F-18 FDG PET-CT for predicting survival in patients with recurrent glioma: a prospective study.

Abstract

Recurrent gliomas are usually histologically high grade; either due to recurrence of a de novo high-grade primary or anaplastic transformation in case of low-grade tumors. Survival in these patients is variable. The objective of the present study is to evaluate the role of FDG PET-CT for predicting survival in a large group of patients with suspected recurrent glioma.
A total of 81 previously treated histopathologically proven glioma patients; with clinical and conventional imaging findings suspicious of recurrence were included in this study. All patients underwent FDG PET-CT study. Based on tumor to white matter (T/W) and tumor to grey matter (T/G) ratios, all lesions were scored on PET-CT (PET scores 0, 1 and 2). Patients were followed up clinically and by repeated imaging. Data was censored, if the patient died of disease or at the end of the study. Survival analysis was done for each variable employing univariate analysis followed by multivariate analysis, using variables found significant on univariate analysis.
PET score was found to be the most significant predictor of survival in univariate and multivariate analysis (p 0.003). Patients having PET score 2 had poorer survival compared to both PET score 0 (p 0.001) and PET score 1 (p 0.004). Other covariates found to have significant correlation with survival were primary treatment modality and clinical symptoms at the time of recurrence.
FDG uptake on PET-CT is a strong predictor of survival in patients with suspected recurrent glioma.

Full text

Page 1

FUNCTIONAL NEURORADIOLOGY
F-18 FDG PET-CT for predicting survival in patients
with recurrent glioma: a prospective study
Amburanjan Santra & Rakesh Kumar & Punit Sharma &
Chandrashekhar Bal & Pramod Kumar Julka &
Arun Malhotra
Received: 17 January 2011 /Accepted: 7 June 2011 /Published online: 8 July 2011
# Springer-Verlag 2011
Abstract
Introduction Recurrent gliomas are usually histologically
high grade; either due to recurrence of a de novo high-grade
primary or anaplastic transformation in case of low-grade
tumors. Survival in these patients is variable. The objective
of the present study is to evaluate the role of FDG PET-CT
for predicting survival in a large group of patients with
suspected recurrent glioma.
Methods A total of 81 previously treated histopathologi-
cally proven glioma patients; with clinical and conventional
imaging findings suspicious of recurrence were included in
this study. All patients underwent FDG PET-CT study.
Based on tumor to white matter (T/W) and tumor to grey
matter (T/G) ratios, all lesions were scored on PET-CT
(PET scores 0, 1 and 2). Patients were followed up
clinically and by repeated imaging. Data was censored, if
the patient died of disease or at the end of the study.
Survival analysis was done for each variable employing
univariate analysis followed by multivariate analysis, using
variables found significant on univariate analysis.
Results PET score was found to be the most significant
predictor of survival in univariate and multivariate
analysis (p 0.003). Patients having PET score 2 had poorer
survival compared to both PET score 0 (p 0.001) and PET
score 1 (p 0.004). Other covariates found to have significant
correlation with survival were primary treatment modality
and clinical symptoms at the time of recurrence.
Conclusion FDG uptake on PET-CT is a strong predictor of
survival in patients with suspected recurrent glioma.
Keywords Glioma.Recurrence.Survival.FDG PET-CT
Introduction
Primary brain tumors are a common cause of cancer related
death in young and middle aged adults [1]. The most
common primary brain tumors are gliomas. Gliomas are
broadly divided into low grade (WHO grades I and II) and
high grade (WHO grades III—anaplastic tumor and IV—
glioblastoma) [2]. All gliomas, particularly the astrocytic
neoplasms, are histologically, genetically, and thus thera-
peutically heterogeneous [3, 4].Glioblastomas are the most
malignant and most common glioma, accounting for 45–
50% of all gliomas [5]. The clinical course of glioblastoma
is usually rapid and fatal, with the median survival of about
1 year. Median survival for anaplastic tumors is 2 to 3 years.
After the initial treatment, these tumors invariably recur.
Low-grade gliomas constitute approximately 15% of all
primary brain tumors [6]. The outcome for these patients is
uncertain. Some patients survive for many years without
clinical or radiologic signs of disease progression, while
others experience a rapid malignant course and die within
1 year of diagnosis. It has been assessed that 49–85% of
low-grade astrocytomas undergo transformation to a higher
grade at some point along their clinical course [7–9].
Another fraction of patients with low-grade gliomas
A. Santra
Department of Nuclear Medicine, Medical College Kolkata,
Kolkata, India
R. Kumar (*):P. Sharma:C. Bal:A. Malhotra
Department of Nuclear Medicine, All India Institute
of Medical Sciences,
E-81, Ansari Nagar (East), AIIMS Campus,
New Delhi 110029, India
e-mail: rkphulia@yahoo.com
P. K. Julka
Department of Radiotherapy, All India Institute
of Medical Sciences,
New Delhi, India
Neuroradiology (2011) 53:1017–1024
DOI 10.1007/s00234-011-0898-3

Page 2

or magnetic resonance imaging (MRI) [11]. Their poor
specificity limits their utility to differentiate recurrent tumor
from therapy related changes which have completely
different management and prognosis.
Positron emission tomography (PET), a noninvasive
imaging technique is helpful in various clinical situations
involving the brain, both benign and malignant [12]. F18
fluorodeoxyglucose (FDG) PET has been proven to be of
value in the assessment of tumor grading and recognizing
malignant degeneration [13, 14], in differentiating tumor
recurrence from tumor necrosis [13–17], in therapy
response assessment [18, 19], and in differentiating
between low-grade tumors and non-tumor lesions [20].
PET provides in vivo metabolic information of the whole
tumor and can guide stereotactic biopsies toward the most
metabolically active part [13, 14, 21]. Several studies have
found FDG PET to be effective in predicting survival in
glioma patients [22, 23].These studies however, differ in
their study design and patient number, with some of them
limited to high-grade or low-grade glioma. Most of these
studies predicted prognosis in primary glioma or a mixed
population of primary and recurrent gliomas. In the
present study, we prospectively evaluated the role of
FDG PET-CT study for predicting survival in a large
group of patients with clinical suspicion and MRI findings
of recurrent glioma.
Materials and methods
This study is a prospective observational study. Patients
were enrolled in the study between August 2006 and
February 2008, after taking prior approval from the
institutional ethical committee. A total of 81 patients were
included in this study, and written informed consent was
taken from all of them. Inclusion criteria for the patients
were histopathologically proven gliomas, previous treat-
ment with surgery or radiotherapy or both, clinical
suspicion of recurrence, and contrast enhanced MRI
suspicious for recurrence. Patients with proven malignancy
of other sites from where brain metastasis can occur and
patients with other comorbid diseases were excluded from
the study. All relevant information of the patients like age,
sex, histopathology (cellular type and grade), primary site,
primary treatment, present clinical status, and size of the
recurrent tumor as in MRI were collected at the time of
inclusion in the study. All patients underwent FDG PET-CT
study at our department. Patients were treated based on
their clinical status and MRI findings. They were followed
up clinically and by repeated MRI and FDG PET-CT.
Patient data was censored if the patient died of disease or at
the end of the study (December, 2009).
FDG PET-CT imaging
FDG PET-CT scans were taken on a dedicated PET-CT
scanner (Biograph 2, Siemens, Germany). All patients
fasted for at least 4 h before the test. Blood glucose level
was below 140 mg/dl in all patients. A dose of 370 MBq
(10 mCi) of FDG was injected intravenously, and the
patients rested in a quiet room. After a 45–60-minute
uptake period, patients were taken for scanning. In the PET-
CTsystem, CTscan acquisition performed on spiral dual slice
CTwith a slice thickness of 4 mm and a pitch of 1. Image was
acquired using a matrix of 512×512 pixels and a pixel
size of about 1 mm. After transmissions scan, 3D PET
acquisition was done for 3–5 min per bed position for
one/two bed position. PET data were acquired using a
matrix of 128×128 pixels with a slice thickness of
1.5 mm. CT-based attenuation correction of the emission
images was employed. PET images were reconstructed
by iterative method ordered subset expectation maximi-
zation (two iterations and eight subsets). Reconstructed
images were displayed and analyzed in transverse,
sagittal, and coronal views.
Interpretation of FDG PET-CT images
Two experienced Nuclear Medicine physicians evaluated
the scans independently, and they were blinded to the
clinical and structural imaging findings. PET-CT images
were interpreted as positive for recurrent tumor when—1)
The lesion has definitely increased FDG accumulation
considering the background and corresponding site on
contralateral hemisphere , 2) if there was a definite lesion
in CT images of PET-CT, which shows FDG uptake that
may be isometabolic or hypometabolic. Corresponding
areas in the CT images and fused PET-CT images were
always corroborated. SUV max values of the tumor
tissue and corresponding white matter and gray matter on
the contralateral hemisphere were calculated. Two FDG
uptake ratios were calculated in all patients—Tumor to
grey matter ratio (T/G ratio) and Tumor to white matter
ratio (T/W ratio).
T/G ratio=SUVmax in Tumor area/SUVmax in contra-
lateral corresponding Gray mater area.
T/W ratio=SUVmax in Tumor area/SUVmax in contra-
lateral corresponding White mater area.
1018Neuroradiology (2011) 53:1017–1024
experiences progressive disease from the start, but the
tumor remains Grade 2 [6]. The range of survival times in
gliomas is broad and unpredictable; some patients die early
and others survive for a decade or more [10]. In previously
treated patients, necrosis induced by therapy or occurring
spontaneously during tumor progression may also show
contrast enhancement and hence cannot be distinguished
reliably from a solid tumor on computed tomography (CT)

Page 3

PET scoring
All lesions were scored according to the intensity of FDG
uptake on PET-CT. PET score 0: no visible uptake or uptake
less than white matter (T/W≤1), PET score 1: Uptake was
more than white matter but less than grey matter (T/W>1
and T/G<1), and PET score 2: Uptake equal to or more than
grey matter (T/G≥1).
Statistical analysis
The primary endpoint of this study was patient survival
after FDG PET-CT. Various descriptive statistics such as
mean, median, range, Standard Deviation (SD), and
Standard Error (SE) of mean have been used to summarize
the baseline clinical and demographic profiles of all the
patients. Survival curves were calculated with Kaplan–
Meier product-limit estimation, and survival in the distri-
butions was compared with the log-rank test for stratified
covariates and Cox proportional hazards interval model for
continuous covariates. A multivariate Cox model was then
constructed using a stepdown technique, starting with a
model containing all variables found to be significant in the
univariate analysis. Variables were retained in the model if
the p value of the variable was significant (p<0.05). For
comparison of hazard ratios for the FDG-PET score
between patient subgroups, the difference between hazard
ratios was computed with 95% confidence intervals and Z
scores based on the standard error estimates for the
individual hazard ratio. Two-sample t test is used for the
parametric data; Mann–Whitney test is used for nonpara-
metric data; and Fisher's exact test is used for categorical
data. All p values reported are two tailed. Statistical
package SPSS 11.0 (SPSS Inc., Chicago, Illinois, USA)
was used for statistical analyses.
Results
Patient and tumor characteristics
A total of 81 patients were enrolled in the study based on
the inclusion and exclusion criteria mentioned above. The
CharacteristicsGroupFrequency Percentage
SexMale
Female
<40 years
>40 years
Left
Right
Central
Single lobe
Multiple lobe
Corpus callosal
Posterior fossa
GBM
Grade III
Grade II
Grade I and others
Sx
Sx+RT
Sx+RT+CT
Debilitating symptoms
Mild symptoms
No treatment
Temozolamide
Surgery+/RT
59
22
48
33
34
40
7
53 (F-24,T-20, P-9)
16
7
5
13
26
34
8
7
42
32
47
34
46
30
5
72.8
27.2
59.3
40.7
42
49.4
8.6
65.4
19.8
8.6
6.2
16
32.1
42
9.9
8.6
51.9
39.5
58
42
56.8
37
6.2
Age
Lateralization of primary
Primary site of tumor
Grade of primary tumor
Primary treatment
Clinical symptoms at recurrence
Treatment after recurrence
Table 1 Patient characteristics
F frontal, T temporal, P parietal,
GBM glioblastoma multiforme,
Sx surgery, RT radiotherapy, CT
chemotherapy
Table 2 PET scoring
PET scoreFDG uptakeFrequency Percentage
0
1
2
No uptake or T/W≤1
T/W>1 and T/G<1
T/G≥1
31
29
21
38.3
35.8
25.9
T/W tumor to white matter ratio, T/G tumor to grey matter ratio
Neuroradiology (2011) 53:1017–10241019

Page 4

clinical data from the study are presented in Table 1. The
mean age at the time of FDG PET-CT scan was 36.8 years
(median—35 years, range: 12–66). Forty seven patients
presented with debilitating clinical symptom like severe
headache, recurrent seizure, and hemiparesis or paraparesis
at the time of evaluation for recurrence, whereas the rest of
Table 3 Results of log-rank test to compare survival within distributions of variables
VariableDistributionsFrequencyNumber of
deaths
Mean survival in
months
Log-rank
statistics
P
SexMale
Female
≤40 years
≥40 years
≤3 cm
≥3 cm
Left
Right
Central
1. Single lobe
2. Multiple lobe
3. Corpus callosal
4. Posterior fossa
1. Grade I
2. Grade II
3. Grade III
4. GBM
1. Sx/RT only
2. Sx±RT
3. Sx±RT±CT
Debilitating
Mild symptoms
1. No treatment
2. Temozolamide
3. Sx±/RT
0
1
2
59
22
48
33
50
31
34
40
7
53
16
7
5
8
34
26
13
7
42
32
47
34
46
30
5
31
29
21
22
8
16
14
14
16
10
15
5
18
5
2
5
2
10
13
5
4
11
5
24
6
18
12
0
4
11
15
22.48
20.13
23.11
20.01
23.25
19.63
22.73
22.31
14.16
23.55
21.25
18.01
15.17
22.04
23.63
18.43
22.29
11.41
24.92
19.95
19.08
25.09
20.14
22.03
0.000.984
Age 0.380.538
Size on MRI2.880.089
Lateralization of
primary
3.930.14
Primary site of tumor1
0.00
0.48
3.56
1
0.07
1.07
0.07
1
5.34
1.51
9.1
231
0.96
0.48
0.06
1
0.79
0.3
0.79
1
0.02
0.22
0.002*
23
0.26
1.39
2
0.61
0.24
2
0.08
3
0.77
3 Grade of primary
tumor
2.71
0.03
2
0.09
0.87
2
1.53 0.21
Primary treatment
received
2.01 0.16
Clinical symptoms
Treatment after
recurrence
1
0.00
2.7
0
3.21
22.3
21
0.98
0.10
0
0.07
0.0001*
2
3.04
1
0.08
1PET score 26.06
23.06
12.578.20.004*
GBM glioblastoma multiforme, Sx surgery, RT radiotherapy, CT chemotherapy
*p<0.05 significant
Fig. 1 Survival after positron emission tomography (FDG-PET)
scanning, stratified by FDG-PET score
Fig. 2 Survival after FDG-PETscan for patients with FDG-PETscores
of less than 2 (solid line) and those with scores of ≥2 (broken line)
1020Neuroradiology (2011) 53:1017–1024

Page 5

the patients had mild symptoms not affecting their daily
activities. Mean lesion size obtained from contrast enhanced
area in MRI was 3.41 cm (±0.40 cm) and median size was
3.3 cm (range 1.1–7.2 cm).
FDG PET-CT findings
FDG PET-CT findings were positive for recurrence in 50
patients and negative in 31 patients based on the interpre-
tation criteria mentioned in the “Materials and methods”.
PET scores of the patients were calculated as detailed in the
“Materials and methods” and are summarized in Table 2.
Follow-up outcome
All patients were followed up. Patient data was censored if
the patient died of disease or till last follow up. Survival
was defined as the time between the FDG PET-CT and the
date when data was censored. Median follow-up period was
16.33 months (range: 0.4–33.1). A total of thirty patients
died within the median follow-up period. Median survival
in this study was 24.97 months.
Survival analysis
Log-rank test and Kaplan–Meyer survival analysis among
distributions of different variables did not show any
significant association of survival with sex, age, size of
the recurrent tumor, site of the primary tumor, and
histological grade of the primary tumor (Table 3). Patients
primarily treated with combination of surgery and radio-
therapy had better survival than those treated with surgery
alone (p 0.02). However, mode of treatment after recurrence
showed no significant survival benefit. Patients having
debilitating clinical symptoms at the time of recurrence had
poorer survival than those having mild symptoms (p 0.002).
Patients having PET score 2 had poorer survival compared
to both PET score 0 (p 0.001) and PET score 1 (p 0.004).
However, there was no statistically significant difference
between survival of patients with PET score 0 and PET
score 1 (p 0.07). So, patients could be stratified into two
groups—PET score<2 and PET score≥2. Survivals accord-
ing to PET score are depicted in Figs. 1 and 2. Even on
multivariate analysis PET Score, clinical symptoms and
primary treatment were found to be significantly associated
with survival (Table 4). PET Score was found to be the
most significant predictor of survival (p 0.003) (Figs. 3, 4,
and 5). Size of lesion, which was a weak predictor on
univariate analysis was found not to be significant on
multivariate analysis.
Discussion
Malignant gliomas that recur after primary treatment carry
poorer prognosis even with aggressive treatment [24, 25].
In contrast, low-grade tumors have variable prognosis and
survival. FDG uptake was reported as a strong predictor of
Table 4 Multivariate analyses of survival
95% CI of HR
Covariatep HRLower Upper
Previous treatment
Clinical symptoms
PET score
0.006
0.011
0.003
0.250
4.617
3.364
0.093
1.411
1.507
0.670
15.109
7.512
P indicates level of significance, HR hazard ratio, CI confidence
interval
Fig. 3 A 40-year-old male patient with right fronto-temporal
anaplastic astrocytoma (Grade 3) primarily treated with surgery and
radiotherapy. The patient presented with severe headache. Transaxial
FDG-PET (a), CT (b), and PET-CT (c) images show a right fronto-
temporal hypermetabolic lesion suggestive of recurrence. T/W was
4.38 and T/G was 2. So, the PET score was 2. The patient was treated
with radiotherapy and temozolamide. The patient died after a follow
up of 9.9 months after PET-CT
Neuroradiology (2011) 53:1017–10241021

Page 6

survival in glioma patients. Reports limited to patients
imaged with FDG-PET at the time of imaging changes
suggesting recurrent tumor [26–29], have reported trends
toward longer survival in patients with tumors with low
glucose uptake, but no study has reported a statistically
significant difference in survival. Other groups have
reported significant correlations between FDG-PET score
and survival in patients with malignant gliomas imaged at
various times in the course of their disease. Alavi et al.
reported median survivals of 7 and 33 months in patients
with hypermetabolic and hypometabolic lesions, respec-
tively (p 0.0007) [23]. Patronas et al. reported median
survivals of 5 and 19 months in patients with hypermeta-
bolic and hypometabolic lesions, respectively (p 0.001)
[30]. However, given the heterogeneity of study designs,
there is still disagreement regarding the value of FDG-PET
in predicting survival in patients with glioma.
The goal of the current study was limited to a single
objective: whether FDG PET-CT predicts survival in
patients with glioma in whom imaging studies after the
initial treatment demonstrate changes that could represent
recurrence. Under this strictly defined setting, FDG-PET
identified two groups of patients with distinctly different
prognoses. The difference in prognosis was both statistically
and clinically significant, even after correction for other
prognostic variables, such as age, sex, grade of primary,
primary treatment, clinical symptoms, and treatment after
recurrence. On univariate analysis, survival was correlated
significantly with the PET score (p 0.001). After adjustment
for variables found to be significant predictors of survival
(clinical symptoms and primary treatment), in a multivariate
Cox model the PET score remained a significant prognostic
factor (p 0.003). PET score was a better predictor of survival
than clinical symptoms and primary treatment, which were
the two other significant predictors on multivariate analysis.
Moreover, our method of PET data analysis based on two
simple ratios (and subsequent three PET scores) is straight-
forward and can be easily reproduced at any clinical PET
centre.
The size of the recurrent lesions as measured in MRI
appeared to be weakly associated with survival in univar-
iate analysis (p=0.08) but not found statistically significant
Fig. 4 A 43-year-old male
patient with right frontal oligo-
dendroglioma (Grade 2) primar-
ily treated with surgery and
radiotherapy. The patient was
clinically asymptomatic. Recur-
rence was suspected on conven-
tional imaging. Transaxial FDG-
PET (a), CT (b), and PET-CT
(c) images did not show any
suspicious lesion suggestive of
recurrence. T/W was 1.14 and
T/G was 0.4. So, the PET score
was 1. The patient was not given
any treatment. The patient was
well till the end of the study
(16.5 months after PET-CT)
Fig. 5 A 66-year-old male patient with right temporal GBM (Grade
4) primarily treated with surgery, radiotherapy, and temozolamide. The
patient presented with mild headache. Recurrence was suspected on
conventional imaging. Transaxial FDG-PET (a), CT (b), and PET-CT
(c) images did not show any suspicious lesion suggestive of
recurrence. T/W was 0.97 and T/G was 0.35. So, the PET score was
0. The patient was not given any treatment. The patient was well till
the end of the study (24.5 months after PET-CT)
1022Neuroradiology (2011) 53:1017–1024

Page 7

in multivariate analysis. In the management of brain
tumors, accurate pathologic grading of the primary lesion
is essential, because histologic features are a major
independent prognostic factor next to patient age and
performance status [14, 31–34]. In our study, the primary
grade of the tumor was not a predictor of survival. This
may be because of the fact that most of the low-grade
tumors undergo anaplastic transformation at recurrence.
Hence, most of the recurrent tumors are high grade even
though the primary has been low grade. As histopathology
of recurrent tumors was not available in all patients, it can
only be assumed that this applied to the present study as
well. Treatment after FDG PET-CT scanning was another
source of heterogeneity in the present study. Patients
received a variety of treatments, guided by their clinical
and radiologic diagnoses. Although mode of treatment after
detection of recurrence did not show any significant
correlation with survival on univariate analysis (p 0.08),
the effect of these treatments on subsequent survival is
difficult to model. Cox models are not adequate to
distinguish the effect of a treatment on survival when the
treatment can be initiated at different times after the start of
the study period [35, 36], and the use of a Cox model in
this situation may produce strongly biased results [36, 37] .
So, we did not attempt to adjust for the effects of
subsequent treatments on survival after FDG PET-CT.
Conclusion
Intensity of FDG uptake on PET-CT is a strong predictor of
survival in patients with imaging findings of recurrent
glioma. FDG uptake in tumors higher than that of grey
matter is the worst prognostic indicator in these patients on
PET-CT imaging.
Conflict of interest
We declare that we have no conflict of interest.
References
1. Levin VA, Leibel SA, Gutin PH (1997) Management of central
nervous system neoplasms. In: DeVita VT Jr, Hellman S,
Rosenberg SA (eds) Cancer: Principles & Practice of Oncology.
Lippincott-Raven Publishers, Philadelphia, pp 2022–2082
2. Kleihurs P, Cavenee WK (eds) (2000) World Health Organization
Classification of Tumors: pathology and genetics of tumors of the
nervous system. Oxford University Press, New York
3. Kleihues P, Ohgaki H (1999) Primary and secondary glioblasto-
mas: from concept to clinical diagnosis. Neurooncology 1:44–51
4. Jung V, Romeike BFM, Henn W, Feiden W, Moringlane JRM,
Zang KD et al (1999) Evidence of focal genetic microheteroge-
neity in glioblastoma multiforme by area-specific CGH on
microdissected tumor cells. J Neuropathol Exp Neurol 58:993–
999
5. Central Brain Tumor Registry of the United States: 2005–2006.
Available at http://www.cbtrus.org
6. Bampoe JO, Bauman G, Cairncross JG (1999) Adult low-grade
gliomas: natural history, prognostic factors, and timing of
treatment. In: Rock JP, Rosenblum ML, Shaw EG, Cairncross
JG (eds) The practical management of low-grade primary brain
tumors. Lippincott Williams & Wilkins, Philadelphia, pp 135–148
7. Piepmeier J, Christopher S, Spencer D, Byrne T, Kim J, Knisel JP
et al (1996) Variations in the natural history and survival of
patients with supratentorial low-grade astrocytomas. Neurosurgery
38:872–879
8. McCormack BM, Miller DC, Budzilovich GN, Voorhees GJ,
Ransohoff J (1992) Treatment and survival of low-grade astrocy-
toma in adults—1977–1988. Neurosurgery 31:636–642
9. Scott JN, Rewcastle NB, Brasher PM, Fulton D, MacKinnon JA,
Hamilton M et al (1999) Which glioblastoma multiforme patient
will become a long-term survivor? A population-based study. Ann
Neurol 46:183–188
10. Bauman G, Lote K, Larson D, Stalpers L, Leighton C, Fisher B et
al (1999) Pretreatment factors predict overall survival for patients
with low-grade glioma: a recursive partitioning analysis. Int J
Radiat Oncol Biol Phys 45:923–929
11. Ricci PE (1999) Imaging of adult brain tumors. Neuroimag Clin N
Am 9:651–669
12. Goldman S, Wikler D, Damhaut P, Monclus M, Levivier M,
Pirotte B et al (1997) Positron emission tomography and brain
tumours. Acta Neurol Belg 97:183–186
13. Mirzaei S, Knoll P, Kohn H (2001) Diagnosis of recurrent
astrocytoma with fludeoxyglucose F18 PET scanning. N Engl J
Med 344:2030–2031
14. Goldman S, Levivier M, Pirotte B, Brucher JM, Wikler D,
Damhaut P et al (1996) Regional glucose metabolism and
histopathology of gliomas. A study based on positron emission
tomography guided stereotactic biopsy. Cancer 78:1098–1106
15. Langleben DD, Segall GM (2000) PET in differentiation of
recurrent brain tumor from radiation injury. J Nucl Med 41:1861–
1867
16. Chao ST, Suh JH, Raja S, Lee SY, Barnett G (2001) The
sensitivity and specificity of FDG PET in distinguishing recurrent
brain tumor from radionecrosis in patients treated with stereotactic
radiosurgery. Int J Cancer 96:191–197
17. Di CG, Oldfield E, Wright DC, De Michele D, Katz DA, Patronas
NJ et al (1988) Cerebral necrosis after radiotherapy and/or
intraarterial chemotherapy for brain tumors: PET and neuropath-
ologic studies. AJR Am J Roentgenol 150:189–197
18. Brock CS, Young H, O’Reilly SM, Matthews J, Osman S, Evans
H et al (2000) Early evaluation of tumour metabolic response
using [18F]fluorodeoxyglucose and positron emission tomogra-
phy: a pilot study following the phase II chemotherapy schedule
for temozolomide in recurrent high-grade gliomas. Br J Cancer
82:608–615
19. Roelcke U, von Ammon K, Hausmann O, Kaech DL, Vanloffeld
W, Landolt H et al (1999) Operated low grade astrocytomas: a
long term PET study on the effect of radiotherapy. J Neurol
Neurosurg Psychiatr 66:644–647
20. Herholz K, Holzer T, Bauer B, Schröder R, Voges J, Ernestus RI et
al (1998) 11C-methionine PET for differential diagnosis of low-
grade gliomas. Neurology 50:1316–1322
21. Pirotte B, Goldman S, Massager N, David P, Wikler D,
Vandesteene A et al (2004) Comparison of 18F-FDG and 11C
methionine for PET-guided stereotactic brain biopsy of gliomas. J
Nucl Med 45:1293–1298
22. De Witte O, Levivier M, Violon P, Salmon I, Damhaut P, Wikler D
Jr et al (1996) Prognostic value positron emission tomography
with [18F]fluoro-2-deoxy-D-glucose in the low-grade glioma.
Neurosurgery 39:470–477
Neuroradiology (2011) 53:1017–1024 1023

Page 8

23. Alavi JB, Alavi A, Chawluk J, Kushner M, Powe J, Hickey W et
al (1988) Positron emission tomography in patients with glioma.
A predictor of prognosis. Cancer 62:1074–1078
24. Levin VA, Prados MD (1992) Treatment of recurrent gliomas and,
metastatic brain tumors with a polydrug protocol designed to
combat nitrosourea resistance. J Clin Oncol 10:766–771
25. Warnick RE, Prados MD, Mack EE, Chandler KL, Doz F, Rabbitt
JE et al (1994) A phase II study of intravenous carboplatin for the
treatment of recurrent gliomas. J Neurooncol 19:69–74
26. Di Chiro G, DeLaPaz RL, Brooks RA, Sokoloff L, Kornblith PL,
Smith BH et al (1982) Glucose utilization of cerebral gliomas
measured by [18F]fluorodeoxyglucose and positron emission
tomography. Neurology 32:1323–1329
27. Ishikawa M, Kikuchi H, Miyatake S, Oda Y, Yonekura Y,
Nishizawa S (1993) Glucose consumption in recurrent gliomas.
Neurosurgery 33:28–33
28. Janus TJ, Kim EE, Tilbury R, Bruner JM, Yung WK (1993) Use of
[18F]fluorodeoxyglucose positron emission tomography in
patients with primary malignant brain tumors. Ann Neurol
33:540–548
29. Kahn D, Follett KA, Bushnell DL, Nathan MA, Piper JG, Madsen
M et al (1994) Diagnosis of recurrent brain tumor: value of 201T1
SPECT vs 18F-fluorodeoxyglucose PET. AJR Am J Roentgenol
163:1459–1465
30. Patronas NJ, Di Chiro G, Kufta C, Bairamian D, Kornblith PL,
Simon R et al (1985) Prediction of survival in glioma patients by
means of positron emission tomography. J Neurosurg 62:816–822
31. DeAngelis LM (2001) Brain tumors. N Engl J Med 344:114–123
32. Berg G, Blomquist E, Cavallin-Stahl E (2003) A systematic
overview of radiation therapy effects in brain tumours. Acta Oncol
42:582–588
33. Barker FG, Chang SM, Gutin PH, Malec MK, McDermott MW,
Prados MD et al (1998) Survival and functional status after
resection of recurrent glioblastoma multiforme. Neurosurgery
42:709–720
34. Kondziolka D, Flickinger JC, Bissonette DJ, Bozik M, Lunsford
LD (1997) Survival benefit of stereotactic radiosurgery for
patients with malignant glial neoplasms. Neurosurgery 41:776–
783
35. Wolfe RA, Strawderman RL (1996) Logical and statistical
fallacies in the use of Cox regression models. Am J Kidney Dis
27:124–129
36. Glesby MJ, Hoover DR (1996) Survivor treatment selection bias
in observational studies: examples from the AIDS literature. Ann
Intern Med 124:999–1005
37. Barker FG II (1996) Analysis of the relationship between long-
term operative success and a transient or delayed operative side
effect. Neurosurgery 39:412–416
1024Neuroradiology (2011) 53:1017–1024