognosis.42-45 Our study evaluated the IDH1 mutation and 1p/19q
8
codeletion for low grade gliomas in addition to EOR, in order to assess the prognostic roles of
9
each variable. We found that the 5-year overall survival was 100% for grade II gliomas for EOR
10
>90%, and 80% for EOR < 90%, while the 2 year overall survival for high grade gliomas was
11
83.7% for EOR >90% and 43.8% for EOR <90%. We also found that EOR was a significant
12
prognostic factor for OS, PFS, and MPFS for insular gliomas, and that for low grade gliomas
13
EOR and the 1p19q co-deletion significantly predicted PFS. Our study shows that the IDH1
14
mutation had a significant association with OS and PFS with the univariate analysis, as has also
15
been presented in the literature,44,45 however, for the multivariate analysis factoring in EOR and
16
1p/19q co-deletion, the IDH1 mutation was not found to be a significant predictor of OS and
17
PFS. This finding suggests that the extent of resection also plays an important role in the
18
prognosis of IDH1 mutation for low grade insular gliomas. Understanding the roles of the
19
insular glioma genetic makeup and extent of resection provides a robust understanding about
20
patient survival that can aid with prognosis and treatment for patients.
AC C
EP
TE D
M AN U
SC
RI PT
1
21
22
23
Patient Selection
ACCEPTED MANUSCRIPT Insular glioma extent of resection
16
With this new understanding about how EOR and molecular genetics play a role in the
2
prognosis for insular glioma patients, how do you choose a proper surgical candidate? Various
3
indications for patient selection have been proposed in the literature including whether there is
4
lenticulostriate arteries involvement, the characteristic of the tumor boundaries, and the cortical
5
involvement.21,23,37 Tumor borders that are well defined are more amenable for complete
6
resection than diffuse borders. Kawaguchi et al.14 experienced a 75.7% gross total resection for
7
sharp tumor borders that had distinct medial borders from the basal ganglia, while only a 19.6%
8
gross total resection for tumors with diffuse borders. Enhancement of the tumor can also be an
9
indication of abundant intratumoral vasculature and has been shown to have increased
10
postoperative neurological deficits, likely due to injury of lenticulostriate arteries that are in
11
proximity to the vascular tumor.14 Care must be taken for tumors that involve the lenticulostriate
12
arteries or the long insular artery, often located in the superior-posterior central insular sulcus, as
13
these tumors have been associated with poorer extent of resection and more postoperative
14
ischemic complications.14,46 We generally include a MR angiogram, or vascular imaging, with
15
our preoperative imaging for these tumors in order to fully assess the location of the arteries. By
16
preoperatively assessing the general locations or the MCA and feeding arteries, we can anticipate
17
regions of higher risk for vascular injury with intraoperative neuro-navigation during tumor
18
resection. The T1-weighted, with contrast, or T2-weighted MRI imaging can be used
19
intraoperatively with neuro-navigation to help identify nearby vasculature (Figure 6).
21
SC
M AN U
TE D
EP
AC C
20
RI PT
1
Limits of the study
22
Our study contains the usual constraints associated with a retrospective study. In addition,
23
this study evaluates a small cohort that contains a heterogeneous group of high and low grade
ACCEPTED MANUSCRIPT Insular glioma extent of resection
17
gliomas. The molecular marker evaluations are based on the low grade population, which adds
2
new insight to our understanding of insular gliomas, but will require larger studies to further
3
investigate this topic. Our institution evaluates the R132H mutation when testing for IDH
4
mutation, leaving the possibility that other IDH mutations could have associations with survival
5
that are not identified in this study.
RI PT
1
6
Conclusion
SC
7
This study suggests that for insular gliomas, EOR is a significant predictor of OS, PFS,
9
and MPFS. For low grade insular gliomas, EOR and the 1p/19q co-deletion were found to be
10
significant predictors of PFS. With an improved understanding about the role of genetics and
11
extent of resection on insular gliomas, a more effective strategy can be developed for providing
12
optimal treatment for insular gliomas.
EP
15
AC C
14
TE D
13
M AN U
8
ACCEPTED MANUSCRIPT Insular glioma extent of resection
18
References
2
1. Duffau H, Capelle L. Preferential brain locations of low-grade gliomas. Cancer. 2004;100(12):2622-
3
2626.
4
2. Duffau H. A personal consecutive series of surgically treated 51 cases of insular WHO grade II glioma:
5
Advances and limitations. J Neurosurg. 2009;110(4):696-708.
6
3. Saito R, Kumabe T, Kanamori M, Sonoda Y, Tominaga T. Insulo-opercular gliomas: Four different
7
natural progression patterns and implications for surgical indications. Neurol Med Chir (Tokyo).
8
2010;50(4):286-290.
9
4. Kumabe T, Higano S, Takahashi S, Tominaga T. Ischemic complications associated with resection of
M AN U
SC
RI PT
1
opercular glioma. J Neurosurg. 2007;106(2):263-269.
11
5. Benet A, Hervey-Jumper SL, Sanchez JJ, Lawton MT, Berger MS. Surgical assessment of the insula.
12
part 1: Surgical anatomy and morphometric analysis of the transsylvian and transcortical approaches to
13
the insula. J Neurosurg. 2016;124(2):469-481.
14
6. Hervey-Jumper SL, Li J, Osorio JA, et al. Surgical assessment of the insula. part 2: Validation of the
15
berger-sanai zone classification system for predicting extent of glioma resection. J Neurosurg.
16
2016;124(2):482-488.
17
7. Signorelli F, Guyotat J, Elisevich K, Barbagallo GM. Review of current microsurgical management of
18
insular gliomas. Acta Neurochir (Wien). 2010;152(1):19-26.
19
8. Kim YH, Kim CY. Current surgical management of insular gliomas. Neurosurg Clin N Am.
20
2012;23(2):199-206, vii.
AC C
EP
TE D
10
ACCEPTED MANUSCRIPT Insular glioma extent of resection
19
9. Isnard J, Guenot M, Sindou M, Mauguiere F. Clinical manifestations of insular lobe seizures: A stereo-
2
electroencephalographic study. Epilepsia. 2004;45(9):1079-1090.
3
10. Guenot M, Isnard J. Epilepsy and insula. Neurochirurgie. 2008;54(3):374-381.
4
11. Signorelli F, Guyotat J, Elisevich K, Barbagallo GM. Review of current microsurgical management of
5
insular gliomas. Acta Neurochir (Wien). 2010;152(1):19-26.
6
12. Tanriover N, Rhoton AL,Jr, Kawashima M, Ulm AJ, Yasuda A. Microsurgical anatomy of the insula
7
and the sylvian fissure. J Neurosurg. 2004;100(5):891-922.
8
13. Sanai N, Polley MY, Berger MS. Insular glioma resection: Assessment of patient morbidity, survival,
9
and tumor progression. J Neurosurg. 2010;112(1):1-9.
M AN U
SC
RI PT
1
14. Kawaguchi T, Kumabe T, Saito R, et al. Practical surgical indicators to identify candidates for radical
11
resection of insulo-opercular gliomas. J Neurosurg. 2014;121(5):1124-1132.
12
15. Pitskhelauri DI, Bykanov AE, Zhukov VY, Kachkov IA, Buklina SB, Tonoyan AS. Review of
13
surgical treatment of insular gliomas: Challenges and opportunities. Zh Vopr Neirokhir Im N N Burdenko.
14
2015;79(2):111-116.
15
16. Ebeling U, Kothbauer K. Circumscribed low grade astrocytomas in the dominant opercular and
16
insular region: A pilot study. Acta Neurochir (Wien). 1995;132(1-3):66-74.
17
17. Duffau H, Capelle L, Lopes M, Faillot T, Sichez JP, Fohanno D. The insular lobe: Physiopathological
18
and surgical considerations. Neurosurgery. 2000;47(4):801-10; discussion 810-1.
19
18. Shankar A, Rajshekhar V. Radiological and clinical outcome following stereotactic biopsy and
20
radiotherapy for low-grade insular astrocytomas. Neurol India. 2003;51(4):503-506.
AC C
EP
TE D
10
ACCEPTED MANUSCRIPT Insular glioma extent of resection
20
19. Zentner J, Meyer B, Stangl A, Schramm J. Intrinsic tumors of the insula: A prospective surgical study
2
of 30 patients. J Neurosurg. 1996;85(2):263-271.
3
20. Heffez DS. Stereotactic transsylvian, transinsular approach for deep-seated lesions. Surg Neurol.
4
1997;48(2):113-124.
5
21. Hentschel SJ, Lang FF. Surgical resection of intrinsic insular tumors. Neurosurgery. 2005;57(1
6
Suppl):176-83; discussion 176-83.
7
22. Moshel YA, Marcus JD, Parker EC, Kelly PJ. Resection of insular gliomas: The importance of
8
lenticulostriate artery position. J Neurosurg. 2008;109(5):825-834.
9
23. Sanai N, Berger MS. Intraoperative stimulation techniques for functional pathway preservation and
M AN U
SC
RI PT
1
glioma resection. Neurosurg Focus. 2010;28(2):E1.
11
24. Simon M, Neuloh G, von Lehe M, Meyer B, Schramm J. Insular gliomas: The case for surgical
12
management. J Neurosurg. 2009;110(4):685-695.
13
25. Cancer Genome Atlas Research Network, Brat DJ, Verhaak RG, et al. Comprehensive, integrative
14
genomic analysis of diffuse lower-grade gliomas. N Engl J Med. 2015;372(26):2481-2498.
15
26. Eseonu CI, Rincon-Torroella J, ReFaey K, et al. Awake craniotomy versus craniotomy under general
16
anesthesia for perirolandic gliomas: Evaluating perioperative complications and extent of resection.
17
Neurosurgery. Accepted September 2016.
18
27. Chaichana KL, Jusue-Torres I, Navarro-Ramirez R, et al. Establishing percent resection and residual
19
volume thresholds affecting survival and recurrence for patients with newly diagnosed intracranial
20
glioblastoma. Neuro Oncol. 2014;16(1):113-122.
AC C
EP
TE D
10
ACCEPTED MANUSCRIPT Insular glioma extent of resection
21
28. Chaichana KL, Jusue-Torres I, Lemos AM, et al. The butterfly effect on glioblastoma: Is volumetric
2
extent of resection more effective than biopsy for these tumors? J Neurooncol. 2014;120(3):625-634.
3
29. Eseonu CI, Rincon-Torroella J, ReFaey K, Quinones-Hinojosa A. The cost of brain surgery: Awake
4
versus asleep craniotomy for perirolandic region tumors. Neurosurgery. Accepted August 2016.
5
30. Wang Y, Wang Y, Fan X, et al. Putamen involvement and survival outcomes in patients with insular
6
low-grade gliomas. J Neurosurg. 2016:1-7.
7
31. Yasargil MG, von Ammon K, Cavazos E, Doczi T, Reeves JD, Roth P. Tumours of the limbic and
8
paralimbic systems. Acta Neurochir (Wien). 1992;118(1-2):40-52.
9
32. Vanaclocha V, Saiz-Sapena N, Garcia-Casasola C. Surgical treatment of insular gliomas. Acta
M AN U
SC
RI PT
1
Neurochir (Wien). 1997;139(12):1126-34; discussion 1134-5.
11
33. Lang FF, Olansen NE, DeMonte F, et al. Surgical resection of intrinsic insular tumors: Complication
12
avoidance. J Neurosurg. 2001;95(4):638-650.
13
34. Barbosa BJ, Dimostheni A, Teixeira MJ, Tatagiba M, Lepski G. Insular gliomas and the role of
14
intraoperative assistive technologies: Results from a volumetry-based retrospective cohort. Clin Neurol
15
Neurosurg. 2016;149:104-110.
16
35. Kumabe T, Nakasato N, Suzuki K, et al. Two-staged resection of a left frontal astrocytoma involving
17
the operculum and insula using intraoperative neurophysiological monitoring--case report. Neurol Med
18
Chir (Tokyo). 1998;38(8):503-507.
19
36. Morantz RA. Radiation therapy in the treatment of cerebral astrocytoma. Neurosurgery.
20
1987;20(6):975-982.
AC C
EP
TE D
10
ACCEPTED MANUSCRIPT Insular glioma extent of resection
22
37. Saito R, Kumabe T, Inoue T, et al. Magnetic resonance imaging for preoperative identification of the
2
lenticulostriate arteries in insular glioma surgery. technical note. J Neurosurg. 2009;111(2):278-281.
3
38. Schatz CR, Kreth FW, Faist M, Warnke PC, Volk B, Ostertag CB. Interstitial 125-iodine radiosurgery
4
of low-grade gliomas of the insula of reil. Acta Neurochir (Wien). 1994;130(1-4):80-89.
5
39. Duffau H, Peggy Gatignol ST, Mandonnet E, Capelle L, Taillandier L. Intraoperative subcortical
6
stimulation mapping of language pathways in a consecutive series of 115 patients with grade II glioma in
7
the left dominant hemisphere. J Neurosurg. 2008;109(3):461-471.
8
40. McGirt MJ, Mukherjee D, Chaichana KL, Than KD, Weingart JD, Quinones-Hinojosa A. Association
9
of surgically acquired motor and language deficits on overall survival after resection of glioblastoma
M AN U
SC
RI PT
1
multiforme. Neurosurgery. 2009;65(3):463-9; discussion 469-70.
11
41. Rahman M, Abbatematteo J, De Leo EK, et al. The effects of new or worsened postoperative
12
neurological deficits on survival of patients with glioblastoma. J Neurosurg. 2016:1-9.
13
42. Cairncross G, Wang M, Shaw E, et al. Phase III trial of chemoradiotherapy for anaplastic
14
oligodendroglioma: Long-term results of RTOG 9402. J Clin Oncol. 2013;31(3):337-343.
15
43. van den Bent MJ, Brandes AA, Taphoorn MJ, et al. Adjuvant procarbazine, lomustine, and vincristine
16
chemotherapy in newly diagnosed anaplastic oligodendroglioma: Long-term follow-up of EORTC brain
17
tumor group study 26951. J Clin Oncol. 2013;31(3):344-350.
18
44. Parsons DW, Jones S, Zhang X, et al. An integrated genomic analysis of human glioblastoma
19
multiforme. Science. 2008;321(5897):1807-1812.
20
45. Yan H, Parsons DW, Jin G, et al. IDH1 and IDH2 mutations in gliomas. N Engl J Med.
21
2009;360(8):765-773.
AC C
EP
TE D
10
ACCEPTED MANUSCRIPT Insular glioma extent of resection
23
46. Iwasaki M, Kumabe T, Saito R, et al. Preservation of the long insular artery to prevent postoperative
2
motor deficits after resection of insulo-opercular glioma: Technical case reports. Neurol Med Chir
3
(Tokyo). 2014;54(4):321-326.
RI PT
1
4
AC C
EP
TE D
M AN U
SC
5
ACCEPTED MANUSCRIPT Insular glioma extent of resection
24
1
Figure Legend:
2
Figure 1: Kaplan-Meier curves revealing the overall survival in patients based on glioma pathology grade. Patients with low grade pathology had a significantly longer overall survival (p=0.0034).
9 10 11 12 13 14 15 16 17 18 19
RI PT
8
Figure 3: Kaplan-Meier curves revealing the progression free survival in patients based on glioma pathology grade. Patients with a lower pathology grade had a significantly longer progression free survival (p<0.0001).
SC
7
Figure 2: Kaplan-Meier curves showing the overall survival in patients with Grade II (left) and Grade III/IV (right) insular gliomas, stratified by Extent of resection (EOR). For the purposes of visual display, patients were grouped by EOR ≥90, 70-89%, and less than 70%.
Figure 4: Kaplan-Meier curves showing the progression free survival in patients with Grade II (left) and Grade III/IV (right) insular gliomas, stratified by Extent of resection (EOR). For the purposes of visual display, patients were grouped by EOR ≥90, 70-89%, and less than 70%.
M AN U
6
Figure 5: Kaplan-Meier curves revealing the malignant progression free survival in patients with Grade II insular gliomas, stratified by Extent of resection (EOR). For the purposes of visual display, patients were grouped by EOR ≥90 and less than 90%. Figure 6: (Left) Coronal MRA and (Right) T1-weighted MRI with gadolinium contrast depicting an insular glioma with nearby MCA and feeding arteries.
TE D
5
EP
4
AC C
3
ACCEPTED MANUSCRIPT
Table 1: Summary of patient and disease characteristics of 74 insular glioma patients No. (%)
RI PT
54 18-80 42 (56.8) 32 (43.2)
SC
80 50-100
AC C
EP
26 (35.1) 14 (18.9) 17 (23.0) 5 (6.8) 12 (16.2)
M AN U
TE D
Characteristic Age (years) median Range Gender Male Female Preoperative KPS median range Preoperative symptoms Seizure Headache Motor deficit Sensory deficit Language deficit Side Left Right WHO tumor classification II III IV Postop chemotherapy Postop radiation Putamen involvement Yes No Preop Tumor Volume (cm3) median range Postop Tumor Volume (cm3) median range EOR ≥ 90% 70-89% < 70% median % Range % IDH1* Mutation Wild type 1p/19q*
43 (58.1) 31 (41.9) 25 (33.8) 16 (21.6) 33 (44.6) 53 (71.6) 33 (44.6) 35 (47.3) 39 (52.7) 43.6 2.2-256.4
3.7 0-109.4 40 (54.0) 15 (20.3) 19 (25.7) 91.7 10.6 - 100 19 (76.0) 6 (24.0)
ACCEPTED MANUSCRIPT
18 (72.0) 7 (28.0) 4.4 1.2-10.1
AC C
EP
TE D
M AN U
SC
RI PT
Codeletion Intact Clinical follow-up (yrs) median range KPS: Karnofsky Performance Status; * From 25 low grade tumors
ACCEPTED MANUSCRIPT
Table 2: Postoperative neurological outcomes following insular glioma surgery Characteristic Postop KPS median range Postop Seizure New Motor Deficit Transient Deficit Permanent Deficit New Sensory Deficit New Language Deficit Transient Deficit, n (%) Permanent Deficit, n (%) Wound Infection, n (%) Meningitis, n (%) Length of Hospital Stay median range Disposition
No (%)
M AN U
SC
RI PT
80 50-100 9 (16.0) 8 (10.8) 6 (8.1) 2 (2.7) 3 (4.1) 5 (6.8) 5 (6.8) 0 (0.0) 0 (0.0) 0 (0.0) 4.5 1-22
50 (67.6) 24 (32.4)
Discharged to Home
AC C
EP
TE D
Discharged to Rehab
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 3: Univariate analyses of survival outcomes in patients with insular low grade gliomas*. EOR, IDH1 mutation, and 1p19q codeletion are significant factors for OS and PFS. Factor p Value, PFS p Value, OS p Value, MPFS Extent of resection 0.028 0.054 0.014 Age 0.349 0.278 0.919 Sex 0.627 0.997 0.932 Preoperative seizures 0.354 0.922 0.896 Preoperative KPS 0.863 0.556 0.914 Left Side tumor 0.212 0.811 0.375 Preoperative Tumor 0.998 0.176 0.314 Volume Putamen involvement 0.368 0.715 0.185 IDH1 mutation 0.548 0.076 0.093 1p19q codeletion 0.261 0.039 0.095 KPS, Karnofsky Performance Status, *Red indicates significance (p<0.15)
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 4: Multivariate analysis (Cox proportional hazards model) of survival outcomes for low grade gliomas*. EOR and 1p19q co-deletion are significant factors for PFS in the multivariable analysis, while EOR is a significant factor for OS in the multivariable analysis. PFS OS Factor p Value HR 95% CI p Value HR 95% CI Extent of Resection 0.017 0.949 0.909-0.991 0.882 0.783-0.994 0.039 IDH1 mutation 0.081 0.163 0.021-1.251 0.738 0.088 0.054-1.531 1p19q codeletion 0.029 0.002-0.491 0.546 1.602 0.001-1.993 0.014 *Red indicates significance (p<0.05)
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 5: Univariate analyses of survival outcomes in patients with insular high grade gliomas*. EOR is a significant factor for PFS and OS. Factor p Value, PFS p Value, OS Extent of resection 0.024 0.020 Age 0.412 0.287 Sex 0.577 0.316 Preoperative seizures 0.804 0.826 Preoperative KPS 0.394 0.229 Left Side tumor 0.811 0.385 Preoperative Tumor Volume 0.913 0.936 Putamen involvement 0.723 0.360 Oligodendroglioma 0.702 0.882 KPS, Karnofsky Performance Status, *Red indicates significance (p<0.15)
ACCEPTED MANUSCRIPT
R2 0.011 0.025 0.007 0.053 < 0.001 0.037 <0.001 0.015 0.010
P-value 0.375 0.181 0.492 0.048 0.818 0.100 0.985 0.268 0.327
SC
Factor Age Gender Left side tumor Preoperative Volume Putamen involvement Pathology Berger-Sanai Zone I IDH1 mutation* 1p/19q codeletion*
RI PT
Table 6: Linear Regression for predictors of EOR. Preoperative tumor volume was found to be a significant predictor of EOR for insular gliomas
AC C
EP
TE D
M AN U
*For low grade gliomas
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Highlights: - Extent of tumor resection and molecular makeup of an insular glioma are two important factors that influence survival
RI PT
- Extent of resection (EOR) is a significant predictor of overall and progression free survival for high and low grade insular gliomas. - For low grade gliomas, EOR significantly predicts malignant progression free survival, and EOR and the 1p/19q co-deletion are significant predictors for overall and progression free survival.
AC C
EP
TE D
M AN U
SC
- Preoperative tumor volume is a significant factor that influences the extent of resection for insular gliomas.
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Abbreviation list LGG - Low grade glioma HGG - High grade glioma EOR - Extent of resection LOS - Length of stay KPS - Karnofsky performance Status
S1878-8750(17)30496-5
DOI:
10.1016/j.wneu.2017.04.002
Reference:
WNEU 5518
To appear in:
World Neurosurgery
Received Date: 24 January 2017 Revised Date:
30 March 2017
Accepted Date: 1 April 2017
Please cite this article as: Eseonu CI, ReFaey K, Garcia O, Raghuraman G, Quinones-Hinojosa A, Volumetric analysis of extent of resection, survival, and surgical outcomes, for insular gliomas, World Neurosurgery (2017), doi: 10.1016/j.wneu.2017.04.002. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Volumetric analysis of extent of resection, survival, and surgical outcomes, for insular gliomas
1 2 3 4 5
Authors: Chikezie I. Eseonu, MDa, Karim ReFaey, MDa,b, Oscar Garcia, MPHa,b, Gugan Raghuramana, Alfredo Quinones-Hinojosa, MDa,b
6
9 10
a
Department of Neurological Surgery and Oncology Outcomes Lab. Johns Hopkins University, Baltimore, Maryland b Neurological Surgery Department, Mayo Clinic, Jacksonville, Florida
11 12 13 14
Conflict of Interest: There is no conflict of interests to report or financial disclosures related to this manuscript
15
Running Title: Insular glioma extent of resection
17 18
M AN U
16
SC
8
RI PT
7
Keywords: Insula, glioma, low grade, high grade, insular tumors, survival
19
Corresponding author:
21
Alfredo Quiñones-Hinojosa, M.D.
22
Brain Tumor Stem Cell Laboratory
23
Department of Neurologic Surgery
24
Mayo Clinic, Florida
25
4500 San Pablo Rd. S
26
Jacksonville, FL 32224
27
Phone: 410-502-2906 / fax: 410-502-5559 / email: [email protected]
30 31 32 33 34 35 36 37 38 39 40 41 42 43
EP
29
AC C
28
TE D
20
ACCEPTED MANUSCRIPT Insular glioma extent of resection
1
2
Abstract
2 3 4 5 6
Background Insular gliomas are challenging tumors to surgically resect due to the anatomy surrounding them. This study evaluates the role of extent of resection(EOR) and molecular markers on surgical outcome and survival for insular gliomas.
8 9 10 11 12
RI PT
7
Methods Seventy-four patients who had undergone an initial resection for an insular glioma by the same surgeon from 2006 to 2016 were analyzed. Low(grade II) and high(grade III/IV) grade gliomas were analyzed for the prognostic role of volumetric EOR and molecular markers on patient survival outcomes.
15 16 17 18 19 20 21 22 23 24
Results The cohort includes 25 low grade gliomas (LGGs) patients(33.8%), and 49 high grade glioma(HGGs) patients(66.2%). The median EOR was 91.7%(range 10-100%). New permanent postoperative deficits were found in 2.7% of patients. LGG patients with a ≥90% EOR had a 5year survival rate of 100% and patients with a <90% EOR had 5-year survival of 80%. HGG patients with a ≥90% EOR had a 2-year survival rate of 83.7%, and patients with a <90% EOR had 2-year survival of 43.8%. For LGGs, the EOR was predictive of OS(p=0.017), progression free survival(PFS, p=0.039), and malignant progression free survival(MPFS, p=0.014), while the 1p/19q co-deletion was predictive for PFS(p=0.014). For HGGs, the EOR was predictive of OS(p=0.020) and PFS(p=0.024). Preoperative tumor volume was a factor that most significantly affected the EOR for insular gliomas(R2=0.053,p=0.048).
M AN U
14
SC
13
25
29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
TE D
28
Conclusion Extensive resections of insular gliomas can be achieved with low morbidity and can improve OS and PFS. In this series of low-grade gliomas, EOR was associated with longer MPFS, and the 1p/19q co-deletion was predictive of PFS.
EP
27
AC C
26
ACCEPTED MANUSCRIPT Insular glioma extent of resection
1
3
Introduction Insular gliomas are the most common intrinsic tumor of the insular cortex and make up
3
around 25% of all low-grade gliomas and 10% of high grade gliomas.1-3 Tumors in this cortical
4
area are surrounded by eloquent brain that controls motor and language function. In addition, the
5
vascular supply for the descending motor pathway also passes through the insular region.4-6
6
Patients with insular tumors can often present with devastating intractable epilepsy or motor
7
dysphasia making it essential to properly treat these lesions.7-10
SC
RI PT
2
Due to the deep location in the Sylvian fissure and complex anatomy that surrounds the
9
area, insular tumors are a challenging tumor to resect.11,12 The standard of care for insular
10
gliomas includes a maximal safe resection followed by chemoradiation therapy.13 Surgery in the
11
insular region may require opening the Sylvian fissure for insular exposure which puts vital
12
functional cortical areas, motor and language subcortical fibers, and vascular structures at risk in
13
this region.12,14 There is a high risk of neurological deficits following resection of insular tumors
14
given the proximity of the insula to vital vascular and neural structures, and severe speech and
15
motor disorders can result in the postoperative period.15
TE D
M AN U
8
Given the high risk of surgery, some authors have reported that insular tumors are
17
inoperable, opting for stereotactic biopsies for diagnosis and then chemotherapy and or
18
radiotherapy as an alternative treatment.2,16-18 However, other authors have reported favorable
19
reports with insular tumor resections based on specialized microsurgical techniques and an
20
understanding of the insular anatomy.13,17,19-22 Despite the high risks of surgery in the insular
21
region, there are limited studies that evaluate the role of the extent of resection (EOR), analyzed
22
volumetrically, on patient outcomes and survival for insular gliomas.23,24 Additionally, with an
23
increasing understanding about glioma genetics and their role in prognosis for low grade gliomas
AC C
EP
16
ACCEPTED MANUSCRIPT Insular glioma extent of resection
4
1
(isocitrate dehydrogenase 1/2 [IDH1/2] mutation and 1p/19q co-deletion), no previous
2
volumetric analysis for insular gliomas have accounted for these variables when evaluating the
3
role of EOR on patient survival outcome.25 In this study, we present a cohort of insular gliomas operated by a single surgeon at a
5
single center and evaluate the role of EOR and molecular markers on patient outcome and
6
survival.
RI PT
4
SC
7
8
10
Methods
M AN U
9
Patient Selection
Seventy-four consecutive patients with insular gliomas resected for the first time by a
12
single surgeon between January 2006 and July 2016 were prospectively collected and
13
retrospectively reviewed. Patients were ≥ 18 years old. Pathology assessment of the tumor was
14
based on WHO guidelines. Patients with gliomatosis cerebri were not included in this study.
15
Clinical data was collected from the patient chart review.
16
18
Surgical approach
All insular tumors were approached via a transcortical approach. The location and side of
AC C
17
EP
TE D
11
19
the hemisphere of the tumor determined whether the tumor was resected under general
20
anesthesia (45 cases) or by awake craniotomy (29 cases). Our method for awake anesthesia has
21
been previously described.26 After the craniotomy was performed and dura opened, cortical and
22
subcortical sensorimotor mapping was conducted. Once functional regions were identified,
23
corticectomies above and below the sylvian fissure in noneloquent regions were made allowing
ACCEPTED MANUSCRIPT Insular glioma extent of resection
5
for tumor resection around the sylvian vessels. Cortical mapping of the medial border of the
2
tumor was also conducted if needed to identify the internal capsule. The goal of all surgeries in
3
this cohort was for a safe, 100% total resection, however, tumor invasion into areas deemed
4
eloquent for motor, sensory, or language areas by intraoperative cortical/subcortical mapping, or
5
extensive involvement of the middle cerebral artery within the tumor, limited the extent of
6
resection.
RI PT
1
8
SC
7
Patient Outcome Data
Patients were examined preoperatively, immediately postoperatively, and during their
10
follow up visits by the senior surgeon (AQH). Neurological deficits were determined by
11
identifying new or worsening deficits related to motor, sensory, or language function. Outpatient
12
examinations were conducted at 1 month, 6 months, and then annually following surgery.
13
Transient deficits were considered new or worsening neurological deficits following surgery that
14
resolved prior to the 6-month follow up clinic visit. Permanent deficits were neurological deficits
15
that persisted beyond the 6-month follow up clinic visit. The length of hospitalization was
16
calculated based on the duration of hospital stay from the day of surgery until the day of
17
discharge.
TE D
EP
Survival data was also analyzed in the form of overall survival (OS), progression free
AC C
18
M AN U
9
19
survival (PFS), and malignant progression free survival (MPFS). Overall survival was the time
20
between initial surgery and death. PFS was the time from initial surgery to unequivocal
21
demonstration of tumor re-growth on follow up imaging. MPFS was defined as the cases where
22
grade II gliomas, determined from pathology from the initial surgery, were found to transform
23
into a higher grade lesion, via histopathology from a subsequent resection/biopsy, or
ACCEPTED MANUSCRIPT Insular glioma extent of resection
6
1
enhancement from follow up imaging. Patients with no progression of their tumor were censored
2
as of their last documented imaging date.
3
Tumor Volumetric Analysis
RI PT
4
The preoperative tumor volume was determined by using the T1-weighted with
6
gadolinium or T2-weighted and fluid-attenuated inversion recovery (FLAIR) MRI (1.5 to 3 mm
7
axial cuts) depending on the tumor type. The OsiriX software (Pixmeo) was used to quantify
8
tumor volume by two clinicians with neuroscience and radiology training who were blinded to
9
the patient information, as we have previously described.26-29 If the inter-observer volume
10
difference was large, then a third clinician, also blinded to the cohorts, would evaluate the
11
volumes, and the mean volume was determined from the two similarly calculated volumes. The
12
postoperative tumor volume was calculated using the MRI images obtained 48 hours following
13
surgery, in the same manner as the preoperative imaging. The extent of resection was calculated
14
with the equation (preoperative - postoperative tumor volume)/preoperative tumor volume.
15
Statistical Analysis
EP
16
TE D
M AN U
SC
5
Univariate statistics were conducted to produce descriptive statistics that were reported as
18
number of patient subjects and percent for categorical variables, median and range for
19
continuous nonparametric variables, and mean and standard deviations for continuous parametric
20
variables. Kaplan-Meier curves were used to generate OS, PFS, and MPFS for all patients. The
21
Kaplan-Meier curves plotted EOR and OS or PFS, using the ≥ 90, 70-89, and <70% EOR groups
22
to maintain consistency with previous insular glioma studies.24,30 For the MPFS, the EOR for the
23
Kaplan-Meier curve was divided into ≥ 90 and < 90% as this has been a common grouping when
AC C
17
ACCEPTED MANUSCRIPT Insular glioma extent of resection
7
assessing the impact of EOR.13 Differences were assessed for statistical significance using the
2
log-rank test. A univariate Cox proportional hazards models were used to determine predictors of
3
survival curves. For survival risk factors, variables that had a p-value <0.15 on the univariate
4
analysis were entered into the multivariate analysis. The multivariate Cox proportional hazards
5
models required a p < 0.05 for significance to be achieved. EOR was treated as a continuous
6
variable in the Cox proportional hazards models to determine if it predicted OS, PFS, MPFS. A
7
linear regression was used to evaluate variables associated with EOR. A biostatistician (O.G)
8
contributed to the statistical analysis. STATA 14 was used for all statistical analyses. The
9
university institutional review board (IRB) approved this study.
M AN U
SC
RI PT
1
10
11
Result
12
Patient Characteristics
The patient cohort included 42 males and 32 females, with the median age of 54 years
14
(range 18-80 years). Tumors were in the left hemisphere in 58.1% of the patients, and awake
15
craniotomy in 39.2% of the cases. Patients most frequently presented with preoperative seizures
16
(35.1%). Grade II gliomas were found in 33.8% of patients, Grade III for 21.6% of patients, and
17
Grade IV in 44.6% of patients. Fifty-three patients had postoperative chemotherapy (71.6%) and
18
33 patients (44.6%) had postoperative radiation.
20
EP
AC C
19
TE D
13
Tumor Characteristics
21
The median preoperative tumor volume was 43.6 cm3 and the median postoperative
22
tumor volume was 3.7 cm3. In this series, according to the Berger-Sanai zone classification for
23
insular tumors, there were 21 Zone I tumors (28.4%), 6 Zone II (8.1%), 7 Zone III (9.5%), 2
ACCEPTED MANUSCRIPT Insular glioma extent of resection
8
Zone IV (2.7%), 7 Zone I+II (9.5%), 11 Zone I+IV (14.9%), 8 Zone II+ III (10.8%), 2 Zone
2
III+IV (2.7%) and 10 giant insular tumors that involved all zones (13.5%).13 Following the first
3
surgery, 19 patients (25.7%) had a EOR of less than 70%, 15 patients (20.3%) had a EOR
4
between 70% and 89%, and 40 patients (54%) had an EOR of 90% or greater. For the patients
5
that had less than 100% total resection (61 patients), 85.2% of the resections were limited by
6
tumor invasion into eloquent cortical/subcortical regions determined by intraoperative mapping,
7
while 14.8% were limited by extensive tumor involvement with the middle cerebral artery or
8
feeding branches of the MCA. For the low grade gliomas, 19 patients (76%) had the IDH1
9
mutation, and 18 patients (72%) had a 1p/19q codeletion (Table 1).
M AN U
SC
RI PT
1
10
11
Postoperative outcome
At the six-month follow up, postoperative seizures were seen in 16% of patients.
13
Transient motor deficits were seen in 8.1% of patients, while permanent deficits were seen in
14
2.7% of patients. From the motor deficit patients, ischemic injury was identified in one patient
15
with a permanent deficit based on postoperative diffusion-weighted imaging. New sensory and
16
language deficits were found in 4.1% and 6.8% of patients, respectively. No ischemic injury was
17
identified on postoperative imaging in any patient with postoperative sensory or language
18
deficits. The median length of hospitalization was 4.5 days (range 1-22 days) (Table 2).
20
EP
AC C
19
TE D
12
Survival Analysis
21
Overall Survival: At the last follow up there were a total of 16 (21.6%) patients that died
22
from their disease. The median follow up time of 4.4 years in surviving patients. Death occurred
ACCEPTED MANUSCRIPT Insular glioma extent of resection
9
1
in 3 (4.1%) Grade II gliomas, 6 (8.1%) Grade III gliomas, and 14 (18.9%) Grade IV gliomas
2
(Figure 1). Low Grade Gliomas Overall Survival:
4
For the Low grade (grade II) gliomas (LGGs), a univariate Cox proportional analysis
5
showed EOR, IDH1 mutation, and 1p19q deletion as significant predictors for OS and PFS
6
(Table 3). A multivariate Cox proportional analysis showed the EOR, treated as a continuous
7
variable, had a significant association with OS, while controlling for IDH1 mutation and
8
presence of the 1p19q codeletion, (HR 0.882, 95% CI 0.783-0.994, p= 0.039) (Table 4). Patients
9
with a ≥ 90% EOR had a 5-year survival rate of 100% and patients with a < 90% EOR had 5
10
year survival of 80%. For graphic display, survival was categorized into three groups ≥90%, 70-
11
89%, and <70%. A stepwise improvement in OS was associated with increasing EOR among our
12
LGG cohort (p= 0.041, Figure 2).
M AN U
SC
RI PT
3
High Grade Gliomas Overall Survival:
14
For patients with high grade gliomas (HGGs), Grade III and Grade IV gliomas, a Cox
15
proportional hazards analysis shows a predictive association between EOR and OS (HR 0.974,
16
95% CI 0.953-0.996, p= 0.020, Table 5). Patients with a ≥ 90% EOR had a 2-year survival rate
17
of 83.7%, and patients with a < 90% EOR had 2 year survival of 43.8%. The stratification of
18
EOR into ≥90%, 70-89%, and <70% showed significant improvement in OS with the Kaplan
19
Meier curves (p=0.0165, Figure 2).
EP
AC C
20
TE D
13
Progression free survival: Forty-nine patients (66.2%) demonstrated radiographic
21
evidence of glioma progression. Eight (10.8%) Grade II gliomas, 8 (10.4%) Grade III gliomas,
22
and 33 (44.6%) Grade IV gliomas (Figure 3).
23
Low Grade Gliomas Progression Free Survival:
ACCEPTED MANUSCRIPT Insular glioma extent of resection
10
For the LGGs, a univariate analysis identified EOR, IDH1 mutation, and 1p/19q co-
2
deletion as significant variables for PFS (Table 3). A multivariate Cox proportional analysis
3
showed the EOR, treated as a continuous variable, (HR 0.949, 95% CI 0.909-0.991, p = 0.017)
4
and the 1p19q co-deletion (HR 0.029, 95% CI 0.002-0.491, p = 0.014) were significant
5
predictors for PFS, while controlling for IDH1 mutation (Table 4). Patients with a ≥ 90% EOR
6
had a 5-year PFS rate of 100%, and patients with a < 90% EOR had 5 year PFS of 41.7% . When
7
stratifying the EOR into three groups ≥90%, 70-89%, and <70%, a stepwise improvement in PFS
8
was associated with increasing EOR among our LGG cohort (p=0.006, Figure 4).
SC
High Grade Gliomas Progression Free Survival:
M AN U
9
RI PT
1
For patients with HGGs, a Cox proportional hazards analysis shows a predictive
11
association between EOR and PFS (HR 0.971, 95% CI 0.946-0.996, p = 0.024, Table 5).
12
Patients with a ≥ 90% EOR had a 2-year PFS rate of 51.4%, and patients with a < 90% EOR had
13
2 year PFS of 33.3%. The stratification of EOR into ≥90%, 70-89%, and <70% showed
14
significant improvement in PFS with the Kaplan Meier curves (p=0.017, Figure 4).
TE D
10
Malignant PFS: Malignant transformation was seen in 6 (8.1%) patients with Grade II
16
gliomas. There was pathological evidence of malignant transformation in 5 of the patients and
17
radiographic evidence of transformation in 1 patient. The Cox proportional hazards analysis
18
showed that malignant progression was predicted by EOR (HR 0.958, 95% CI 0.926 -0.991, p =
19
0.014). The 5-year MPFS rate was 100% for EOR ≥90% and 57.7% for EOR <90% (Figure 5).
20
The IDH1 mutation (p=0.548) or the 1p/19q co-deletion (p=0.261) were not found to be
21
significant predictors of MPFS.
AC C
EP
15
22
23
Extent of resection predictors
ACCEPTED MANUSCRIPT Insular glioma extent of resection
11
1
A linear regression was used to evaluate factors that influenced the extent of resection for
2
insular gliomas (Table 6). Preoperative tumor volume was found to be a significant predictor of
3
extent of resection (R2=0.053, p= 0.048).
5
RI PT
4
Discussion
Insular gliomas have been shown to be distinct from other gliomas, where the
7
microenvironment of the insula gives rise to gliomas with unique growth patterns and anatomical
8
locations than other gliomas.8,13 With these intrinsic differences, it is important to understand the
9
factors that affect the prognosis for insular gliomas. This study provides the first volumetric
10
analysis of insular glioma resection that also analyzes the genetic characteristics of the low grade
11
gliomas in order to evaluate survival outcome. Our study finds that EOR is a significant predictor
12
of OS and PFS for high and low grade insular gliomas. For low grade gliomas, EOR significantly
13
predicts MPFS, and EOR and the 1p/19q co-deletion are significant predictors for OS and PFS.
14
We also find that the extent of resection is associated with preoperative tumor volume.
M AN U
TE D
16
Improving the Extent of resection
EP
15
SC
6
Given the eloquent anatomy in proximity to the insula, achieving an extensive resection
18
in this area can be difficult. Yasargil et al.31 presented an early study that showed the possibility
19
of removal of insular tumors with good neurological outcomes and substantial resections done in
20
most cases, although the exact extent of resection was not reported. Subsequent studies followed
21
that grouped extent of resections into categories of 'gross total' and 'subtotal' resection in an
22
effort to evaluate the efficacy of surgery in the insular region, although the results varied.
23
Vanaclocha et al.32 reported 23 patients with insular tumors where 20 (87%) had gross total
AC C
17
ACCEPTED MANUSCRIPT Insular glioma extent of resection
12
resections and 70% of the lesions were in the left hemisphere. In contrast, Zentner et al.19
2
reported gross total resection in 17% of 30 patients with insular tumors, subtotal in 70%, and
3
partial in 13%. The introduction of volumetric analysis for tumor resection allowed for more
4
precise evaluations of surgical outcome and survival. For insular gliomas extents of resection of
5
≥ 90% has been reported at a rate of 23-42%.23,24 Our study found similar results with 40% of
6
our cohort having a EOR ≥ 90%. For the high grade gliomas, we found a significantly improved
7
overall survival at an EOR ≥ 81%, while there was an significantly better PFS for at an EOR ≥
8
71%. Our low grade glioma cohort was not large enough to analyze a similar threshold EOR for
9
survival.
M AN U
SC
RI PT
1
In order to achieve a high EOR for insular tumors while maintaining low morbidity,
11
additional tools are need for surgery. Several instruments aid in improving the extent of resection
12
while minimizing the damage to surrounding structures. Preoperative MRI or CT anigiography
13
has been used to evaluate the relationship of lenticulostriate arteries to the insular tumor, to aid in
14
reducing the risk of violating the arteries during the tumor resection.2,33 Intraoperative functional
15
monitoring for dominant side lesions for transcortical approaches has also been used.13 Duffau et
16
al.2,17 used cortical stimulation during awake craniotomy to identify eloquent regions for his
17
transcortical approach for insular tumors and reported 16% total resections, with 61% subtotal,
18
and 23% partial. Additional assistive devices include fluorescence-guided resection,
19
neurophysiological monitoring, tractography, neuro-navigation, and ultrasound.34 Lang et al.33
20
utilized ultrasound to help guide their insular tumor resection and found that from 22 patients
21
45% had gross total resections, 27% had subtotal resections and 27% had partial resections. For
22
our cohort, we used preoperative MRI and neuro-navigation for operative planning,
AC C
EP
TE D
10
ACCEPTED MANUSCRIPT Insular glioma extent of resection
13
1
neurophysiological monitoring, and direct cortical mapping with awake craniotomy for dominant
2
hemispheric lesions that involved eloquent regions. The surgical approach that is chosen for an insular glioma can also affect the extent of
4
resection. Many approaches exist to resect insular tumors, including the transsylvian,
5
transcortical, and both approaches combined, but there is no consensus on which approach is
6
considered the safest or provides the maximal extent of resection.31,33,35-37 The transsylvian
7
approach can be associated with injury to the arteries and veins of the sylvian fissure which can
8
cause ischemia in the insular region. In addition, retraction of the opercular region during this
9
approach can also lead to neurological damage, such as dysphasia and short term memory
10
deficits.38,39 With the transcortical approach, motor function can be at risk, or for dominant
11
hemisphere lesions near Broca's or Wernicke's area, language can also be affected with this
12
approach. In a cadaveric study, Benet et al.5 showed that for approaches to the insula, the
13
transcortical approach can provide better exposure and surgical freedom for most insular tumors
14
compared to the transsylvian approach. The transcortical approach can avoid iatrogenic spasm of
15
branches of the MCA.13 Duffau et al.2 proposed a multi-staged approach for tumors that involved
16
the lateral lenticulostriate arteries or posterior segment of the insula, claiming that the initial
17
incomplete resection would allow for functional remapping of the insula so that the second
18
surgery would be safer.
SC
M AN U
TE D
EP
AC C
19
RI PT
3
Once the surgical approach has gained access to the insula, caution must be taken when
20
resecting at the insular tumor boundaries. Retraction of the frontal operculum (dominant side), to
21
obtain superficial tumor access, can lead to dysphasia with dysnomia, or damage to the
22
horizontal fibers of the arcuate fasciculus can cause conduction aphasia, impaired perception, or
23
short-term memory loss.39 Dissecting the medial aspect of the tumor can lead to damage to the
ACCEPTED MANUSCRIPT Insular glioma extent of resection
14
external capsule, leading to semantic paraphasias, or to the lateral lenticulostriate or long
2
perforators of M2 arteries, causing hemiparesis or hemiplegia.39 Dissecting the superior aspect of
3
the insular tumor may cause damage to the medial and lateral fibers of the corona radiata,
4
causing sensorimotor impairment to the lower and upper limbs.21,33 Our general practice is to use
5
the transcortical approach and direct cortical and subcortical stimulation mapping to avoid
6
eloquent regions.
RI PT
1
8
SC
7
Surgical Morbidity
The benefits of an extensive resection must be weighed against the risks of morbidities.
10
With insular glioma resections, injury to nearby cortical and subcortical structures can lead to
11
neurological deficits. Transient and permanent postoperative deficits following insular surgery
12
has ranged in the literature from 14-59% (transient) to 0 to 20% (permanent).2,13,24,32,33
13
Hemiparesis is a common deficit seen postoperatively and occurs in 2-37% of patients.2,19,23,31,32
14
Our study found that 10.8% of patients had new motor deficits with 2.7% being permanent
15
deficits. Motor deficits following insular surgery have been associated with worse
16
prognosis.40,41Maintaining good motor function postoperatively is important to help patients
17
maintain better performance status and help with overall recovery.
19
20
TE D
EP
AC C
18
M AN U
9
Factors for Survival
Few studies have produced the survival and prognostic assessments with extent of
21
resection for insular gliomas.13,24 Simon et al.24found that the 5-year overall and progression free
22
survival was 68 and 58% for grade II gliomas and 83 and 80% for anaplastic astrocytomas, as
23
well as determined that prognostic factors for OS and PFS were age, tumor pathology, Yasargil
ACCEPTED MANUSCRIPT Insular glioma extent of resection
15
type 5 tumor, and EOR. Sanai et al.13 incorporated volumetric analysis of tumor resection and
2
found that the 5-year overall survival was 100% for grade II gliomas for EOR >90%, and 84%
3
for EOR < 90%, while the 2 year overall survival for high grade gliomas was 91% in for EOR
4
>90% and 75% for EOR <90%. EOR was also a significant predictor of OS and PFS in the Sanai
5
et al.23 study. Both the Simon and Sanai studies presented informative results regarding insular
6
gliomas, however, neither study evaluated the genetic makeup of the tumors which has been
7
shown to also influence prognosis.42-45 Our study evaluated the IDH1 mutation and 1p/19q
8
codeletion for low grade gliomas in addition to EOR, in order to assess the prognostic roles of
9
each variable. We found that the 5-year overall survival was 100% for grade II gliomas for EOR
10
>90%, and 80% for EOR < 90%, while the 2 year overall survival for high grade gliomas was
11
83.7% for EOR >90% and 43.8% for EOR <90%. We also found that EOR was a significant
12
prognostic factor for OS, PFS, and MPFS for insular gliomas, and that for low grade gliomas
13
EOR and the 1p19q co-deletion significantly predicted PFS. Our study shows that the IDH1
14
mutation had a significant association with OS and PFS with the univariate analysis, as has also
15
been presented in the literature,44,45 however, for the multivariate analysis factoring in EOR and
16
1p/19q co-deletion, the IDH1 mutation was not found to be a significant predictor of OS and
17
PFS. This finding suggests that the extent of resection also plays an important role in the
18
prognosis of IDH1 mutation for low grade insular gliomas. Understanding the roles of the
19
insular glioma genetic makeup and extent of resection provides a robust understanding about
20
patient survival that can aid with prognosis and treatment for patients.
AC C
EP
TE D
M AN U
SC
RI PT
1
21
22
23
Patient Selection
ACCEPTED MANUSCRIPT Insular glioma extent of resection
16
With this new understanding about how EOR and molecular genetics play a role in the
2
prognosis for insular glioma patients, how do you choose a proper surgical candidate? Various
3
indications for patient selection have been proposed in the literature including whether there is
4
lenticulostriate arteries involvement, the characteristic of the tumor boundaries, and the cortical
5
involvement.21,23,37 Tumor borders that are well defined are more amenable for complete
6
resection than diffuse borders. Kawaguchi et al.14 experienced a 75.7% gross total resection for
7
sharp tumor borders that had distinct medial borders from the basal ganglia, while only a 19.6%
8
gross total resection for tumors with diffuse borders. Enhancement of the tumor can also be an
9
indication of abundant intratumoral vasculature and has been shown to have increased
10
postoperative neurological deficits, likely due to injury of lenticulostriate arteries that are in
11
proximity to the vascular tumor.14 Care must be taken for tumors that involve the lenticulostriate
12
arteries or the long insular artery, often located in the superior-posterior central insular sulcus, as
13
these tumors have been associated with poorer extent of resection and more postoperative
14
ischemic complications.14,46 We generally include a MR angiogram, or vascular imaging, with
15
our preoperative imaging for these tumors in order to fully assess the location of the arteries. By
16
preoperatively assessing the general locations or the MCA and feeding arteries, we can anticipate
17
regions of higher risk for vascular injury with intraoperative neuro-navigation during tumor
18
resection. The T1-weighted, with contrast, or T2-weighted MRI imaging can be used
19
intraoperatively with neuro-navigation to help identify nearby vasculature (Figure 6).
21
SC
M AN U
TE D
EP
AC C
20
RI PT
1
Limits of the study
22
Our study contains the usual constraints associated with a retrospective study. In addition,
23
this study evaluates a small cohort that contains a heterogeneous group of high and low grade
ACCEPTED MANUSCRIPT Insular glioma extent of resection
17
gliomas. The molecular marker evaluations are based on the low grade population, which adds
2
new insight to our understanding of insular gliomas, but will require larger studies to further
3
investigate this topic. Our institution evaluates the R132H mutation when testing for IDH
4
mutation, leaving the possibility that other IDH mutations could have associations with survival
5
that are not identified in this study.
RI PT
1
6
Conclusion
SC
7
This study suggests that for insular gliomas, EOR is a significant predictor of OS, PFS,
9
and MPFS. For low grade insular gliomas, EOR and the 1p/19q co-deletion were found to be
10
significant predictors of PFS. With an improved understanding about the role of genetics and
11
extent of resection on insular gliomas, a more effective strategy can be developed for providing
12
optimal treatment for insular gliomas.
EP
15
AC C
14
TE D
13
M AN U
8
ACCEPTED MANUSCRIPT Insular glioma extent of resection
18
References
2
1. Duffau H, Capelle L. Preferential brain locations of low-grade gliomas. Cancer. 2004;100(12):2622-
3
2626.
4
2. Duffau H. A personal consecutive series of surgically treated 51 cases of insular WHO grade II glioma:
5
Advances and limitations. J Neurosurg. 2009;110(4):696-708.
6
3. Saito R, Kumabe T, Kanamori M, Sonoda Y, Tominaga T. Insulo-opercular gliomas: Four different
7
natural progression patterns and implications for surgical indications. Neurol Med Chir (Tokyo).
8
2010;50(4):286-290.
9
4. Kumabe T, Higano S, Takahashi S, Tominaga T. Ischemic complications associated with resection of
M AN U
SC
RI PT
1
opercular glioma. J Neurosurg. 2007;106(2):263-269.
11
5. Benet A, Hervey-Jumper SL, Sanchez JJ, Lawton MT, Berger MS. Surgical assessment of the insula.
12
part 1: Surgical anatomy and morphometric analysis of the transsylvian and transcortical approaches to
13
the insula. J Neurosurg. 2016;124(2):469-481.
14
6. Hervey-Jumper SL, Li J, Osorio JA, et al. Surgical assessment of the insula. part 2: Validation of the
15
berger-sanai zone classification system for predicting extent of glioma resection. J Neurosurg.
16
2016;124(2):482-488.
17
7. Signorelli F, Guyotat J, Elisevich K, Barbagallo GM. Review of current microsurgical management of
18
insular gliomas. Acta Neurochir (Wien). 2010;152(1):19-26.
19
8. Kim YH, Kim CY. Current surgical management of insular gliomas. Neurosurg Clin N Am.
20
2012;23(2):199-206, vii.
AC C
EP
TE D
10
ACCEPTED MANUSCRIPT Insular glioma extent of resection
19
9. Isnard J, Guenot M, Sindou M, Mauguiere F. Clinical manifestations of insular lobe seizures: A stereo-
2
electroencephalographic study. Epilepsia. 2004;45(9):1079-1090.
3
10. Guenot M, Isnard J. Epilepsy and insula. Neurochirurgie. 2008;54(3):374-381.
4
11. Signorelli F, Guyotat J, Elisevich K, Barbagallo GM. Review of current microsurgical management of
5
insular gliomas. Acta Neurochir (Wien). 2010;152(1):19-26.
6
12. Tanriover N, Rhoton AL,Jr, Kawashima M, Ulm AJ, Yasuda A. Microsurgical anatomy of the insula
7
and the sylvian fissure. J Neurosurg. 2004;100(5):891-922.
8
13. Sanai N, Polley MY, Berger MS. Insular glioma resection: Assessment of patient morbidity, survival,
9
and tumor progression. J Neurosurg. 2010;112(1):1-9.
M AN U
SC
RI PT
1
14. Kawaguchi T, Kumabe T, Saito R, et al. Practical surgical indicators to identify candidates for radical
11
resection of insulo-opercular gliomas. J Neurosurg. 2014;121(5):1124-1132.
12
15. Pitskhelauri DI, Bykanov AE, Zhukov VY, Kachkov IA, Buklina SB, Tonoyan AS. Review of
13
surgical treatment of insular gliomas: Challenges and opportunities. Zh Vopr Neirokhir Im N N Burdenko.
14
2015;79(2):111-116.
15
16. Ebeling U, Kothbauer K. Circumscribed low grade astrocytomas in the dominant opercular and
16
insular region: A pilot study. Acta Neurochir (Wien). 1995;132(1-3):66-74.
17
17. Duffau H, Capelle L, Lopes M, Faillot T, Sichez JP, Fohanno D. The insular lobe: Physiopathological
18
and surgical considerations. Neurosurgery. 2000;47(4):801-10; discussion 810-1.
19
18. Shankar A, Rajshekhar V. Radiological and clinical outcome following stereotactic biopsy and
20
radiotherapy for low-grade insular astrocytomas. Neurol India. 2003;51(4):503-506.
AC C
EP
TE D
10
ACCEPTED MANUSCRIPT Insular glioma extent of resection
20
19. Zentner J, Meyer B, Stangl A, Schramm J. Intrinsic tumors of the insula: A prospective surgical study
2
of 30 patients. J Neurosurg. 1996;85(2):263-271.
3
20. Heffez DS. Stereotactic transsylvian, transinsular approach for deep-seated lesions. Surg Neurol.
4
1997;48(2):113-124.
5
21. Hentschel SJ, Lang FF. Surgical resection of intrinsic insular tumors. Neurosurgery. 2005;57(1
6
Suppl):176-83; discussion 176-83.
7
22. Moshel YA, Marcus JD, Parker EC, Kelly PJ. Resection of insular gliomas: The importance of
8
lenticulostriate artery position. J Neurosurg. 2008;109(5):825-834.
9
23. Sanai N, Berger MS. Intraoperative stimulation techniques for functional pathway preservation and
M AN U
SC
RI PT
1
glioma resection. Neurosurg Focus. 2010;28(2):E1.
11
24. Simon M, Neuloh G, von Lehe M, Meyer B, Schramm J. Insular gliomas: The case for surgical
12
management. J Neurosurg. 2009;110(4):685-695.
13
25. Cancer Genome Atlas Research Network, Brat DJ, Verhaak RG, et al. Comprehensive, integrative
14
genomic analysis of diffuse lower-grade gliomas. N Engl J Med. 2015;372(26):2481-2498.
15
26. Eseonu CI, Rincon-Torroella J, ReFaey K, et al. Awake craniotomy versus craniotomy under general
16
anesthesia for perirolandic gliomas: Evaluating perioperative complications and extent of resection.
17
Neurosurgery. Accepted September 2016.
18
27. Chaichana KL, Jusue-Torres I, Navarro-Ramirez R, et al. Establishing percent resection and residual
19
volume thresholds affecting survival and recurrence for patients with newly diagnosed intracranial
20
glioblastoma. Neuro Oncol. 2014;16(1):113-122.
AC C
EP
TE D
10
ACCEPTED MANUSCRIPT Insular glioma extent of resection
21
28. Chaichana KL, Jusue-Torres I, Lemos AM, et al. The butterfly effect on glioblastoma: Is volumetric
2
extent of resection more effective than biopsy for these tumors? J Neurooncol. 2014;120(3):625-634.
3
29. Eseonu CI, Rincon-Torroella J, ReFaey K, Quinones-Hinojosa A. The cost of brain surgery: Awake
4
versus asleep craniotomy for perirolandic region tumors. Neurosurgery. Accepted August 2016.
5
30. Wang Y, Wang Y, Fan X, et al. Putamen involvement and survival outcomes in patients with insular
6
low-grade gliomas. J Neurosurg. 2016:1-7.
7
31. Yasargil MG, von Ammon K, Cavazos E, Doczi T, Reeves JD, Roth P. Tumours of the limbic and
8
paralimbic systems. Acta Neurochir (Wien). 1992;118(1-2):40-52.
9
32. Vanaclocha V, Saiz-Sapena N, Garcia-Casasola C. Surgical treatment of insular gliomas. Acta
M AN U
SC
RI PT
1
Neurochir (Wien). 1997;139(12):1126-34; discussion 1134-5.
11
33. Lang FF, Olansen NE, DeMonte F, et al. Surgical resection of intrinsic insular tumors: Complication
12
avoidance. J Neurosurg. 2001;95(4):638-650.
13
34. Barbosa BJ, Dimostheni A, Teixeira MJ, Tatagiba M, Lepski G. Insular gliomas and the role of
14
intraoperative assistive technologies: Results from a volumetry-based retrospective cohort. Clin Neurol
15
Neurosurg. 2016;149:104-110.
16
35. Kumabe T, Nakasato N, Suzuki K, et al. Two-staged resection of a left frontal astrocytoma involving
17
the operculum and insula using intraoperative neurophysiological monitoring--case report. Neurol Med
18
Chir (Tokyo). 1998;38(8):503-507.
19
36. Morantz RA. Radiation therapy in the treatment of cerebral astrocytoma. Neurosurgery.
20
1987;20(6):975-982.
AC C
EP
TE D
10
ACCEPTED MANUSCRIPT Insular glioma extent of resection
22
37. Saito R, Kumabe T, Inoue T, et al. Magnetic resonance imaging for preoperative identification of the
2
lenticulostriate arteries in insular glioma surgery. technical note. J Neurosurg. 2009;111(2):278-281.
3
38. Schatz CR, Kreth FW, Faist M, Warnke PC, Volk B, Ostertag CB. Interstitial 125-iodine radiosurgery
4
of low-grade gliomas of the insula of reil. Acta Neurochir (Wien). 1994;130(1-4):80-89.
5
39. Duffau H, Peggy Gatignol ST, Mandonnet E, Capelle L, Taillandier L. Intraoperative subcortical
6
stimulation mapping of language pathways in a consecutive series of 115 patients with grade II glioma in
7
the left dominant hemisphere. J Neurosurg. 2008;109(3):461-471.
8
40. McGirt MJ, Mukherjee D, Chaichana KL, Than KD, Weingart JD, Quinones-Hinojosa A. Association
9
of surgically acquired motor and language deficits on overall survival after resection of glioblastoma
M AN U
SC
RI PT
1
multiforme. Neurosurgery. 2009;65(3):463-9; discussion 469-70.
11
41. Rahman M, Abbatematteo J, De Leo EK, et al. The effects of new or worsened postoperative
12
neurological deficits on survival of patients with glioblastoma. J Neurosurg. 2016:1-9.
13
42. Cairncross G, Wang M, Shaw E, et al. Phase III trial of chemoradiotherapy for anaplastic
14
oligodendroglioma: Long-term results of RTOG 9402. J Clin Oncol. 2013;31(3):337-343.
15
43. van den Bent MJ, Brandes AA, Taphoorn MJ, et al. Adjuvant procarbazine, lomustine, and vincristine
16
chemotherapy in newly diagnosed anaplastic oligodendroglioma: Long-term follow-up of EORTC brain
17
tumor group study 26951. J Clin Oncol. 2013;31(3):344-350.
18
44. Parsons DW, Jones S, Zhang X, et al. An integrated genomic analysis of human glioblastoma
19
multiforme. Science. 2008;321(5897):1807-1812.
20
45. Yan H, Parsons DW, Jin G, et al. IDH1 and IDH2 mutations in gliomas. N Engl J Med.
21
2009;360(8):765-773.
AC C
EP
TE D
10
ACCEPTED MANUSCRIPT Insular glioma extent of resection
23
46. Iwasaki M, Kumabe T, Saito R, et al. Preservation of the long insular artery to prevent postoperative
2
motor deficits after resection of insulo-opercular glioma: Technical case reports. Neurol Med Chir
3
(Tokyo). 2014;54(4):321-326.
RI PT
1
4
AC C
EP
TE D
M AN U
SC
5
ACCEPTED MANUSCRIPT Insular glioma extent of resection
24
1
Figure Legend:
2
Figure 1: Kaplan-Meier curves revealing the overall survival in patients based on glioma pathology grade. Patients with low grade pathology had a significantly longer overall survival (p=0.0034).
9 10 11 12 13 14 15 16 17 18 19
RI PT
8
Figure 3: Kaplan-Meier curves revealing the progression free survival in patients based on glioma pathology grade. Patients with a lower pathology grade had a significantly longer progression free survival (p<0.0001).
SC
7
Figure 2: Kaplan-Meier curves showing the overall survival in patients with Grade II (left) and Grade III/IV (right) insular gliomas, stratified by Extent of resection (EOR). For the purposes of visual display, patients were grouped by EOR ≥90, 70-89%, and less than 70%.
Figure 4: Kaplan-Meier curves showing the progression free survival in patients with Grade II (left) and Grade III/IV (right) insular gliomas, stratified by Extent of resection (EOR). For the purposes of visual display, patients were grouped by EOR ≥90, 70-89%, and less than 70%.
M AN U
6
Figure 5: Kaplan-Meier curves revealing the malignant progression free survival in patients with Grade II insular gliomas, stratified by Extent of resection (EOR). For the purposes of visual display, patients were grouped by EOR ≥90 and less than 90%. Figure 6: (Left) Coronal MRA and (Right) T1-weighted MRI with gadolinium contrast depicting an insular glioma with nearby MCA and feeding arteries.
TE D
5
EP
4
AC C
3
ACCEPTED MANUSCRIPT
Table 1: Summary of patient and disease characteristics of 74 insular glioma patients No. (%)
RI PT
54 18-80 42 (56.8) 32 (43.2)
SC
80 50-100
AC C
EP
26 (35.1) 14 (18.9) 17 (23.0) 5 (6.8) 12 (16.2)
M AN U
TE D
Characteristic Age (years) median Range Gender Male Female Preoperative KPS median range Preoperative symptoms Seizure Headache Motor deficit Sensory deficit Language deficit Side Left Right WHO tumor classification II III IV Postop chemotherapy Postop radiation Putamen involvement Yes No Preop Tumor Volume (cm3) median range Postop Tumor Volume (cm3) median range EOR ≥ 90% 70-89% < 70% median % Range % IDH1* Mutation Wild type 1p/19q*
43 (58.1) 31 (41.9) 25 (33.8) 16 (21.6) 33 (44.6) 53 (71.6) 33 (44.6) 35 (47.3) 39 (52.7) 43.6 2.2-256.4
3.7 0-109.4 40 (54.0) 15 (20.3) 19 (25.7) 91.7 10.6 - 100 19 (76.0) 6 (24.0)
ACCEPTED MANUSCRIPT
18 (72.0) 7 (28.0) 4.4 1.2-10.1
AC C
EP
TE D
M AN U
SC
RI PT
Codeletion Intact Clinical follow-up (yrs) median range KPS: Karnofsky Performance Status; * From 25 low grade tumors
ACCEPTED MANUSCRIPT
Table 2: Postoperative neurological outcomes following insular glioma surgery Characteristic Postop KPS median range Postop Seizure New Motor Deficit Transient Deficit Permanent Deficit New Sensory Deficit New Language Deficit Transient Deficit, n (%) Permanent Deficit, n (%) Wound Infection, n (%) Meningitis, n (%) Length of Hospital Stay median range Disposition
No (%)
M AN U
SC
RI PT
80 50-100 9 (16.0) 8 (10.8) 6 (8.1) 2 (2.7) 3 (4.1) 5 (6.8) 5 (6.8) 0 (0.0) 0 (0.0) 0 (0.0) 4.5 1-22
50 (67.6) 24 (32.4)
Discharged to Home
AC C
EP
TE D
Discharged to Rehab
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 3: Univariate analyses of survival outcomes in patients with insular low grade gliomas*. EOR, IDH1 mutation, and 1p19q codeletion are significant factors for OS and PFS. Factor p Value, PFS p Value, OS p Value, MPFS Extent of resection 0.028 0.054 0.014 Age 0.349 0.278 0.919 Sex 0.627 0.997 0.932 Preoperative seizures 0.354 0.922 0.896 Preoperative KPS 0.863 0.556 0.914 Left Side tumor 0.212 0.811 0.375 Preoperative Tumor 0.998 0.176 0.314 Volume Putamen involvement 0.368 0.715 0.185 IDH1 mutation 0.548 0.076 0.093 1p19q codeletion 0.261 0.039 0.095 KPS, Karnofsky Performance Status, *Red indicates significance (p<0.15)
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 4: Multivariate analysis (Cox proportional hazards model) of survival outcomes for low grade gliomas*. EOR and 1p19q co-deletion are significant factors for PFS in the multivariable analysis, while EOR is a significant factor for OS in the multivariable analysis. PFS OS Factor p Value HR 95% CI p Value HR 95% CI Extent of Resection 0.017 0.949 0.909-0.991 0.882 0.783-0.994 0.039 IDH1 mutation 0.081 0.163 0.021-1.251 0.738 0.088 0.054-1.531 1p19q codeletion 0.029 0.002-0.491 0.546 1.602 0.001-1.993 0.014 *Red indicates significance (p<0.05)
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 5: Univariate analyses of survival outcomes in patients with insular high grade gliomas*. EOR is a significant factor for PFS and OS. Factor p Value, PFS p Value, OS Extent of resection 0.024 0.020 Age 0.412 0.287 Sex 0.577 0.316 Preoperative seizures 0.804 0.826 Preoperative KPS 0.394 0.229 Left Side tumor 0.811 0.385 Preoperative Tumor Volume 0.913 0.936 Putamen involvement 0.723 0.360 Oligodendroglioma 0.702 0.882 KPS, Karnofsky Performance Status, *Red indicates significance (p<0.15)
ACCEPTED MANUSCRIPT
R2 0.011 0.025 0.007 0.053 < 0.001 0.037 <0.001 0.015 0.010
P-value 0.375 0.181 0.492 0.048 0.818 0.100 0.985 0.268 0.327
SC
Factor Age Gender Left side tumor Preoperative Volume Putamen involvement Pathology Berger-Sanai Zone I IDH1 mutation* 1p/19q codeletion*
RI PT
Table 6: Linear Regression for predictors of EOR. Preoperative tumor volume was found to be a significant predictor of EOR for insular gliomas
AC C
EP
TE D
M AN U
*For low grade gliomas
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Highlights: - Extent of tumor resection and molecular makeup of an insular glioma are two important factors that influence survival
RI PT
- Extent of resection (EOR) is a significant predictor of overall and progression free survival for high and low grade insular gliomas. - For low grade gliomas, EOR significantly predicts malignant progression free survival, and EOR and the 1p/19q co-deletion are significant predictors for overall and progression free survival.
AC C
EP
TE D
M AN U
SC
- Preoperative tumor volume is a significant factor that influences the extent of resection for insular gliomas.
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Abbreviation list LGG - Low grade glioma HGG - High grade glioma EOR - Extent of resection LOS - Length of stay KPS - Karnofsky performance Status