Cerebral Blood Volume Calculated by Dynamic Susceptibility Contrast-Enhanced Perfusion MR Imaging: Preliminary Correlation Study with Glioblastoma Genetic Profiles
et al. (2013) Cerebral Blood Volume Calculated by Dynamic Susceptibility Contrast-Enhanced Perfusion MR
Imaging: Preliminary Correlation Study with Glioblastoma Genetic Profiles. PLoS ONE 8(8): e71704. doi:10.1371/journal.pone.0071704
Cerebral Blood Volume Calculated by Dynamic Susceptibility Contrast-Enhanced Perfusion MR Imaging: Preliminary Correlation Study with Glioblastoma Genetic Profiles
Inseon Ryoo 0
Seung Hong Choi 0
Ji-Hoon Kim 0
Chul-Ho Sohn 0
Soo Chin Kim 0
Hwa Seon Shin 0
Jeong A. Yeom 0
Seung Chai Jung 0
A. Leum Lee 0
Tae Jin Yun 0
Chul-Kee Park 0
Sung-Hye Park 0
Alexander Annala, City of Hope, United States of America
0 1 Department of Radiology, Seoul National University College of Medicine , Seoul , Korea , 2 Department of Neurosurgery, Seoul National University College of Medicine , Seoul , Korea , 3 Department of Pathology, Seoul National University College of Medicine , Seoul , Korea
Purpose: To evaluate the usefulness of dynamic susceptibility contrast (DSC) enhanced perfusion MR imaging in predicting major genetic alterations in glioblastomas. Materials and Methods: Twenty-five patients (M:F = 13:12, mean age: 52.1615.2 years) with pathologically proven glioblastoma who underwent DSC MR imaging before surgery were included. On DSC MR imaging, the normalized relative tumor blood volume (nTBV) of the enhancing solid portion of each tumor was calculated by using dedicated software (Nordic TumorEX, NordicNeuroLab, Bergen, Norway) that enabled semi-automatic segmentation for each tumor. Five major glioblastoma genetic alterations (epidermal growth factor receptor (EGFR), phosphatase and tensin homologue (PTEN), Ki67, O6-methylguanine-DNA methyltransferase (MGMT) and p53) were confirmed by immunohistochemistry and analyzed for correlation with the nTBV of each tumor. Statistical analysis was performed using the unpaired Student t test, ROC (receiver operating characteristic) curve analysis and Pearson correlation analysis. Results: The nTBVs of the MGMT methylation-negative group (mean 9.567.5) were significantly higher than those of the MGMT methylation-positive group (mean 5.461.8) (p = .046). In the analysis of EGFR expression-positive group, the nTBVs of the subgroup with loss of PTEN gene expression (mean: 10.368.1) were also significantly higher than those of the subgroup without loss of PTEN gene expression (mean: 5.662.3) (p = .046). Ki-67 labeling index indicated significant positive correlation with the nTBV of the tumor (p = .01). Conclusion: We found that glioblastomas with aggressive genetic alterations tended to have a high nTBV in the present study. Thus, we believe that DSC-enhanced perfusion MR imaging could be helpful in predicting genetic alterations that are crucial in predicting the prognosis of and selecting tailored treatment for glioblastoma patients.
Funding: This study was supported by a grant from the National R&D Program for Cancer Control, Ministry of Health & Welfare, Republic of Korea (1120300), the
Korea Healthcare technology R&D Projects, Ministry for Health, Welfare& Family Affairs (A112028 and Cover Letter HI13C0015) and by the Research Center
Program of IBS (Institute for Basic Science) in Korea. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of
Competing Interests: The authors have declared that no competing interests exist.
Malignant gliomas are the most common and lethal cancers
originating in the brain. Glioblastoma multiforme (GBM, World
Health Organization [WHO] Grade IV), the most aggressive
subtype of glioma, has a dismal prognosis. Nevertheless, survival
for patients with GBM has improved from an average of 10
months to 14 months after diagnosis in the last several years due to
the development of various treatment options .
Improvement in treatment strategies has largely been based on
the substantial progress in the identification of genetic alterations
or profiles in GBMs, which enables tailored therapy. Inarguably,
the most accurate method for the identification of genetic
alterations in GBM is pathologic analysis. Nevertheless, the
procedure for obtaining the brain tumor tissue for pathologic
analysis is time consuming, costly, physician intensive and not
always feasible, given its innate invasiveness [2,3]. For example, it
is potentially dangerous to sample tissue from patients who are in
poor general condition for brain surgery, who have tumors in a
critical portion of the brain or who have recurrent tumors on
follow-up images after treatment involving surgery. These
tissuebased methods are also often associated with sampling errors due
to improper resection or biopsy of tumor tissues and heterogeneity
of tumors, resulting in some cases with false genetic profile
assignation. Moreover, preoperative insight into the genetic
composition of the tumor is sometimes, if not often, necessary to
guide preoperative chemotherapies to shrink the tumor.
Due to the need for less- or non-invasive means of predicting
GBM genetic alterations, along with the recent tremendous
advances in imaging techniques, there have been many attempts
to use the imaging features of GBM with conventional MR and
perfusion-weighted imaging (PWI) techniques as noninvasive
radiophenotypic surrogates for genetic alterations . However,
similar attempts to correlate tumor blood volume (TBV) from
dynamic susceptibility contrast (DSC) enhanced perfusion MR
imaging with various genetic alterations in GBMs have not yet
been conducted, except for a very recent study correlating TBV
with epidermal growth factor receptor variant III (EGFRvIII)
expression status in GBMs .
Our hypothesis is that GBMs with genetic profiles associated
with poor prognosis would show high TBV values on
DSCenhanced perfusion MR imaging, because it is well known that the
TBV value can predict the progression of GBMs . Thus, the
purpose of our study is to evaluate the usefulness of the TBV value
from DSC-enhanced perfusion MR imaging in predicting major
genetic alterations in glioblastomas.
Materials and Methods
This retrospective study was approved by the institutional
review board of Seoul National University Hospital, and informed
consent was waived.
Using a computerized search of our hospitals medical records
and pathology files from November 2009 to February 2012, we
identified 102 patients with pathologically proven GBM (World
Health Organization [WHO] Grade IV) who had undergone
surgery in our hospital. Sixty-one patients without preoperative
DSC-enhanced perfusion MR imaging performed with a 3-tesla
MR machine were excluded. An additional two patients, one with
no immunohistochemistry (IHC) results and another with severe
paramagnetic artifacts on MR imaging, were also excluded.
Among the remaining 39 patients, 10 patients had GBMs with
an oligodendroglial component and four patients had a 1p 19q
deletion, which are known to be associated with a more favorable
prognosis than in those patients having general GBMs .
Finally, a total of 25 patients were included in our retrospective
study (M:F = 13:12, mean age: 52.1615.2, age range: 2671 years)
(Figure 1). All examinations were performed within 15 days before
the surgery (mean: 4.5263.97 days, range: 115 days).
MR Imaging Technique
All 25 patients underwent conventional MR imaging and
DSCenhanced perfusion MR imaging using a 3 T-scanner (Verio;
Siemens Healthcare Sector, Erlangen, Germany) with a
32channel head coil. The conventional MR imaging included
T1weighted imaging such as transverse spin-echo imaging before and
after contrast enhancement, multi-planar reconstructed transverse
and coronal imaging from a sagittal three-dimensional
magnetization prepared rapid acquisition gradient echo (3D-MPRAGE)
sequence before and after contrast enhancement and transverse
T2-weighted turbo spin-echo sequences. Contrast-enhanced
T1weighted imaging was acquired after the intravenous
administration of gadobutrol (GadovistH, Bayer Schering Pharma, Berlin,
Germany) with 0.1 mmol per kilogram (mmol/kg) of body weight.
Transverse spin-echo T1-weighted images were obtained with the
following parameters: repetition time (TR), 558 ms; echo time
(TE), 9.8 ms; flip angle (FA), 70u; matrix, 3846187; field-of-view
(FOV), 1756220 mm; section thickness, 5 mm; and number of
excitations (NEX), 1. We obtained 3D-MPRAGE sequences using
the following parameters: TR, 1500 ms; TE, 1.9 ms; FA, 9u;
matrix, 2566232; FOV, 2206250 mm; section thickness, 1 mm;
and NEX, 1. The parameters of transverse T2-weighted images
were as follows: TR, 5160 ms; TE, 91 ms; FA, 124130u; matrix,
6406510580; FOV, 1751996220 mm; section thickness,
5 mm; and NEX, 3.
Figure 2. Clustering method was used as a semi-automatic segmentation of glioblastoma. (A) Contrast-enhanced T1-weighted images
were used as structural images. (B) Volume of interest (VOI) was determined by a neuroradiologist. (C, D) VOIs were divided into seven tissue classes
according to pixel values and class 6 and 7 were selected as enhancing tumor tissue. (C: class 6, D: class 6 and 7 together).
The transverse DSC-enhanced perfusion MR imaging was
obtained with a single-shot gradient-echo echo-planar sequences
during the intravenous administration of gadobutrol with
0.1 mmol/kg of body weight at a rate of 4 mL/sec by use of a
EGFR expression(+) group (n = 22)
PTEN loss (n = 7)
PTEN no loss (n = 15)
Methylation (+) (n = 16)
Methylation (2) (n = 9)
MGMT methylation status (n = 25)
power injector (Spectris; Medrad, Pittsburgh, PA). A 30-mL bolus
injection of saline was followed at the same injection rate. For each
section, 60 images were acquired at intervals equal to the TR. The
parameters were as follows: TR, 1500 ms; TE, 30 ms; FA, 90u;
matrix, 1286128; section thickness, 56 mm; intersection gap,
1 mm; FOV, 2406240 mm; sections, 1520; voxel size,
1.87561.87565 mm3; pixel bandwidth, 1563 Hz; and total
acquisition time, 1 minute 30 seconds.
Image Postprocessing and Perfusion Data Analysis
The DSC-enhanced perfusion MR images were processed by
use of an MR perfusion analysis method including semi-automatic
segmentation (Nordic TumorEx, NordicNeuroLab, Bergen,
Norway), in which contrast-enhanced T1-weighted images (CE-T1WI)
were used as structural images. There are three algorithms used
for tissue segmentation in this program. (Threshholding, Seed
growing and Clustering). Clustering was used in this study. The
volume of interests (VOIs) determined by a neuroradiologist with 7
year-experience in neuroradiology based on structural images (i.e.,
CE-T1 WI) were divided into seven tissue classes. This clustering
or classification is performed using an Expectation-Maximization
algorithm. Pixels are assigned to the class with the highest
probability of finding the vector whose elements are the pixel
values in that class. Class 6 and 7 among the seven classes were
chosen as tumor tissue in this study excluding nontumorous
portions such as tissues associated with necrosis and edema
(Figure 2). The relative cerebral blood volume (rCBV) maps were
generated by use of established tracer kinetic models applied to the
first-pass data [14,15]. To reduce the recirculation effects, the
DR2*(1/T2*) curves were fitted to a gamma-variate function,
which is an approximation of the first-pass response as it would
appear in the absence of recirculation or leakage. The dynamic
curves were mathematically corrected to reduce contrast-agent
leakage effects . Normalization of rCBV maps is automatically
performed using the mean value of the blood volume values
outside the tumor without any intervention of observers. The
normalized rCBV (nCBV) maps were presented as color overlays
on structural images. The coregistration between the structural
images and nCBV maps was performed based on geometric
information stored in the respective data sets, automatically
[17,18]. However, manual correction was also performed before
final coregistration to minimize the misresgistration caused by
geometric distortions on echo-planar sequences. Finally, the
normalized relative tumor blood volume (nTBV ) values of the
tissue classified as tumor tissue (i.e., class 6 and 7 in this study) were
calculated based on coregistered nCBV maps.
Genetic Profile Analysis
In IHC studies of GBM, five genetic alterations, previously
documented as important genetic markers for the grading of
malignancy in gliomas, were identified [4,6,19,20]. Among the
genetic alterations, there were two proliferation markers
(epidermal growth factor receptor [EGFR] and Ki-67), one DNA repair
gene (O6-methylguanine-DNA methyltransferase [MGMT]) and
two tumor-suppressor genes (phosphatase and tensin homologue
[PTEN] and p53).
Each case was evaluated to estimate the positivity of neoplastic
cells, according to the characteristics of each biomarker. The
expression status of EGFR protein was visually scored using a
three-tiered scale (1+, 2+, 3+) and a medium-power field (x200).
To evaluate the expression status of p53 and the labeling index of
Ki-67, an Aperio image analyzer (Aperio Technologies, Vista, CA)
was utilized to calculate the percentage of positive tumor cells. For
MGMT, its promoter methylation status was investigated using a
methylation-specific polymerase chain reaction (MS-PCR)
technique. The MGMT promoter-methylation status of a tumor did
not always correspond to its MGMT immunohistochemical
expression status (MGMT-loss or MGMT-no loss groups). IHC
staining for PTEN was also described in terms of loss (PTEN-loss
or PTEN-no loss group). Finally, the pattern of the genetic
alterations was analyzed for correlation with nTBV values as
determined by DSC-enhanced perfusion MR imaging.
All statistical analyses were performed using a statistical software
program (MedCalc, version 184.108.40.206; MedCalc, Mariakerke,
Belgium). To evaluate the correlation between EGFR expression
status and nTBV values and the correlation between PTEN
expression status and nTBV, an ANOVA test and an unpaired
Student t test were used, respectively. In the EGFR
expressionpositive subgroup analysis, the unpaired Student t test and receiver
operating characteristic (ROC) curve analysis were applied to
correlate nTBV parameters and PTEN expression status. The
unpaired Student t test and ROC curve analysis were also used to
determine whether the nTBV values of the MGMT
methylationpositive group significantly differed from those of the MGMT
methylation-negative group. In the ROC analysis, cutoff values
that provided a balance between sensitivity and specificity for
prediction of the genetic profiles were determined by selecting the
coordinate that was nearest the left upper corner (ie, [0,1 ]) on
each ROC curve, and values higher than the cutoff values were
considered to indicate loss of PTEN expression or nonmethylation
of MGMT promoter .
The diagnostic accuracies based on the nTBV values were
estimated by calculating the area under the ROC curves (Az). The
sensitivities and specificities at these cutoff levels were calculated,
along with 95% confidence intervals (CIs).
Pearsons correlation analysis was performed to measure the
significance of the association between the nTBV parameters and
two genetic alterations, specifically, Ki-67 labeling index and p53
A multivariable stepwise logistic regression model was used to
determine which genetic alteration is most strongly associated with
Leave-one-out cross-validation (LOOCV) test was also done to
evaluate the accuracy of nTBV value in predicting the genetic
The data for each parameter were assessed for normality with
the Kolmogorov-Smirnov test. In all tests, P values less than.05
were considered statistically significant.
There were no significant differences in nTBV values among
the groups with different EGFR expression statuses (p = .667). In
terms of the loss of PTEN, we found no significant difference
between the groups with the loss of PTEN (mean: 9.368.0, n = 8)
and without the loss of PTEN (mean: 5.862.2, n = 17), (p = .09).
However, in subgroup analysis with the EGFR expression-positive
group (n = 22), the nTBV values of the subgroup (n = 7) with loss
of PTEN gene expression (mean: 10.368.1) were significantly
higher than those of the subgroup (n = 15) without loss of PTEN
gene expression (mean: 5.662.3) (p = .046). The nTBV values
according to the EGFR expression and PTEN expression statuses
along with the results of Kolmogorov-Smirnov test are
summarized in table 1 and figure 3. In the ROC curve analysis of EGFR
expression-positive group, the Az value of nTBV for predicting
PTEN loss was 0.733. The cutoff value that provided a balance
between sensitivity and specificity was 8.05 and the corresponding
sensitivity and specificity were 100% and 44.4%, respectively
The nTBV values were significantly lower in the MGMT
methylation-positive group (mean:5.461.8, n = 16)) than those in
the MGMT methylation-negative group (mean: 9.567.5, n = 9)
(p = .046) (table 1 and figure 4). In the ROC curve analysis, Az of
nTBV for predicting MGMT methylation was 0.677. The
sensitivity and specificity at the cutoff value of 6.19 were 73.3%
and 85.7%, respectively (table 2).
With LOOCV test, the accuracies of nTBV in predicting PTEN
loss in EGFR positive group and MGMT promoter methylation
were 72.7% (16/22) and 76.0% (19/25), respectively.
The Ki-67 labeling index also presented a strong positive
correlation with nTBV parameters (r = 0.483, p = .014) (Figure 5).
However, p53 expression status (r = 0.299, p = .166) did not
present a significant correlation with the nTBV values (Figure 6).
In LOOCV test for evaluating Ki-67 labeling index with nTBV
values, Pearson correlation coefficient was 0.38 (p = .06).
From the univariate analysis, we found that the genetic
alterations associated with an increase in nTBV included EGFR
overexpression with PTEN loss, a non-methylated MGMT
promoter and the Ki-67 labeling index. Thus, we performed a
multivariable stepwise logistic regression analysis using the genetic
alterations, which revealed that the Ki-67 labeling index was most
strongly correlated with nTBV (p = .014) values.
Figure 7 and 8 present the representative cases with high and
low nTBV values, respectively, in which the genetic profiles are
correlated with nTBV values.
In this study, we analyzed DSC-enhanced perfusion MR
imaging of GBMs, which was correlated with genetic alterations
of the tumors. We found that the EGFR expression-positive GBMs
with loss of PTEN expression showed significantly higher nTBV
values than those without loss of PTEN expression, and GBMs
without MGMT promoter methylation also showed significantly
higher nTBV values than those with MGMT promoter
methylation. In addition, the Ki-67 labeling index had a positive
correlation with nTBV parameter. Interestingly, the Ki-67 index
Figure 7. Glioblastomas with aggressive genetic profiles show high normalized relative tumor blood volume (nTBV). (A, B) A
66-yearold woman had glioblastoma with EGFR expression (3+), PTEN loss (+), MGMT methylation (2), and a Ki-67 index of 39%. The tumor showed high
nTBV (27.5).(C, D) A 70-year-old man had glioblastoma with EGFR expression (3+), PTEN loss (+), MGMT methylation (2), and a Ki-67 index of 27%. The
tumor showed high nTBV (12.84). (A, C) Contrast-enhanced T1-weighted axial image, (B, D) normalized relative cerebral blood volume (nCBV) map
overlaid on structural contrast-enhanced T1-weighted axial image.
was most strongly correlated with nTBV in GBMs. These genetic
alterations have been known to be poor prognostic factors in
GBMs, and the increase in nTBV is also a well-known predictor
for progression in GBMs. Finally, we demonstrated that nTBV
values were able to reflect some genetic alterations related to GBM
EGFR is the most frequently amplified gene and a primary
contributor to tumor initiation and progression in GBM .
EGFR amplification is associated with EGFR overexpression. The
increased activity of EGFR promotes tumor growth through many
different mechanisms, promoting survival, invasion, and
angiogenesis. Many previous studies have demonstrated that EGFR
stimulation activates many downstream signal pathways, including
the phosphoinositide-3-kinase (PI3K)/Akt pathway. The EGFR
becomes activated upon binding to epidermal growth factor (EGF)
and recruits PI3K to the cell membrane. PI3K converts
phosphatidylinositol-4,5-bisphosphate (PIP2) to the
second-messenger molecule phosphatidylinositol-3,4,5-triphosphate (PIP3),
which activates Akt. Activated Akt, in turn, activates the
mammalian target of rapamycin (mTOR), which helps induce
cellular proliferation and block apoptosis. Here is the point where
PTEN plays an important role in inhibiting tumor proliferation.
Figure 8. Glioblastomas with favorable genetic profiles show low normalized relative tumor blood volume (nTBV). (A, B) A
67-yearold man had glioblastoma with EGFR expression (2+), PTEN loss (2), MGMT methylation (+), and a Ki-67 index of 5%. The tumor showed low nTBV
(4.3). (C, D) A 36-year-old woman had glioblastoma with EGFR expression (+), PTEN loss (2), MGMT methylation (+), and a Ki-67 index of 9%. The
tumor showed low nTBV (4.59) (A, C) Contrast-enhanced T1-weighted axial image, (B, D) normalized relative cerebral blood volume (nCBV) map
overlaid on structural contrast-enhanced T1-weighted axial image.
PTEN terminates the PIP3 signal that results in blocking the
downstream pathways of EGFR-mediated cellular proliferation
[5,22,23]. Therefore, GBMs with EGFR expression combined
with PTEN gene loss present poor prognosis, and this may
influence responses to EGFR-targeted therapy. The increased
EGFR signaling by amplified EGFR expression with PTEN loss
promotes increased angiogenesis, and increased TBV is a measure
of microvasculature density and has previously been shown to
reflect an underlying angiogenetic mechanism ; thus, tumors
with EGFR overexpression with PTEN loss can have increased
MGMT is a repair protein that specifically removes
promutagenic alkyl groups from the O6 position of guanine in DNA,
thereby protecting tumor cells against alkylating agents [25,26].
Loss of MGMT expression may be caused by the methylation of
promoter CpG islands. MGMT promoter methylation is
frequently present in GBM (4575%) and has been associated with
longer survival of GBM patients treated with temozolomide [27
31]. Therefore, the evaluation of the methylation status of the
MGMT promoter is important for selecting treatment options.
Mutations in p53, a famous tumor suppressor gene, are found in
a diverse variety of human tumors . However, in this study
there was no significant correlation with nTBV value.
Tumor cellularity and proliferation index have been shown to
be associated with patient prognosis and survival [33,34]. Ki-67 is
a well-known biomarker representing the proliferative activity of
tumor cells. An increase in the Ki-67 proliferation index has been
found to be related with unfavorable prognosis in high-grade
gliomas [33,35,36]. Considering that increased proliferation also
means increased angiogenesis, an increased proliferation index
may lead to increased TBV.
Rising interest and the need for non- or less-invasive techniques
for the identification of the genetic alterations mentioned above in
GBM, accompanied with exponential growth in the power and
utility of imaging techniques such as MRI over the past 20 years,
has led radiologists to search for imaging features or biomarkers
reflecting genetic profiles of GBM. There have been several studies
correlating various imaging features such as tumor border,
enhancement pattern, peritumoral edema, multiplicity, T2 signal
and intratumoral cystic change with genetic profiles of GBM,
including EGFR and MGMT expression [4,6,3739]. Recently,
there has been increased interest in the EGFR status of GBM and
EGFR-targeted therapy, including EGFRvIII (a mutated variant
of EGFR with continuous autoactivation without EGF found
frequently (2030%) in GBMs) targeted therapy, which has been
the subject of several studies relating imaging and EGFR status,
especially using perfusion MR images [13,20]. A very recent study
revealed that EGFRvIII-expressing GBMs presented significantly
higher relative TBV compared with those tumors lacking
EGFRvIII expression . When employing several promising
therapies directed specifically at EGFRvIII, including EGFRvIII
peptide vaccination and EGFRvIII-directed monoclonal
antibodies, that have emerged recently, imaging-based identification of
EGFR status becomes increasingly important [40,41]. However,
studies correlating perfusion MR imaging with other genetic
alterations in GBM, including the genetic profiles analyzed in this
study, have not yet been conducted.
Aside from the intrinsic limits of any retrospective study, several
other limitations of this study should be mentioned. First of all, the
sample size was rather small to generalize our findings and some
results showed subtle statistical significance. Furthermore,
LOOCV test demonstrated only borderline significance (p = .06)
for the positive correlation between Ki-67 labeling index and
nTBV, even though the p-value was 0.01 for their correlation on
Pearson correlation analysis. The result may also be attributed to
our small sample size. Thus further investigations with larger
populations and more precise analysis such as histogram analysis
are warranted to strengthen the statistical power. Secondly, we did
not evaluate EGFRvIII and vascular endothelial growth factor
(VEGF) expression status which are also known to be related to
rTBV , as currently there are no available IHC methods and
cytokine analysis (reverse transcription polymerase chain reaction,
RT-PCR) for such analyses. Therefore, we could not assess the
relationship between the perfusion parameters and EGFRvIII or
VEGF expression status along with PTEN status. Lastly, the
Nordic TumorEX software enabled only calculation of normalized
rCBV values, other perfusion parameters such as cerebral blood
flow could not be evaluated in this study.
Despite all these limitations, our study revealed that GBMs with
aggressive genetic alterations tended to have high relative TBV
values. These results indicate the potential of DSC-enhanced
perfusion MR imaging as a noninvasive radiophenotypic
biomarker for genetic profiles of GBMs that are crucial in predicting
the prognosis and response to specific treatment and selecting a
tailored therapy for GBM patients. Although pathologic analysis is
still the gold standard in confirming genetic profiles,
DSCenhanced perfusion MR imaging, which may serve as a relatively
non-invasive and convenient mean to predict the genetic profiles,
could be a potentially useful alternative to surgery, especially for
patients who cannot undergo invasive procedures. Furthermore,
given that MR imaging is widely used in current clinical practice,
our study could be a cornerstone of further research focused on
determining the imaging biomarkers for various tumors, in
addition to GBMs.
In conclusion, our results suggest that DSC-enhanced perfusion
MR imaging could be a potentially useful alternative to surgery for
prediction of the genetic profiles, thus serving as a guide to
appropriate treatment selection, especially for patients who cannot
undergo invasive procedures.
Conceived and designed the experiments: IR SHC. Performed the
experiments: IR SHC SCJ. Analyzed the data: IR SHC SCJ SCK HSS
JAY ALL TJY. Contributed reagents/materials/analysis tools: IR SHC
JHK CHS CKP SHP. Wrote the paper: SHC IR SCJ.
1. Van Meir EG , Hadjipanayis CG , Norden AD , Shu HK , Wen PY , et al. ( 2010 ) Exciting new advances in neuro-oncology: the avenue to a cure for malignant glioma . CA Cancer J Clin 60 : 166 - 193 .
2. Jackson RJ , Fuller GN , Abi-Said D , Lang FF , Gokaslan ZL , et al. ( 2001 ) Limitations of stereotactic biopsy in the initial management of gliomas . Neuro Oncol 3 : 193 - 200 .
3. Tykocinski ES , Grant RA , Kapoor GS , Krejza J , Bohman LE , et al. ( 2012 ) Use of magnetic perfusion-weighted imaging to determine epidermal growth factor receptor variant III expression in glioblastoma . Neuro Oncol 14 : 613 - 623 .
4. Diehn M , Nardini C , Wang DS , McGovern S , Jayaraman M , et al. ( 2008 ) Identification of noninvasive imaging surrogates for brain tumor gene-expression modules . Proc Natl Acad Sci U S A 105 : 5213 - 5218 .
5. Mellinghoff IK , Wang MY , Vivanco I , Haas-Kogan DA , Zhu S , et al. ( 2005 ) Molecular determinants of the response of glioblastomas to EGFR kinase inhibitors . N Engl J Med 353 : 2012 - 2024 .
6. Drabycz S , Roldan G , de Robles P , Adler D , McIntyre JB , et al. ( 2010 ) An analysis of image texture, tumor location, and MGMT promoter methylation in glioblastoma using magnetic resonance imaging . Neuroimage 49 : 1398 - 1405 .
7. Law M , Young RJ , Babb JS , Peccerelli N , Chheang S , et al. ( 2008 ) Gliomas: predicting time to progression or survival with cerebral blood volume measurements at dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging . Radiology 247 : 490 - 498 .
8. Cairncross JG , Macdonald DR ( 1988 ) Successful chemotherapy for recurrent malignant oligodendroglioma . Ann Neurol 23 : 360 - 364 .
9. Quinn JA , Reardon DA , Friedman AH , Rich JN , Sampson JH , et al. ( 2004 ) Phase 1 trial of irinotecan plus BCNU in patients with progressive or recurrent malignant glioma . Neuro Oncol 6 : 145 - 153 .
10. Burton EC , Lamborn KR , Feuerstein BG , Prados M , Scott J , et al. ( 2002 ) Genetic aberrations defined by comparative genomic hybridization distinguish long-term from typical survivors of glioblastoma . Cancer Res 62 : 6205 - 6210 .
11. Schmidt MC , Antweiler S , Urban N , Mueller W , Kuklik A , et al. ( 2002 ) Impact of genotype and morphology on the prognosis of glioblastoma . J Neuropathol Exp Neurol 61 : 321 - 328 .
12. Chawla S , Krejza J , Vossough A , Zhang Y , Kapoor GS , et al. ( 2013 ) Differentiation between Oligodendroglioma Genotypes Using Dynamic Susceptibility Contrast Perfusion-Weighted Imaging and Proton MR Spectroscopy . AJNR Am J Neuroradiol.
13. Whitmore RG , Krejza J , Kapoor GS , Huse J , Woo JH , et al. ( 2007 ) Prediction of oligodendroglial tumor subtype and grade using perfusion weighted magnetic resonance imaging . J Neurosurg 107 : 600 - 609 .
14. Rosen BR , Belliveau JW , Vevea JM , Brady TJ ( 1990 ) Perfusion imaging with NMR contrast agents . Magn Reson Med 14 : 249 - 265 .
15. Ostergaard L , Sorensen AG , Kwong KK , Weisskoff RM , Gyldensted C , et al. ( 1996 ) High resolution measurement of cerebral blood flow using intravascular tracer bolus passages. Part II: Experimental comparison and preliminary results . Magn Reson Med 36 : 726 - 736 .
16. Boxerman JL , Schmainda KM , Weisskoff RM ( 2006 ) Relative cerebral blood volume maps corrected for contrast agent extravasation significantly correlate with glioma tumor grade, whereas uncorrected maps do not . AJNR Am J Neuroradiol 27 : 859 - 867 .
17. Bjornerud A ( 2003 ) The ICE software package: direct co-registration of anatomical and functional datasets using DICOM image geometry information . Proc Hum Brain Mapping 19.
18. Emblem KE , Nedregaard B , Nome T , Due-Tonnessen P , Hald JK , et al. ( 2008 ) Glioma grading by using histogram analysis of blood volume heterogeneity from MR-derived cerebral blood volume maps . Radiology 247 : 808 - 817 .
19. ( 2008 ) Comprehensive genomic characterization defines human glioblastoma genes and core pathways . Nature 455 : 1061 - 1068 .
20. Kapoor GS , Gocke TA , Chawla S , Whitmore RG , Nabavizadeh A , et al. ( 2009 ) Magnetic resonance perfusion-weighted imaging defines angiogenic subtypes of oligodendroglioma according to 1p19q and EGFR status . J Neurooncol 92 : 373 - 386 .
21. Park SH , Kim PN , Kim KW , Lee SW , Yoon SE , et al. ( 2006 ) Macrovesicular hepatic steatosis in living liver donors: use of CT for quantitative and qualitative assessment . Radiology 239 : 105 - 112 .
22. Endersby R , Baker SJ ( 2008 ) PTEN signaling in brain: neuropathology and tumorigenesis . Oncogene 27 : 5416 - 5430 .
23. Koul D ( 2008 ) PTEN signaling pathways in glioblastoma . Cancer Biol Ther 7 : 1321 - 1325 .
24. Maia AC , Jr., Malheiros SM , da Rocha AJ , da Silva CJ , Gabbai AA , et al. ( 2005 ) MR cerebral blood volume maps correlated with vascular endothelial growth factor expression and tumor grade in nonenhancing gliomas . AJNR Am J Neuroradiol 26 : 777 - 783 .
25. Bobola MS , Silber JR , Ellenbogen RG , Geyer JR , Blank A , et al. ( 2005 ) O6- methylguanine-DNA methyltransferase, O6-benzylguanine, and resistance to clinical alkylators in pediatric primary brain tumor cell lines . Clin Cancer Res 11 : 2747 - 2755 .
26. Silber JR , Mueller BA , Ewers TG , Berger MS ( 1993 ) Comparison of O6- methylguanine-DNA methyltransferase activity in brain tumors and adjacent normal brain . Cancer Res 53 : 3416 - 3420 .
27. Esteller M , Garcia-Foncillas J , Andion E , Goodman SN , Hidalgo OF , et al. ( 2000 ) Inactivation of the DNA-repair gene MGMT and the clinical response of gliomas to alkylating agents . N Engl J Med 343 : 1350 - 1354 .
28. Hegi ME , Diserens AC , Godard S , Dietrich PY , Regli L , et al. ( 2004 ) Clinical trial substantiates the predictive value of O-6-methylguanine-DNA methyltransferase promoter methylation in glioblastoma patients treated with temozolomide . Clin Cancer Res 10 : 1871 - 1874 .
29. Hegi ME , Diserens AC , Gorlia T , Hamou MF , de Tribolet N , et al. ( 2005 ) MGMT gene silencing and benefit from temozolomide in glioblastoma . N Engl J Med 352 : 997 - 1003 .
30. Brell M , Tortosa A , Verger E , Gil JM , Vinolas N , et al. ( 2005 ) Prognostic significance of O6-methylguanine-DNA methyltransferase determined by promoter hypermethylation and immunohistochemical expression in anaplastic gliomas . Clin Cancer Res 11 : 5167 - 5174 .
31. Dunn J , Baborie A , Alam F , Joyce K , Moxham M , et al. ( 2009 ) Extent of MGMT promoter methylation correlates with outcome in glioblastomas given temozolomide and radiotherapy . Br J Cancer 101 : 124 - 131 .
32. Yin S VME ( 2009 ) p53 pathway alterations in brain tumors .; EG VM, editor. New York : Humana Press (Springer).
33. Karkavelas G , Mavropoulou S , Fountzilas G , Christoforidou V , Karavelis A , et al. ( 1995 ) Correlation of proliferating cell nuclear antigen assessment, histologic parameters and age with survival in patients with glioblastoma multiforme . Anticancer Res 15 : 531 - 536 .
34. Cruz-Sanchez FF , Ferreres JC , Figols J , Palacin A , Cardesa A , et al. ( 1997 ) Prognostic analysis of astrocytic gliomas correlating histological parameters with the proliferating cell nuclear antigen labelling index (PCNA-LI) . Histol Histopathol 12 : 43 - 49 .
35. Donato V , Papaleo A , Castrichino A , Banelli E , Giangaspero F , et al. ( 2007 ) Prognostic implication of clinical and pathologic features in patients with glioblastoma multiforme treated with concomitant radiation plus temozolomide . Tumori 93 : 248 - 256 .
36. Habberstad AH , Gulati S , Torp SH ( 2011 ) Evaluation of the proliferation markers Ki-67/MIB-1, mitosin , survivin, pHH3, and DNA topoisomerase IIalpha in human anaplastic astrocytomas-an immunohistochemical study . Diagn Pathol 6 : 43 .
37. Pope WB , Chen JH , Dong J , Carlson MR , Perlina A , et al. ( 2008 ) Relationship between gene expression and enhancement in glioblastoma multiforme: exploratory DNA microarray analysis . Radiology 249 : 268 - 277 .
38. Van Meter T , Dumur C , Hafez N , Garrett C , Fillmore H , et al. ( 2006 ) Microarray analysis of MRI-defined tissue samples in glioblastoma reveals differences in regional expression of therapeutic targets . Diagn Mol Pathol 15 : 195 - 205 .
39. Aghi M , Gaviani P , Henson JW , Batchelor TT , Louis DN , et al. ( 2005 ) Magnetic resonance imaging characteristics predict epidermal growth factor receptor amplification status in glioblastoma . Clin Cancer Res 11 : 8600 - 8605 .
40. Hadjipanayis CG , Machaidze R , Kaluzova M , Wang L , Schuette AJ , et al. ( 2010 ) EGFRvIII antibody-conjugated iron oxide nanoparticles for magnetic resonance imaging-guided convection-enhanced delivery and targeted therapy of glioblastoma . Cancer Res 70 : 6303 - 6312 .
41. Sampson JH , Archer GE , Mitchell DA , Heimberger AB , Bigner DD ( 2008 ) Tumor-specific immunotherapy targeting the EGFRvIII mutation in patients with malignant glioma . Semin Immunol 20 : 267 - 275 .