close

Вход

Забыли?

вход по аккаунту

?

870

код для вставкиСкачать
1098
Regional Glucose Metabolism and Histopathology
of Gliomas
A Study Based on Positron Emission Tomography-Guided Stereotactic Biopsy
Serge Goldman, M.D.'
Marc Levivier, M.D?
Benoit Pirotte, M.D?
Jean-Marie Brucher, M.D.~
David Wikler, M.s.~
Philippe Damhaut Ph.0.'
Etienne Stanus, m o . '
Jacques Brotchi, M.D.'
Jerzy Hildebrand, M.O."
BACKGROUND. Positron emission tomography (PET) with '8FF-2-fluoro-2-deoxy-Dglucose (FDG) is widely applied to the study of gliomas. The histology of most
'
PET/Biomedical Cyclotron Unit, ULB-HBpital
Erasme, Brussels, Belgium.
Service de Neurochirurgie,
Erasrne, Brussels, Belgium.
ULB-HBpital
Department of Neuropathology, Cliniques Universitaires Saint-Luc, Brussels, Belgium.
Service de Neurologie, ULB-HBpital Erasme,
Brussels, Belgium.
A portion of this article has been presented at
the 45th Annual Meeting of the American Academy of Neurology, New York, April 25-May 1,
1993. The abstract of this presentation appeared in Neurology 1993; 43:A249.
Supported by research grants 9.4503.91 and
9.4513.93F from the Belgian National Lottery
and 3.4509.92, 3.4508.92 and 3.4532.94 from
the National Funds for Scientific Research
(FNRS, Belgium).
Address for reprints: Serge Goldman, M.D.,
PET/Biomedical Cyclotron Unit, ULB-HBpital
Erasme, 808, route de Lennik, B-1070 Brussels,
Belgium.
Received April 2,1996; accepted May 21,1996.
0 1996 American Cancer Society
gliomas is regionally heterogeneous. The relationship between histologic features
and glucose metabolism evaluated by PET with FDG may therefore vary within
the limits of the tumor. PET with FDG integrated in the planning of stereotactic
brain biopsy allows precise comparison between local FDG uptake and histology.
Using this approach, the authors investigated whether glucose metabolism of gliomas is related to anaplasia, and whether PET with FDG detects metabolic heterogeneity that parallels histologic heterogeneity of gliomas.
METHODS. A total of 161 biopsy samples collected from 20 PET-guided procedures
performed in patients with gliomas (8 low grade astrocytomas, 8 anaplastic astrocytomas, 1 anaplastic oligoastrocytoma, and 3 glioblastomas) were analyzed for the
presence or absence of 8 histologic features. Stereotactic coordinates were used
to calculate the metabolic rate of glucose (MRGlu) in the region of each biopsy
sample. Gray and white matter MRGlu were used to define four metabolic grades
that were compared with local histology.
RESULTS. The difference in MRGlu expressed as micromoles per 100 g per minute
was highly significant between anaplastic and nonanaplastic samples; the median
c quartile deviation was 23 f 16 in anaplastic samples and 18 2 5 in nonanaplastic
samples ( P < 0.005). Even more significant differences were found when MRGlu
was normalized to the cortex or to the white matter. Metabolic grades were different in anaplastic and nonanaplastic samples ( P < 0.0001). Approximately 75% of
samples metabolically graded 3 or 4 demonstrated signs of anaplasia. compared
with 10% of samples graded 0 or 1.
CONCLUSIONS. FDG uptake in gliomas is anatomically heterogeneous and is regionally related to the presence of anaplasia. Cancer 1996; 78:1098-1106.
0 1996 American Cancer Society.
KEYWORDS glioma, brain, brain neoplasm, positron emission tomography, glucose
metabolism, fluorodeoxyglucose, stereotaxy, brain biopsy, neuropathology.
P
ositron emission tomography (PET) with '*F-2-fluoro-2-deoxy-Dglucose (FDG) is applied to the management of gliomas and is
especially useful for detection of tumor recurrence and prognosis
estimation.'-' In most PET with FDG (FDG-PET) studies, metabolic
grading of brain tumors is based on the highest level of FDG uptake
within the presumed limits of the tumor.'"' This approach takes into
account the well established cellular, histologic, and metabolic heterogeneity of brain tumors.""* It assumes that local distribution of
FDG uptake in a single tumor parallels the heterogeneity of biologic
characteristics, such as proliferative activity or cellular density, which
Local FDG Uptake and Histology in Gliorna/Goldrnan et al.
1099
TABLE 1
Patient Characteristics
Patient
no.
Age
(F)
Gender
1
2
3
4
5
6
7
8
52
40
58
32
53
26
46
47
9
65
10
11
12
13
14
15
16
17
18
19
20
41
78
33
M
M
M
F
M
F
M
F
hl
M
F
F
F
F
F
M
M
M
M
M
50
32
33
62
44
48
76
23
No. of
samplesa
11
10
10
9
1
6
11
7
9
7
13
5
7
10
8
6
7
8
9
I
Location of lesion
Diagnosis
Left temporoparietal
Left thalamus
Right frontoparietal
Left temporoparietal
Left cerebellar peduncle
Right frontal
Left thalamus
Right frontal
Right temporoparietal
Right thalamus
Left parietal
Right cerebellum
Left frontoparietal
Right subinsular
Left parietal
Left frontal, corpus callosum
Left frontotemporal
Right temporal
Left parietal, corpus callosum
Left temporal
AA
AA
LG
LG
GB
LG
k4
LG
GB
GB
AA
AA
LG
LG
AA
AAt
AA
LG
AA
LG
M: male: F: female; LG: low grade glionia; M:anaplastic astrocytoma or tanaplastic oligoastrocytoma. GB: glioblastoma
Total number of samoles obtained froin I to 3 tragectories.
a
might be significant in terms of diagnosis and prognosis. This assumption derives from general postulates
relating level of tumor malignancy and degree of glycolysis' '-'(' and from animal autoradiographic experiments on the local distribution of FDG uptake in systemic ~ a n c e r s . ' ~The
~'~~
relationship
'~
between FDG
uptake and local biologic features of human gliomas
cannot be precisely studied in resected tumors because anatomic matching of regions defined on PET
and in resected tumors is difficult. Stereotactic PET
performed for the guidance of stereotactic biopsy offers a unique opportunity to study the local relationship between metabolic and histologic features in
brain tumors. We have applied this approach to a large
series of biopsy samples obtained during PET-guided
stereotactic biopsies in patients with glioma.
PATIENTS AND METHODS
A consecutive series of 22 patients suspected of having
a brain tumor gave their informed consent to undergo
combined FDG-PET- and computed tomography
((=TI-guided stereotactic biopsies after a procedure
that allowed imaging data acquisition (stereotactic CT
and PET), surgical planning, and biopsies on the same
day, as described in detail e1~ewhere.l~
Patients reported in this study are part of a larger series of patients in whom a study of the diagnostic yield of ste-
reotactic biopsy guided by FDG-PET was conducted
in accordance with the ethical guidelines of our institution." In all patients from the current study, quantification of glucose metabolism was rendered possible
thanks to the insertion of an arterial line allowing
tracer kinetics determination in arterial blood (input
function). A diagnosis of glial tumor was made on reliable histologic data in 20 patients whose characteristics are summarized in Table 1. One patient with a
final diagnosis of anaplastic astrocytoma was excluded
from the analysis because of poor quality of histologic
samples due to overfixation. The other patient excluded from the analysis had a nontumorous lesion
(necrotic lesion from vascular origin). Tumors were
graded based on the histologic criteria of the World
Health Organization classification."
In all patients, we performed FDG-PET in stereotactic conditions following a protocol previously deThe patients were fasted and conscious,
in a supine resting state with eyes closed and ears
unplugged. The PET tomograph was a CTI-Siemens
933/08-12 (CTI-Siemens, Knoxville, TN). The 15 6.75mm thick adjacent slices (8 direct and 7 crossed slices)
covered the entire brain. In-plane spatial resolution
(full width at half-maximum) was about 5 mm. During
PET acquisition, the stereotactic head ring was secured
to the clamp that fits into both the CT and the PET
1100
CANCER September 1, 1996 / Volume 78 I Number 5
couch. The fiducial reference system used for stereotactic PET consisted of V-shape tubing filled with an
[‘8Flflu~ridesolution adapted on four localization
plates originally designed for stereotactic magnetic
resonance imaging. Before each emission scan, a
transmission scan was obtained using a ring source
filled with a ”F-fluoride solution and allowing a measured correction of the images for attenuation. The
patients were injected intravenously with a bolus of
2260 megabecquerel (MBq) ( 2 7 millicurie [mCi]) of
FDG prepared following the method of Hamacher et
a1.22A 25-frame dynamic scan was acquired from 0 to
60 minutes postinjection with the last 20-minute
frame starting 40 minutes postinjection. Between 15
seconds to 50 minutes after intravenous injection, 21
serial blood samples were taken from a contralateral
radial artery. Blood samples were centrifuged and
plasma was counted in a gamma counter. Glucose
concentration was determined in the plasma of 4
blood samples at 0, 10, 20, and 40 minutes and used
to calculate metabolic rates of glucose (MRGlu).Every
day, the PET camera was calibrated using a cylindrical
phantom filled with a known concentration of 68Ge/
“Ga to convert brain counts into brain radioactivity
content. Using corrected emission scan data, plasma
radioactivity time course, and plasma glucose concentration, MRGlu were measured using both a graphic
method previously applied in the literature to the
study of tumor glucose metabolismz3and with the CTISiemens software according to the model of Sokoloff et
al.24adapted by Phelps et a1.,25which has been recently
validated and used for the study of brain tumor glucose rnetaboli~m.~~.~’
Because our analysis revealed a
high correlation between MRGlu values calculated by
the 2 methods (correlation coefficient (r) = 0.98; P =
O.OOOl), we only used the values obtained by the
graphic method for the presentation of the results. The
lumped constant that corrects for the differences in
transport and phosphorylation rates of glucose and
FDG was assumed to be 0.42 and MRGlu values were
expressed in micromoles (pmol) per 100 g per minute.
All PET data analyses were performed blinded to the
histology of the tumor. To calculate MRGlu at the level
of each biopsy sample site, circular 0.3-cm2regions of
interest (ROI) were centered on the coordinates of the
actual biopsy recorded postoperatively and transferred on the stereotactic PET using a local implementation of the PET processing software (Fig. 1). The
main advantage of this procedure, compared with
usual neurooncologic PET data analyses, is that it
eliminates most subjective influences of the PET investigator with regard to ROI size or placement.
As previously described,” we evaluated MRGlu in
noninvolved gray and white matter with ROIs deline-
ated in the frontal and temporal cortex of the hemisphere contralateral to the tumor and in white matter
areas. Cortical and subcortical areas in the involved
hemisphere were not included in the analysis because,
as described in the literature,’ they were inconstantly
hypometabolic for multiple reasons (mass effect, deafferentation, and infiltration). We calculated the surface-weighted mean MRGlu value in the ROIs distributed in the cortex and in the ROIs distributed in the
white matter. A “metabolic grade” was defined to classify the samples based on MRGlu relative to the patient’s cortex and white matter values (Fig. 2).
In the 20 patients included in the analysis, a total
of 161 cylindrical biopsy specimens (diameter: 1 mm,
length: 1 cm) were obtained from the serial stereotactic biopsies performed with a side-cutting cannula,
following the technique described by Kelly et al.29Trajectory targets were selected as previously described.”
In short, we selected, when present, areas of FDG uptake higher than in the surrounding normal-appearing
white and gray matter, or foci of relatively increased
FDG uptake in an otherwise hypometabolic lesion.
Targets selected on PET were projected onto the corresponding stereotactic CT slice to control the reliability
and safety of the target selection and the trajectory.
The serial sampling along the biopsy trajectory ensured the collection of samples outside the areas of
high FDG uptake, nearer and farther than the metabolic targets. Samples from nonhypermetabolic areas
were also obtained from CT-guided trajectories, chosen as second or third trajectories when a PET-guided
trajectory was defined, or as the primary trajectory
when no metabolic target could be defined. After formalin fixation and embedding of the samples, serial
sections were obtained and stained with hematoxylin
and eosin, Masson’s trichrome, and some immunostains (glial fibrillary acidic protein, S-100 protein,
neuron specific enolase, vimentin, and others) when
ne~essary.~’
Presence or absence of eight histologic
features was recorded on all samples: tumor tissue,
anaplastic tumor tissue, infiltrating tumor cells, brain
tissue included in the specimen, edema, gliosis, focal
necrosis, and extensive necrosis. Several features
could be present in the same sample.
Statistical analysis involved Mann-Whitney U
tests and Kruskal-Wallis tests. For variables with nongaussian distribution, measurement of central tendency and measurement of variation were the median
and semiinterquartile range (quartile deviation), respectively. To test ordinal by ordinal contingency tables, we used an adapted chi-square test that evaluated the linearly increasing frequency component (linear regression).31 Bonferroni correction for multiple
comparisons was applied when mentioned.
Local FDG Uptake and Histology in Gliorna/Goldrnan et al.
1101
FIGURE 1. Anaplastic astrocytoma. Two successive positron emission tomography (PET) planes (A and B, from rostra1 to caudal) with
regions of interest centered on the coordinates of four biopsy samples that all comprised tumor tissue. In this patient, three samples of the
same trajectory were on a single PET plane (B) because of the particular orientation of the trajectory close to the imaging angle.
I Grade
I
0
I
2
1
WMI2
WM
3
CX
4
2CX
I
I
I
FIGURE 2. Metabolic grade scale: a metabolic grade from 0 to 4 is
attributed to each sample based on its metabolic rate of glucose (MRGIu)
compared with MRGlu in the white matter (WM) and the cortex (CX). Four
thresholds delimit the metabolic grades; WM/2 indicates CMRGlu half the
value in the WM and 2 CX indicates a value twice the value in the CX.
RESULTS
The 161 samples analyzed, 34 originated from noninvaded brain tissue. Eight of these samples had a metaholic grade of l , 20 samples had a metabolic grade of
2,, and 6 samples had a metabolic grade of 3.
The remaining 127 histologic specimens sampled
tumor tissue with (n = 31) or without (n = 96) typical
anaplastic changes. The other histopathologic features
were less frequently encountered: 11 samples were in
surrounding brain tissue infiltrated by tumor cells, 6
samples demonstrated microfoci of necrosis, 5 samples showed extensive necrosis, 10 samples contained
gliosis, and 4 samples demonstrated signs of edema.
Our analysis concentrated on the relationship between
glucose metabolism and the presence of anaplasia in
tumor samples; none of the other histopathologic features, which usually occurred in combination, were
frequent enough to allow independent statistical analysis.
By pathologic definition, all 57 tumor samples in
low grade gliomas (LG) were nonanaplastic. Thirtynine nonanaplastic and 26 anaplastic samples were
obtained from the group “AA”, which was comprised
of 8 anaplastic astrocytomas and one anaplastic oligoastrocytoma. All five samples obtained in glioblastomas (GB) were anaplastic.
The statistical difference between the distribution
of anaplastic and nonanaplastic samples according to
the metabolic grade was highly significant (chi-square
= 26.5; degree of freedom [DFI = 2; P = 0.0001) and
essentially related to a highly significant linear regression (chi-square linear regression = 20.4; DF = 1; P =
0.0001; Figure 3). The percent of anaplastic samples
increased from metabolic Grade 0 (0%) to 4 (100%).
Approximately 12% (6 of 49) of samples metabolically
CANCER September 1, 1996 / Volume 78 / Number 5
1102
70
0
1
Non Anaplastic
Anaplastic
“j
50
cu
0
Yl
20
10
0
1
0
3
2
4
Metabolic Grade
FIGURE 3. For each metabolic grade, number of samples separated by
the presence (gray boxes) or absence (white boxes) of histologic signs
of anaplasia. Proportion of anaplastic samples increases significantly with
increasing metabolic grade.
TABLE 2
Metabolic Grades of Glucose Metabolism in Tumor Samples
According to Pathology
Metabolic
grade’
LG
AA
GB
Total
0
1
0
29
0
25
3
0
57
0
2
3
2
47
2
2
18
35
8
2
65
5
3
4
Total
0
62
14
2
127
LG: low grade glioma; AA: anaplasric astrocytoma or oligoastroqtoma; G B glioblastoma.
’See text and Figure 2 for definition of metabolic grade and statistical analysis.
graded 0 or 1were anaplastic, whereas this proportion
reached 75% (12 of 16) for samples metabolically
graded 3 or 4. The distribution of sample metabolic
grades also vaned significantly in the pathologic diagnosis, essentially because of a linear regression between metabolic grade progression and diagnostic
progression from the nonanaplastic astrocytomas to
the anaplastic astrocytomas (chi-square linear regression = 5.3; DF = 1; P = 0.05; Table 2).
The difference between MRGlu values in anaplastic and nonanaplastic samples was highly significant
when expressed in absolute values (Mann-Whitney U
test, P < 0.005; Table 3) or in values normalized to the
noninvaded gray or white matter ( P = 0.0001). Despite
these statistical differences, there was an overlap in
absolute and normalized values between anaplastic
and nonanaplastic samples (illustrated by TUlCX in
Figure 4).
MRGlu also varied in absolute (Kruskal-Wallistest,
P < 0.01) or normalized values ( P = 0.0001) when
tumor samples were grouped according to pathologic
grading (Table 3). In MRGlu absolute values, 2-by-2
Mann-Whitney U tests with Bonferroni correction revealed a significant difference between LG and GB ( P
< 0.05). In values normalized to the cortex or the white
matter, differences between LG and GB were also
highly significant ( P < 0.005). Differences between LG
and AA were significant for values normalized to the
cortex and to the white matter ( P < 0.0005). Differences between AA and GB were significant for values
normalized to the cortex ( P < 0.05).
DISCUSSION
This study demonstrates a local relationship between
the level of FDG uptake detected by PET and the presence of anaplastic changes in human gliomas.
Our analysis is based on in vivo metabolic measurements in brain sites in which the presence of tumor has been controlled histologically. Therefore, it
provides a reliable estimation of glucose metabolism
in glial tumors. The study demonstrates statistical differences between the MRGlu values calculated in anaplastic and nonanaplastic tumor zones. These differences are detected for absolute values and for values
normalized to the cortex or to the white matter, validating the simplified semiquantification for this type
of analysis. Visual inspection of FDG-PET images,
which relies on contrast between the tumor and the
white and gray matters, should similarly allow adequate localization of anaplastic regions in heterogeneous brain neoplasms. However, the overlap between
the distributions of metabolic activities in anaplastic
and nonanaplastic areas indicates that some nonanaplastic areas present with elevated metabolism and
that, conversely, anaplastic areas may present with a
low level of FDG uptake. With regard to the presence
of high uptake of FDG in brain tumor areas that are
histologically considered as nonanaplastic, preliminary data from our center indicate that this type of
finding may relate to a prognostic value of PET in LG
because nonanaplastic gliomas with foci of high FDG
uptake lead to statistically shorter survival than homogeneously hypometabolic turn or^.^' Variation of FDG
uptake in anaplastic samples was large and part of
these samples demonstrated low FDG uptake. This
might indicate that factors other than those that define
the histologic characteristics of anaplasia influence the
uptake of the tracer. To elucidate this point, the procedure described in this article should be used to further
compare PET data with biochemical features accessi-
local FDG Uptake and Histology in Glioma/Goldman et al.
1103
TABLE 3
Glucose Metabolism in Tumor Samples According to Anaplasia and Pathologic Grading (Median 2 Quartile Deviation)
~
MRGlu
TU / cx
TUIWM
~
No anaplasia
Anaplasia
PA
LG
AA
GB
Pb
18 5 5
0.5 ? 0.1
1.1 5 0.3
23 i 16
0.9 t 0.4
1.6 ? 0.7
< 0.005
18 5 4
0.5 i 0.1
0.9 i- 0.2
20 5 7
0.6 i 0.2
1.3 i 0.3
50 5 15
1.6 2 0.4
3.0 ? 0.8
< 0.01
< 0.0001
< 0.0001
<0.0001
<0.0001
hlHGlu: metabolic rates ofglecose [pno11100g/min): TUICX MRGlu ratio of sample 10 cortex; TUIWM: MRGlu ratio of sample to white matter. LG: low grade glioma; AA: anaplastic astrocyloma or oligoastrocytoma;
C.B: ~lioblastoma.
“ P valurs tram of Mann-Whitney U test.
P valuer from Kruskal-Wallis test.
2.5
0
0
x
0
1.5-
Y;>
b
1-
0.5
I
I
Anaplasia
No Anaplasia
FIGURE 4. Scatterplot of the metabolic rates of glucose normalized to
the cortex (TU/CX) in samples with and without histologic signs of anaplasia. Bars indicate the interquartile range (Mann-Whitney U test: P <
0.0001).
ble in vitro on biopsy samples. The overlap between
metabolic values calculated in PET ROIs centered on
{.he stereotactic location of anaplastic and nonanaplastic tumor zones may also partly result from technical limitations inherent to comparisons of in vivo metabolic results with in vitro histologic data. The correspondence between the center of the PET ROIs and
the center of the biopsy sample may be estimated from
the phantom experiments that have validated the stereotactic PET procedure.lg This validation has demonrstrated that the localization accuracy of stereotactic
PET is within 2.5 mm, the in-plane pixel size of the
‘system.To maintain this level of localization accuracy,
no reslicing of the PET images is performed so that
the angle between the biopsy trajectories and the PET
planes is variable. This variable angle influences the
relationship between the PET volume in which the
quantification is performed and the sample volume
that is studied histologically. Consequently, the PET
volumes analyzed may contain a variable proportion
of tissue not submitted to histologic analysis and, in
some cases, part of the histologic specimen will not
be included in the PET volume of interest. Nevertheless, this type of error is expected to occur randomly
and, therefore, probably play a role in the overlap between the anaplastic and nonanaplastic samples values but does not invalidate the statistical differences
found.
The study emphasizes the need for careful definition and description of tumor zones in which PET
quantification of glucose metabolism is performed. In
particular, the results indicate that measurements on
“whole tumor area” (defined on PET or on other imaging modalities) and “hot spot” analysis may lead to
different results and interpretation, mainly in the more
heterogeneous anaplastic astrocytomas. In agreement
with this view, a recent study showed better correlation between FDG uptake and tumor grade with “hot
spot” than with “whole tumor” analysis.33As another
consequence, our analysis confirms that PET-guidance of stereotactic biopsies toward zones of high FDG
uptake might increase the diagnostic yield of these
procedures by reducing the risk of grade underestimation attributable to histologic heterogeneity of glial tumors.11”2
We have previously described our experience
in 43 patients with this procedure,“’ which has also
been applied or proposed by other a ~ t h o r s . ~One
~~~~-~
previous study based the potential usefulness of PETguidance of brain biopsies on results obtained by
matching PET with stereotactic CT scans performed
for the guidance of stereotactic biop~ies.‘~
In a carefully selected group of eight patients with nonanaplastic astrocytoma and two patients with anaplastic astrocytoma, this study showed a correlation between glucose consumption and cell density. Due to the small
number of patients with anaplastic changes, no rela-
1104
CANCER September 1, 1996 / Volume 78 / Number 5
tionship with anaplasia was searched for in this study.
Therefore, our study represents an original attempt
to precisely establish which is the local relationship
between glucose metabolism and a major histologic
criterion of tumor classification: the presence or absence of anaplastic characteristics.
Anaplasia represents a major predictor of survival
in glial tumors.'"-'" A prognostic value independent of
the histologic grade has also been claimed for FDGPET in brain neoplasm~.""~'~'The relationship between local glucose consumption and anaplasia found
in this study might relate, at least partially, the prognostic value of FDG-PET to the abundance or the intensity of some aspects of the anaplastic changes. The
link between glucose consumption and anaplasia
probably originates in the abnormal induction of glycolysis in highly malignant cells. This phenomenon,
attributed to overexpression of glucose transporters, 15,16,4i seems to be related to the proliferative activity because, in various models, FDG uptake was correlated with the proportion of proliferating
It
is worth mentioning that local FDG uptake in tumors
may be influenced by factors other than tumor cell
metabolism. Several studies on experimental neoplasms have indeed emphasized the major contribution of prenecrotic and inflammatory cells to the increase in FDG ~ p t a k e , 'and
~ . ~FDG
~ uptake detected by
PET in brain lesions with inflammatory components
support these experimental data.45-48
Cellular changes
in the prenecrotic phase may be undetectable by light
microscopy and require electron microscopic demonstration." Therefore, these changes may have been
overlooked in our analysis. Furthermore, because our
series includes a minority of patients with GB, samples
in which focal or extensive necrosis is found are too
rare to reveal the effect of this type of feature on FDG
uptake. As expected, glucose metabolism in samples
containing noninvaded brain tissue is essentially
within the range of white matter to gray matter glucose
metabolism. This level of metabolism probably depends mainly on the variable proportion of gray and
white matter, as well as on neural disconnection and
mass effects.',4' The influence of sparse infiltrating
cells on this variable metabolism could not be assessed
due to the small number of samples with this histologic feature.
The simple metabolic grading scale based on glucose metabolism in the white and gray matter statistically discriminates anaplastic from nonanaplastic
samples. The same metabolic grading also differentiated samples coming from LG, AA, or GB. These results
support recent data from the literature demonstrating
differentiation of low grade and high grade gliomas
using cutoff levels based on white matter and cortical
FDG uptake.'" Nevertheless, the overlap between the
metabolic activity in samples from the different groups
of tumors calls for caution in defining metabolic cutpoints. It confirms the absolute need for histologic
confirmation of diagnoses based on morphologic and
metabolic imaging. MRGlu values in all tumor samples, in particular when normalized to the white matter or to the cortex, were also statistically different in
tumors with a pathologic Grade 2 (LG), 3 (AA),or 4
(GB). The distinction between low grade and high
grade tumors directly derived from the absence of anaplasia in the former group of tumors, but the differences found between the AA and the GB (which should
be confirmed in a larger group of patients) pointed to
the possible influence of other cellular or histologic
factors that remained to be identified. Again, if verified
in further studies, this difference in FDG uptake between AA and GB might partially explain the prognostic value of FDG-PET found in several studies which
included, in the same group, patients with Grade 3
and Grade 4 tumors. Survival of patients treated for
an anaplastic astrocytoma is indeed more variable and
statistically better than survival of patients treated for a
gliobla~toma.~~~""""
Metabolic grades of samples from
Grade 2 and Grade 3 tumors largely overlaped, mainly
because a substantial proportion of samples from
Grade 3 tumors presented no characteristics of anaplasia, a fact that again underlines the important metabolic and histologic heterogeneity of this type of tumor and emphasizes the need for the use of the highest level of tracer uptake within the limits of a tumor
to define its metabolic status.
In conclusion, this study demonstrates anatomic
heterogeneity in FDG uptake and the influence of anaplastic changes on regional glucose metabolism detected by FDG-PET in human gliomas. A simple semiquantitative metabolic grading scale helps to differentiate anaplastic from nonanaplastic samples. These
results support the use of stereotactic FDG-PET to improve the diagnostic yield of stereotactic biopsies of
brain tumors.
REFERENCES
Di Chiro G, DeLaPaz RL, Brooks RA, Sokoloff L, Kornblith
PL, Smith BH, et al. Glucose utilization of cerebral gliornas
measured by ["FI fluorodeoxyglucose and positron emission tomography. Neurology 1982;321323-9.
Di Chiro G, Brooks RA,Patronas NJ, Bairamian D, Kornblith
PL, Smith BH, et al. Issues in the in vivo measurement of
glucose metabolism of human central nervous system tumors. A H HNeurol 1984; 15:S138-46.
Patronas NJ, Di Chiro G, Kufta C, Bairarnian I), Kornblith
PL, Simon R, et al. Prediction of survival in gliorna patients
by means of positron emission tomography. J Nezrrosurg
1985;62:816-22.
Local FDG Uptake and Histology in Glioma/Goldman et al.
4.
5.
6.
'7.
8.
9.
10
11
12
13
14
15
16.
17.
18.
19.
20.
21.
c,c,
LL.
Alavi JB, Alavi A, Chawluk J, Kushner M, Powe J, Hickey W,
et al. Positron emission tomography in patients with glioma.
A predictor of prognosis. Cancer 1988;62:1074-8.
Di Chiro G. Positron emission tomography using [18Fl fluorodeoxyglucose in brain tumors. A powerful diagnostic and
prognostic tool. lnuesf Radiol 1987;22:360-71.
Francavilla TL, Miletich RS, Di Chiro G, Patronas NJ, Rizzoli
HV, Wright DC. Positron emission tomography in the detection of malignant degeneration of low-grade gliomas. Neurosurgeiy 1989;24:1-5.
Glantz MJ. Hoffman JM, Coleman RE, Friedman AH, Hanson
MW, Burger PC, et al. Identification of early recurrence of
primary central nervous system tumors by [18F] fluorodeoxyglucose positron emission tomography. Ann Neurol
1991;29:347-55.
Kim C K , Alavi JB, Alavi A, Reivich M. New grading system
of cerebral gliomas using positron emission tomography
with €:-I8 fluorodeoxyglucose. 1Neurooncol 1991; 10:85-91.
Coleman RE, Hoffman JM, Hanson MW, Sostman HD,
Schold SC. Clinical application of PET for the evaluation of
brain tumors. J Nucl Med 1991;32:616-22.
Delbeke D, Meyerowitz C, Lapidus RL, Maciunas R1, Jennings MT. Moots PL, et al. Optimal cutoff levels of F-18
fluorodeoxyglucose uptake in the differentiation of lowgrade from high-grade brain tumors with PET. Radiology
1995; 195:47-52.
Paulus W, Peiffer J. Intratumoral histologic heterogeneity of
gliomas: a quantitative study. Cancer 1989;64:442-7.
Russel DS, Rubinstein LJ. Pathology of tumours of the nervous system. 5th edition. London: Arnold, 1989.
Paul R, Johansson R, Kellokumpu-Lehtinen PL, Soderstrom
KO, Kangas L. Tumor localization with 18F-2-fluoro-2deoxy-D-glucose: comparative autoradiography, glucose 6phosphatase histochemistry, and histology of renally implanted sarcoma of the rat. Res Exp Med (Berl) 1985;185:8794.
Minn H, loensuu H, Ahonen A, Klemi P. Fluorodeoxyglucose
imaging: a method to assess the proliferative activity of human cancer in viva. Comparison with DNA flow cytometry
in head and neck tumors. Cancter 1988;61:1776-81.
Warburg 0. The metabolism of tumors. London: Arnold
Constable, 1930:75-327.
Warburg 0. On the origin of cancer cells. Science
1956; 123:309-14.
Kubota R, Kubota K, Yamada S, Tada M, Ida T, Tamahashi
N. Active and passive mechanisms of [fluorine-I8]fluorodeoxyglucose uptake by proliferating and prenecrotic
cancer cells in viva: a microautoradiography study. J Nucl
Metl 1994;35:1067-75.
Kuboxa R, Kubota K, Yamada S, Tada M, Takahashi T, Iwata
R, et al. Methionine uptake by tumor tissue: a microautoradiogriiphic comparison with FDG. 1Nucl Med 1995;36:48492.
Levivier M. Goldman S. Bidaut LM, Luxen A, Stanus E,
Przedborski S, et al. Positron emission tomography-guided
stereotactic brain biopsy. Neurosurgery 1992;31:792-7.
Levivier M, Goldman S, Pirotte B, Brucher J, Baleriaux D,
Luxen A, et al. Diagnostic yield of stereotactic brain biopsy
guided by positron emission tomography with [18F]fluorodeoxyglucose. J Neurosurg 1995;82:445-52.
Kleihues P, Burger PC, Scheithauer BW. Histological typing
of tuttiours of the central nervous system. 2nd edition. Berlin: Springer-Verlag, 1993:l-37.
Hamacher K, Coenen HH, Stocklin G. Efficient stereospecific
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
1105
synthesis of no-carrier-added 2-["'F]fluoro-2-deoxy-D-glucase using aminopolyether supported nucleophilic substitution. J Niccl Med 1986;27:235-8.
Messa C. Choi Y, Hoh CK, Jacobs EL, Glaspy ]A, Rege S, et
al. Quantification of glucose utilization in liver metastases:
parametric imaging of FDG uptake with PET. J Cornput Assist
Tomogr 1992;16:684-9.
Sokoloff L, Reivich M, Kennedy C, Des Rosiers MH, Patlak
CS, Pettigrew KD, et al. The [ 1 4 C ] d e ~ ~ g method
l u ~ ~ ~for
e
the measurement of local cerebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. 1Neurochem 1977;28:897-916.
Phelps ME, Huang SC, Hoffman EH, Selin C, Sokoloff L,
Kuhl D. Tomographic measurement of local cerebral glucose
metabolic rate in humans with (F-18)2-fluoro-2-deoxy-Dglucose: validation of method. Ann Neurol 1979;6:371-88.
Herholz K, Rudolf J, Heiss WD. FDG transport and phosphorylation in human gliomas measured with dynamic PET.
J Neurooricol 1992; 2 1 5 9 - 6 5 .
Herholz K, Pietrzyk U , Voges 1, Schroder R, Halber M, Treuer
H, et al. Correlation of glucose consumption and tumor cell
density in astrocytomas-a stereotactic PET study. J Neurosurg 1993;79:853-8.
De Witte 0, Hildebrand J, Luxen A, Goldman S. Acute effect
of carmustine o n glucose metabolism in brain and glioblastoma. Cancer 1994;74:2836-42.
Kelly PI, Daumas-Duport C, Kispert DB, Kall BA, Scheithauer
BW, Illig JJ. Imaging-based stereotactic serial biopsies in untreated intracranial glial neoplasms. J Neurosurg 1987;
66:865-74.
Brucher 1-M. Neuropathological diagnosis with stereotactic
biopsies. Possibilities. difficulties and requirements. Acta
Neuroclzir (Wien) 1993;124:37-9.
Sachs L. Applied statistics. A handbook of techniques. 2nd
edition. New York Springer-Verlag, 1984:462-74.
Goldman S, Dewitte 0, Levivier M, Violon P, Damhaut P,
Wikler D, et al. Glucose uptake in glioma: relationship to
local histology and prognostic value in low grade tumor labstractl. Proceedings of the Eleventh International Conference on Brain Tumor Research and Therapy, Napa, California, October 31-November 3, 1995.
Junck L, McKeever PE, Ross DA, Greenberg HS, Brunberg
JA, Betley AT. The significance of fluorodeoxyglucose uptake
in gliomas [abstract]. J CereO Blood Flow Metab 1995;
155737.
Bergstrom M, Collins VP, Ehrin E, Ericson K, Eriksson L,
Greitz T, et al. Discrepancies in brain tumor extent as shown
by computed tomography and positron emission tomography using [68GalEDTA, [ l lC]glucose, and il 1Clmethionine.
1 Comput Assist Tornogr 1983;7:1062-6.
Hanson MW, Glantz MI. Hoffman JM, Friedman AH, Burger
PC, Schold SC, et al. FDG-PET in the selection of brain lesions for biopsy. J C o n ~ p uAssist
t
Tornogr 1991;15:796-801.
Thomas DGT, Gill SS, Wilson C. Current and future utilization of PET scanning in the evaluation and management of
malignant cerebral glioma. In: Appuzzo MLJ, editor. Malignant cerebral glioma. Park Ridge, 111: American Association
of Neurological Surgeons, 1991:79-89.
Mosskin M, van-Holst H, Bergstrom M, Collins VP,Eriksson
L, Johnstrom P. et al. Positron emission tomography with
11C-methionine and computed tomography of intracranial
tumours compared with histopathologic examination of
multiple biopsies. Acta Radiol 1987;28:673-81.
1106
CANCER September 1, 1996 / Volume 78 / Number 5
38. Maciunas RJ, Kessler RM, Maurer C, Mandava V, Watt G,
Smith G. Positron emission tomography imaging-directed
stereotactic neurosurgery. Stereotact Funct Neurosurg 1992;
58:134-40.
39. Daumas-Duport C, Scheithauer BW, O'Fallon I, Kelly PJ.
Grading of astrocytomas: a simple and reproductible
method. Cancer 1988;62:2152-65.
40. Black KL. Brain tumors (2 parts). N Engl J Med 1991;
324: 1471-6, 1555-64.
41. Hildebrand 1, Delecluse F. Malignant glioma. In: Slavin ML,
Staquet MI, editors. Randomized trials in cancer. New York:
Raven Press, 1986:583-604.
42. Mineura K, Sasajima T, Kowada H, Ogawa T, Hatazawa J,
Shishido F, et al. Perfusion and metabolism in predicting
the survival of patients with cerebral gliomas. Cancer
1994;7312386-94.
43. Yamamoto T, Seino Y, Fukomoto H, Koh G, Yano H, Inagchi
N. et al. Over-expression of facilitate glucose transporter
genes in human cancer. Biochem Biopkys Res Commun
1990; 170:223-30.
44. Kubota R, Yamada S, Kubota K, Ishiwata K, Tamahashi N,
Ido T. Intratumoral distribution of Fluorine- 18-fluorodeoxyglucose in vivo: high accumulation in macrophages and
45.
46.
47.
48.
49.
50.
granulation tissues studied by microautoradiography. I Nucl
Med 1992;33: 1972-80.
Sasaki M, Ichiya Y, Kuwabara Y, Otsuka M, Tahara T, Fukumura T, et al. Ring-like uptake of ["'FIFDG in brain abscess:
a PET study. J Comput Assist Tomogr 1990;14:486-7.
Hoffman JM, Waskin HA, Schifter T, Hanson MW, Gray L,
Rosenfeld S, et al. FDG-PET in differentiating lymphoma
from nonmalignant central nervous system lesions in patients with AIDS. J Nucl Med 1993;34:567-75.
Dethy S, Goldman S, Blecic S, Luxen A, Levivier M, Hildeu~o~e
brand J. ["CIMethionine and [ ' 8 F ] F l ~ ~ r ~ d e o x y g lPET
study in brain hematoma. J Nucl Med 1994;35:1162-6.
Dethy S, Manto M, Kentos A, Konopnicki D, Pirotte B, Goldman s, et al. PET findings in a brain abscess associated with
a silent atrial septa1 defect. Clin Neurol Neurosurg 1995;
97:349-53.
Hino A, Imahori Y,Tenjin H, Mizukawa N, Ueda S, Hirakawa
K, et al. Metabolic and hemodynamic aspects of peritumoral
low-density areas in human brain tumor. Neurosurgery
1990;26:615-21.
Burger PC, Vogel FS, Green SB, Strike TA. Glioblastoma multiforme and anaplastic astrocytoma: pathologic criteria and
prognostic implications. Cancer 1985;56:1106-11.
Документ
Категория
Без категории
Просмотров
5
Размер файла
846 Кб
Теги
870
1/--страниц
Пожаловаться на содержимое документа