Clinical Chemistry / OSTEOPONTIN IN ATYPICAL TERATOID/RHABDOID TUMOR
Elevation of Plasma and Cerebrospinal Fluid Osteopontin
Levels in Patients With Atypical Teratoid/Rhabdoid Tumor
Chung-Lan Kao, MD,1,9 Shih-Hwa Chiou, MD, PhD,2,10 Donald Ming-Tak Ho, MD,3
Yann-Jang Chen, MD, PhD,6 Ren-Shyan Liu, MD,4 Chih-Wen Lo,2 Fu-Ting Tsai,2
Chi-Hung Lin, MD, PhD,7 Hung-Hai Ku, PhD,8 Shang-Ming Yu, PhD,8 and Tai-Tong Wong, MD5
Key Words: Atypical teratoid/rhabdoid tumor; Primitive neuroectodermal tumor; Medulloblastoma; Osteopontin
DOI: 10.1309/0FTKBKVNK4T5P1L1
Abstract
Osteopontin, a cancer metastasis–associated gene,
is specifically up-regulated in central nervous system
(CNS) atypical teratoid/rhabdoid tumor (AT/RT), but its
biological behavior in the progression of CNS AT/RT
has never been studied. We obtained plasma,
cerebrospinal fluid (CSF), and brain tissue specimens
from lobectomy or hemispherectomy samples from 39
patients (medulloblastoma, 16; AT/RT, 8; epilepsy, 6;
hydrocephalus, 9). By enzyme-linked immunosorbent
assay, the median osteopontin levels in plasma and CSF
in AT/RT (852.0 and 1,175.0 ng/mL, respectively) were
significantly higher than in medulloblastoma (492.5
and 524.5 ng/mL, respectively) and hydrocephalus and
epilepsy (208.0 and 168.0 ng/mL, respectively) (P <
.05). The results of real-time reverse
transcriptase–polymerase chain reaction and
immunohistochemical analysis demonstrated that
osteopontin expression in AT/RT (n = 5) was
significantly higher than in medulloblastoma (n = 8)
samples. The differences in osteopontin expression in
plasma, CSF, and tumor samples in AT/RT and
medulloblastoma correlated with survival differences.
In 5 patients with AT/RT, plasma osteopontin levels
decreased after treatment but increased with relapse.
Osteopontin might be a potential marker to aid in
identifying AT/RT recurrence.
Primary central nervous system (CNS) atypical teratoid/
rhabdoid tumor (AT/RT) is an extremely malignant neoplasm
in infancy and childhood.1-4 Despite aggressive surgical and
adjuvant radiochemotherapy, the outcome of CNS AT/RT has
been uniformly poor.4-6 Because CNS AT/RT could be difficult to distinguish from primitive neuroectodermal tumor and
medulloblastoma, the differential diagnosis of these tumors is
very important. Histologically, AT/RTs consist of a unique
combination of cells, including rhabdoid cells, and peripheral
epithelial and mesenchymal elements.2-6 As the word teratoid
indicates, AT/RTs do not show divergent tissue development
characteristic of malignant teratomas. AT/RTs react to a wide
range of immunohistochemical markers, such as vimentin,
glial fibrillary acidic protein, epithelial membrane antigen,
cytokeratin, synaptophysin, and smooth muscle actin.1,2,5-10
With the use of complementary DNA microarray
assays,11,12 osteopontin was found to be up-regulated in
AT/RT primary cell culture compared with Daoy metastatic
medulloblastoma (HTB-186) and astrocyte (SVG-12) cell
lines obtained from the American Type Culture Collection
(Manassas, VA) in our previous study. Osteopontin is a bone
matrix glycoprotein that modulates mineralization and bone
resorption.13 It also might contribute to angiogenesis or neovascularization through the effects on expression of vascular
endothelial growth factor.13-15 Osteopontin was implicated as
a cancer- and metastasis-related gene by studies detecting elevated osteopontin expression in neoplastically transformed
cells and in several experimental models.16-18 Clinical studies
identified osteopontin as a potential diagnostic marker in ovarian, breast, colon, prostate, and lung cancers.19-21
The purpose of the present study was to further determine the role of osteopontin in AT/RT. We hypothesized that
Am J Clin Pathol 2005;123:297-304
© American Society for Clinical Pathology
297
DOI: 10.1309/0FTKBKVNK4T5P1L1
297
297
Kao et al / OSTEOPONTIN IN ATYPICAL TERATOID/RHABDOID TUMOR
the osteopontin concentration might be associated with
tumor progression or recurrence in AT/RT. We evaluated the
expression levels of osteopontin in the plasma, cerebrospinal
fluid (CSF), and tumor lesions from patients with AT/RT and
compared the results with those for medulloblastoma and
hydrocephalus and epilepsy cases. We measured osteopontin
levels in patients with AT/RT at different times, including
before and after treatment and with tumor relapse. We also
analyzed whether different levels of osteopontin expression
correlated with overall survival in patients with medulloblastoma and AT/RT.
Materials and Methods
This research followed the tenets of the Declaration of
Helsinki. Pathologic specimens and CSF samples from
patients with AT/RT, medulloblastoma, and hydrocephalus or
epilepsy were obtained from the time of surgery, between
January 1997 and April 2004. We obtained plasma, CSF, and
tumor tissue samples from 24 brain tumors, including 16
medulloblastomas and 8 CNS AT/RTs; these samples constituted the tumor group. The nontumor group included samples
from 6 patients who had undergone surgery for epilepsy and 9
who had undergone shunt surgery for hydrocephalus. Brain
tissue samples also were obtained from the same 6 patients
with epilepsy after lobectomy or hemispherectomy for epilepsy surgery. Tumor types and grades were classified by a senior neuropathologist according to the World Health
Organization histologic assessment criteria.
Samples from 39 patients were included in the study. The
mean ± SD age of patients at diagnosis was 7.1 ± 6.2 years
(range, 0.01 to 19.9 years). Tissue samples were obtained at
the time of surgery from 16 patients with medulloblastoma
(mean ± SD age, 8.5 ± 6.7 years; mean ± SD survival time,
48.8 ± 29.2 months; range, 14-99 months), 8 patients with
AT/RT (mean ± SD age, 6.6 ± 5.2 years; mean ± SD survival
time, 26.0 ± 11.4 months; range, 3-39 months), and 6 patients
with epilepsy. Plasma and CSF samples were obtained from
all patients with AT/RT, medulloblastoma, or epilepsy and
from the 9 patients with hydrocephalus.
Osteopontin Levels in Plasma and CSF by EnzymeLinked Immunosorbent Assay
Plasma and CSF samples were obtained on the day of surgery from all patients. Plasma samples were obtained again after
surgery and adjuvant therapy from 5 patients with AT/RT. CSF
samples were obtained from 3 recurrent AT/RT cases at the time
of the second surgery. All samples were kept on ice and frozen
at –70°C immediately. The concentration of soluble osteopontin in plasma and CSF samples was determined by using the
Human Osteopontin ELISA (enzyme-linked immunosorbent
298
298
Am J Clin Pathol 2005;123:297-304
DOI: 10.1309/0FTKBKVNK4T5P1L1
assay) EIA (enzyme immunoassay) kit (code No. 17158,
Immuno-Biological Laboratories, Gumma, Japan), which has
an interassay coefficient of variation varying from 0.7% to
2% and an intra-assay coefficient of variation varying from
3.7% to 4.7%, depending on the level of the marker. The
ELISA is designed to detect human osteopontin in serum
with a detection limit of 5 ng/mL or more. The developed
reaction was quantified by reading at 490 nm (MRX,
Dynatech Laboratories, Chantilly, VA). Each sample was
analyzed in triplicate.
Osteopontin Level in Tumor Specimens by Real-Time
Reverse Transcriptase–Polymerase Chain Reaction
Tissue RNA samples from 8 patients with medulloblastoma (mean ± SD age, 11.1 ± 8.5 years; mean ± SD survival
time, 45.5 ± 27.3 months; range, 16-98 months), 5 with AT/RT
(mean ± SD age, 4.7 ± 3.6 years; mean ± SD survival time,
28.2 ± 8.3 months; range, 20-39 months), and 6 with epilepsy
were obtained for real-time reverse transcriptase–polymerase
chain reaction (RT-PCR).
Total RNA from samples was isolated by using TRIzol
Reagent (Life Technologies, Gaithersburg, MD). The ratio
between the weight of tissue and the volume of TRIzol was
controlled by 100 mg of tissue per milliliter of reagent, as
described in the manufacturer’s protocol. The concentrations
of extracted RNA were calculated by spectrometer measurements (Ultrospec 3100 pro, Amersham Pharmacia Biotech,
Hong Kong), and by using the optical density 260/280 ratio
measurement and agarose gel examination. To exclude the
possibility of genomic DNA contamination, the GeneStrips
Hybridization Tube (RNAture, Irvine, CA) was used for messenger RNA purification: 2 mg of total RNA was used in each
case, and a 40-µL volume of messenger RNA was purified.
Relative quantitation by real-time PCR was performed
using the LightCycler (Roche Molecular Systems, Alameda,
CA). Amplification was carried out in a total volume of 20 µL
containing 0.5 µmol/L of each primer, 4 mmol/L of magnesium chloride, 2 µL of LightCycler-Fast Start DNA Master
SYBR grade I (Roche Diagnostics, Mannheim, Germany), and
2 µL of 10× diluted complementary DNA as the template.
GAPDH, the housekeeping gene, was amplified as a reference
standard in each experiment. The primer sequences used to
detect osteopontin and GADPH gene amplification were as follows: osteopontin: forward, 5'-TGAGAGCAATGAGCATTCCGATG-3'; reverse, 5'-CAGGGAGTTTCCATGAAGCCAC3'; GAPDH: forward, 5'-GAAGGTGAAGGTCGGAGTC-3';
reverse, 5'-CCCGAATCACATTCTCCAAGAA-3'.
PCR conditions were as follows: 1 cycle of 10 minutes
at 95°C followed by 40 cycles of denaturation at 95°C for 10
seconds, annealing at 55°C for 5 seconds, and extension at
72°C for 20 seconds. Bundled LightCycler quantification
software (version 3.3, Roche Molecular Systems) was used.
© American Society for Clinical Pathology
Clinical Chemistry / ORIGINAL ARTICLE
Standard curves (cycle threshold values vs template concentration) were prepared, and the osteopontin/GAPDH ratio
was calculated as an indicator of osteopontin transcript in
each case.
analysis was used whenever appropriate for comparison of subgroups. P values less than .05 were considered significant.
Osteopontin Expressions in Tumor Specimens by
Immunohistochemical Analysis
The 4-µm paraffin sections were deparaffinized in xylene,
rehydrated in a series of graded alcohols, and immunostained
with antibodies against vimentin (ChemMate, DAKO,
Glostrup, Denmark), smooth muscle actin (HHF 35,
ChemMate, DAKO), glial fibrillary acidic protein
(ChemMate, DAKO), epithelial membrane antigen (DAKO
ChemMate Detection Kit), cytokeratin (AE1/AE3;
ChemMate, DAKO), neuron-specific enolase (ChemMate,
DAKO), synaptophysin (ChemMate, DAKO), chromogranin
(ChemMate, DAKO), placental alkaline phosphatase (PLAP)
(ChemMate, DAKO), and MIB-1 (ChemMate, Immunotech,
Marseille, France), as described previously,6 and osteopontin
(dilution 1:100; 10A16; Immuno-Biological Laboratories).
Immunoreactive signals were detected with a mixture of
biotinylated IgG antibody and peroxidase-conjugated streptavidin (LSAB2 system, DAKO). Brain tissue samples from 6
patients with epilepsy who had undergone lobectomy or hemispherectomy were used as control samples. The immunostaining results were graded based on the percentage of reactive
cells and scaled as follows: –, absent; 1+, 1% to 10%; 2+, 10%
to 20%; 3+, 20% to 40%; and 4+, more than 40%.
Results
Osteopontin Levels in Plasma and CSF samples by
ELISA
The median and interquartile range of plasma osteopontin levels in patients with AT/RT were 852.0 ng/mL (range,
715.5-1,024.0 ng/mL); in patients with medulloblastoma,
492.5 ng/mL (range, 392.5-582.0 ng/mL); and in patients
without tumors, 208.0 ng/mL (range, 157.0-275.0 ng/mL)
❚Figure 1A❚. The median and interquartile range of CSF osteopontin levels in patients with AT/RT were 1,175.0 ng/mL
(range, 859.5-1,599.0 ng/mL); with medulloblastoma, 524.5
ng/mL (range, 456.0-607.5 ng/mL); and patients without
tumor, 168.0 ng/mL (range, 145.3-211.8 ng/mL ) ❚Figure 1B❚.
By using Kruskal-Wallis 1-way analysis, we found that plasma and CSF osteopontin levels in patients with AT/RT were
significantly higher than those in patients with medulloblastoma or no tumor (P < .05). Regression analysis revealed statistically significant correlations between the plasma and CSF
osteopontin levels in patients with AT/RT and with medulloblastoma. In contrast, the plasma osteopontin levels correlated fairly with the CSF osteopontin levels in patients without
tumor ❚Figure 2❚.
To further determine whether osteopontin levels were
correlated with clinical treatment and disease progression,
we measured the plasma osteopontin levels in 5 AT/RT
cases. Our findings indicated that after radical surgery and
adjuvant chemoradiotherapy, the osteopontin plasma levels
in all cases dropped significantly (P < .05). Rebound of the
Statistical Methods
Overall survival was calculated by using the Kaplan-Meier
method from the date of diagnosis until last contact or death due
to any cause. Differences between Kaplan-Meier curves were
assessed by using the log-rank test. The Kruskal-Wallis 1-way
A
B
2,000
1,200
CSF Osteopontin (ng/mL)
Plasma Osteopontin (ng/mL)
1,400
P < .05
P < .05
P < .05
1,000
800
600
400
200
0
Hydrocephalus
(n = 15)
Medulloblastoma
(n = 16)
AT/RT
(n = 8)
P < .05
P < .05
P < .05
1,600
1,200
800
400
0
Hydrocephalus
(n = 15)
Medulloblastoma
(n = 16)
AT/RT
(n = 8)
❚Figure 1❚ Levels of the soluble
form of osteopontin in plasma
(A) and in cerebrospinal fluid (CSF)
(B) of patients with atypical
teratoid/rhabdoid tumor (AT/RT;
n = 8), medulloblastoma (n = 16),
and no tumor (hydrocephalus
[n = 9] or epilepsy [n = 6]) (mean ±
SD, indicated by square and error
bars). Kruskal-Wallis 1-way analysis
showing the average osteopontin
concentration in patients with
AT/RT was significantly higher than
that in patients with medulloblastoma
(P < .05) or hydrocephalus or
epilepsy (P < .05).
Am J Clin Pathol 2005;123:297-304
© American Society for Clinical Pathology
299
DOI: 10.1309/0FTKBKVNK4T5P1L1
299
299
Kao et al / OSTEOPONTIN IN ATYPICAL TERATOID/RHABDOID TUMOR
A
1,000
300
250
200
150
100
50
0
0 50 100 150 200 250 300 350
Plasma Osteopontin (ng/mL)
2,000
CSF Osteopontin (ng/mL)
CSF Osteopontin (ng/mL)
350
CSF Osteopontin (ng/mL)
C
B
800
1,500
600
1,000
400
200
0
0 200
400
600
800
1,000
Plasma Osteopontin (ng/mL)
500
0
0 600
800
1,000
1,200
Plasma Osteopontin (ng/mL)
❚Figure 2❚ Regression analysis showing statistically significant correlation between plasma and cerebrospinal fluid (CSF)
osteopontin levels in patients with atypical teratoid/rhabdoid tumor (AT/RT) (n = 8; r = 0.91; P = .001) (A), medulloblastoma (n =
16; r = 0.95; P < .001) (B), or no tumor (hydrocephalus [n = 9] or epilepsy [n = 6]) (n = 15; r = 0.73; P = .002) (C).
plasma osteopontin levels was noted with tumor recurrence
just before surgery for relapse ❚Figure 3A❚. The CSF osteopontin levels were abnormally high at initial diagnoses of
AT/RT and at tumor recurrence in 3 cases ❚Figure 3B❚.
Osteopontin Expression in Tissues Specimens by RealTime RT-PCR
Real-time, quantitative RT-PCR analysis revealed amplification of the osteopontin transcript in tissue samples of patients
with AT/RT or medulloblastoma and in patients with epilepsy
❚Figure 4❚. The average osteopontin/GAPDH ratio in the tissue
samples of patients with AT/RT (mean ratio, 29.3; n = 5) was
significantly higher than that in patients with medulloblastoma (mean ratio, 7.1; n = 8) or epilepsy (mean ratio, 0.95; n
= 6) (P < .05). There was no significant difference in mean age
and survival time of selected patients with tumor vs all
patients with tumor (AT/RT and medulloblastoma; P > .1). In
A
Histologic Findings and Immunohistochemical Studies of
Tumor Specimens
The histologic features of AT/RT include the presence of
“rhabdoid” cells ❚Image 1A❚. The MIB-I index ranged from 40
B
1,800
CSF Osteopontin (ng/mL)
1,200
Plasma Osteopontin (ng/mL)
all specimens of brain tissues from 6 epilepsy resections, the
osteopontin/GAPDH ratios were less than 5. In 3 (38%) of 8
medulloblastoma tissue samples, osteopontin/GAPDH ratios
were less than 5; in 3 of (38%) 8 they were between 5 and 10;
and 2 (25%) of 8 were more than 10. In contrast, the osteopontin/GAPDH ratios in all 5 AT/RT tissue samples were more
than 19. Furthermore, we found that the osteopontin ratios of
AT/RT vs medulloblastoma and epilepsy in tissue samples
were more than the ratios in CSF and plasma (osteopontin
ratios of AT/RT vs medulloblastoma and normal tissues, 4.12
and 16.5, respectively; in CSF, 2.33 and 6.96, respectively;
and in plasma, 1.72 and 4.37, respectively).
900
600
300
300
300
0
1st
Surgery
After
Treatment
(2-4 mo)
Am J Clin Pathol 2005;123:297-304
DOI: 10.1309/0FTKBKVNK4T5P1L1
2nd
Surgery
(Recurrent)
1,500
1,200
900
600
300
0
1st
Surgery
2nd
Surgery
❚Figure 3❚ Plasma osteopontin levels
in 5 patients with atypical
teratoid/rhabdoid tumor (AT/RT)
before and after 2-4 months of
treatment and with tumor recurrence
after 1-9 months. A, A significant
decrease in osteopontin levels
occurred following treatment (P <
.05). Plasma osteopontin levels
increased with tumor relapse.
B, Cerebrospinal fluid (CSF)
osteopontin levels increased in 3
patients with AT/RT with tumor
recurrence.
© American Society for Clinical Pathology
Clinical Chemistry / ORIGINAL ARTICLE
Survival Analysis
Of the 24 patients with brain tumors included in the survival analysis, the mean ± SD follow-up period was 41.8 ±
26.4 months (range, 3-98 months). Kaplan-Meier survival
analysis revealed a more favorable overall survival (P =
.003) for patients with medulloblastoma compared with
patients with AT/RT ❚Figure 5A❚. Patients were grouped as
having high or low osteopontin levels based on the 10th percentile of osteopontin levels in AT/RT (osteopontin, >667 vs
≤667 ng/mL in plasma and >787 vs ≤787 ng/mL in CSF).
Patients with high osteopontin levels (>667 ng/ml) had significantly poorer survival than those with low osteopontin
levels (≤667 ng/mL) ❚Figure 5B❚. The log-rank test showed a
difference in survival between these 2 groups of patients (P
< .0001). Patients with a CSF osteopontin level of more than
787 ng/mL had less favorable overall survival than patients
with a CSF osteopontin level of 787 ng/mL or less (P =
.0003) ❚Figure 5C❚.
Discussion
AT/RTs usually exhibit more unfavorable outcomes than
medulloblastomas owing to their high frequency of leptomeningeal dissemination.1-6,8 The present study offers the
first evidence that patients with AT/RT have higher plasma
and CSF osteopontin levels in comparison with patients with
medulloblastoma, hydrocephalus, or epilepsy. Overexpression
of osteopontin was confirmed in pathologic sections of
AT/RT by real-time RT-PCR and immunohistochemical
analysis. We further found a significant correlation between
plasma osteopontin levels and the risk of tumor relapse in 5
patients with AT/RT.
35
Osteopontin/GAPDH Ratio
to 62.8. Of the 8 patients with AT/RT, immunohistochemical
results revealed immunoreactivity for vimentin in 7 (88%), for
smooth muscle actin in 8 (100%), for synaptophysin in 7
(88%), for epithelial membrane antigen in 3 (38%), for glial
fibrillary acidic protein in 4 (50%), for neuron-specific enolase in 2 (25%), and for cytokeratin in 3 (38%). None of the
cases were immunoreactive for PLAP and chromogranin.
Electron microscopic examination of 1 specimen revealed that
the majority of tumor cells were large, with relatively abundant cytoplasm ❚Image 1B❚ and ❚Image 1C❚. Ultrastructural
features included the presence of large, whorled masses of
intermediate filaments within the cytoplasm. Dilated, rough
endoplasmic reticulum was observed in most tumor cells.
Furthermore, in accordance with our real-time RT-PCR
results, immunohistochemical studies showed expression of
osteopontin in all 8 cases of AT/RT tumors (1+ in 1 patient
[13%], 2+ in 2 patients [25%], 3+ in 4 patients [50%], 4+ in 1
patient [13%]) ❚Image 1D❚.
30
25
20
15
10
5
0
1 2 3 4 5 6
Epilepsy
1 2 3 4 5 6 7 8
Medulloblastoma
1 2 3 4 5
AT/RT
❚Figure 4❚ Osteopontin expression in tissue specimens as
shown by real-time reverse transcriptase–polymerase chain
reaction. The average osteopontin/glyceraldehyde-3phosphate dehydrogenase (GAPDH) ratio in patients with
atypical teratoid/rhabdoid tumor (AT/RT; n = 5) was
significantly higher than the ratios in medulloblastoma (n = 8)
and epilepsy (n = 6) (P < .05)
Osteopontin, a bone matrix glycoprotein, originally was
thought to function as a modulator for bone resorption and
remodeling.13,14 The mechanisms of osteopontin in the progression of malignancy possibly might involve the binding of
integrin via integrin-mediated signaling. The integrin-mediated signaling pathway was believed to be responsible for adhesion, migration, and angiogenesis of cancers.14,15,22 Inducing
the activities of metalloproteinase family members by osteopontin also might enhance dissemination from the primary
tumor because the metalloproteinase family is a major component of extracellular matrix degradation.23,24
Previous studies have shown that osteopontin can support
in vitro attachment for a variety of cell types and promote
migration of inflammatory and tumor cells.13,25,26 A recent
study also demonstrated that osteopontin promoted the attachment of malignant astrocytoma cells and caused these cells to
become more migratory and invasive.27 Osteopontin gene
knockout studies further indicated that osteopontin inhibits
macrophage functions and enhances growth and metastasis,
suggesting that osteopontin is involved in tumor protection
from immune surveillance and enhancement of tumor cell survival.28,29 Thus, the high expression of osteopontin in plasma,
CSF, and tumor tissue samples in CNS AT/RT in our study
might be associated with increased invasiveness or metastatic
potential, leading to the aggressive clinical characteristics and
poor outcome in AT/RT.
A recent study pointed out that the expression level of osteopontin differed significantly between individual astrocytoma
tissue samples and seemed to correlate with their malignancy
Am J Clin Pathol 2005;123:297-304
© American Society for Clinical Pathology
301
DOI: 10.1309/0FTKBKVNK4T5P1L1
301
301
Kao et al / OSTEOPONTIN IN ATYPICAL TERATOID/RHABDOID TUMOR
A
B
C
D
❚Image 1❚ Pathologic findings for an atypical teratoid/rhabdoid tumor (AT/RT) specimen. A, Medium-sized to large cells with a
moderate amount of cytoplasm and large oval, polygonal, and elongated nuclei with prominent nucleoli (H&E, bar, 100 µm).
B and C, Electron microscopy showed a tumor cell with a prominent nucleus and bundles of intermediate filaments (bar, 2 µm).
D, Immunohistochemical study showed expression of osteopontin protein in AT/RT tissue sections. Osteopontin signals were
identified by the chromogen 3-amino-9-ethyl-carbazole as reddish brown (bar, 100 µm).
grades and invasive potential.30 Other studies found that osteopontin might be a stage-related prognostic factor and a potential tumor marker in intrahepatic metastasis and head and neck
squamous cell carcinomas. The presence of plasma osteopontin levels in tumor was shown to correlate with tumor relapses and death.31,32 In the present study, 5 patients with AT/RT
had a significant decrease in plasma and CSF osteopontin levels after extensive tumor resection and adjuvant treatment,
but the level increased with tumor recurrence. As demonstrated in previous studies, elevation of the osteopontin level
was associated with increased tumor burden in metastatic
breast and prostate carcinoma.33,34 Our study shared the same
302
302
Am J Clin Pathol 2005;123:297-304
DOI: 10.1309/0FTKBKVNK4T5P1L1
observation, ie, that the osteopontin level might be a marker
for evaluating tumor status in patients with AT/RT.
Another interesting finding of our study is that the osteopontin ratio of AT/RT vs medulloblastoma and epilepsy was
more profound in tissue samples than in CSF and plasma
samples. Although the exact localization of osteopontin in
tumor cells is still debated, present knowledge supports the
synthesis of osteopontin in tumor cells.35 Because osteopontin produced by tumors is primarily soluble,35 we assume that
the plasma and CSF osteopontin levels in patients with
AT/RT might partly reflect the production of osteopontin in
brain tumor lesions, which possibly also represents the
© American Society for Clinical Pathology
Clinical Chemistry / ORIGINAL ARTICLE
A
1.0
0.8
0.6
0.4
AT/RT
0.8
0.6
0.4
0 12 24 36 48 60 72 84 96
Months
⬎667 ng/mL
0.2
0.0
0.0
1.0
ⱕ 667 ng/mL
0 12 24 36 48 60 72 84 96
Months
Cumulative Survival
Medulloblastoma
Cumulative Survival
Cumulative Survival
1.0
0.2
C
B
ⱕ 787 ng/mL
0.8
0.6
0.4
⬎787 ng/mL
0.2
0.0
0 12 24 36 48 60 72 84 96
Months
❚Figure 5❚ A, Patients with medulloblastoma (n = 16) had better overall survival than patients with atypical teratoid/rhabdoid
tumor (AT/RT; n = 8) (P = .0033). B, Patients with plasma osteopontin levels of 667 ng/mL or less (n = 14) had more favorable
overall survival than those with levels greater than 667 ng/mL (n = 10) (P < .0001). C, Patients with osteopontin levels in the
cerebrospinal fluid of 787 ng/mL or less (n = 16) had more favorable overall survival than those with levels greater than 787
ng/mL (n = 8) (P = .0003).
changes in tumor metabolism and turnover. The strong positive correlations between tumor plasma and CSF osteopontin
levels have pointed to the possibility that monitoring the plasma osteopontin level, along with the clinical manifestations
and image studies, might provide an additional marker for
detection of recurrence of AT/RT.
An important question still to be addressed is the association of osteopontin level with survival. From the reported
literature, there is an evident association between osteopontin and decreased survival times in breast, prostate, gastric,
colon, and ovarian metastatic neoplasms, head and neck
squamous carcinoma, and hepatocellular carcinoma.1820,27,30-34,36 Our survival analysis revealed a statistically significant correlation between elevated osteopontin concentrations and decreased overall survival between AT/RT and
medulloblastoma in plasma (P < .0001) and CSF (P =
.0003). The cutoff points for osteopontin in survival measurements in the present study were set at the 10th percentile
of AT/RT osteopontin levels (667 ng/mL for plasma and 787
ng/mL for CSF). It seemed that patients with higher osteopontin levels had less favorable survival. Because AT/RT has
been documented to have poorer survival than medulloblastoma, our findings suggest that osteopontin might be a candidate for prediction of prognosis. A larger sample for intragroup and intergroup comparison is needed for future clinical applications.
We demonstrated higher osteopontin levels in plasma
and CSF in CNS AT/RT than in medulloblastoma, hydrocephalus, and epilepsy. We also found that elevation of
osteopontin levels was associated with unfavorable overall
survival in CNS AT/RT compared with medulloblastoma.
The levels reflected tumor status in AT/RTs, suggesting that
osteopontin might be a potential prognostic marker for
monitoring tumor recurrence and survival in CNS AT/RTs.
Additional follow-up and a larger sample are necessary to
confirm our findings.
From the Departments of 1Physical Medicine and Rehabilitation,
2Education and Medical Research, 3Pathology, and 4Nuclear
Medicine and 5Division of Pediatric Neurosurgery, the Neurological
Institute, Taipei Veterans General Hospital and National Yang-Ming
University; 6Faculty of Life Science and Institutes of 7Microbiology
and Immunology, 8Anatomy and Cell Biology, and 9Clinical
Medicine, National Yang-Ming University; and 10National Research
Institute of Chinese Medicine, Taipei, Taiwan.
Supported by grants 92 and 93 from the Taipei Veterans
General Hospital, the joint projects of VTY (92-P1-07/08) and
UTVGH (93-P1-04/06/10), Yen Tjing Ling Medical Foundation,
and the National Science Council.
Address reprint requests to Dr Wong: Division of Pediatric
Neurosurgery, the Neurological Institute, Taipei Veterans General
Hospital and National Yang-Ming University, Taipei, Taiwan.
Acknowledgment: We thank the National Microarray and
Gene Expression Analysis Core Facility of National Yang-Ming
University for excellent technical support.
References
1. Rorke LB, Packer RJ, Biegel JA. Central nervous system
atypical teratoid/rhabdoid tumors of infancy and childhood:
definition of an entity. J Neurosurg. 1996;85:56-65.
2. Burger PC, Yu IT, Tihan T, et al. Atypical teratoid/rhabdoid
tumor of the central nervous system: a highly malignant tumor
of infancy and childhood frequently mistaken for
medulloblastoma. Am J Surg Pathol. 1998;22:1083-1092.
3. Rorke LB, Packer R, Biegel J. Central nervous system atypical
teratoid/rhabdoid tumors of infancy and childhood. J
Neurooncol. 1995;24:21-28.
4. Buchino JJ. Atypical teratoid/rhabdoid tumor: a stereogram
unveiled. Adv Anat Pathol. 1999;6:97-102.
Am J Clin Pathol 2005;123:297-304
© American Society for Clinical Pathology
303
DOI: 10.1309/0FTKBKVNK4T5P1L1
303
303
Kao et al / OSTEOPONTIN IN ATYPICAL TERATOID/RHABDOID TUMOR
5. Utsuki S, Oka H, Tanaka S, et al. Importance of reexamination for medulloblastoma and atypical
teratoid/rhabdoid tumor. Acta Neurochir (Wien).
2003;154:663-666.
6. Ho DMT, Hsu CY, Wong TT, et al. Atypical teratoid/rhabdoid
tumor of the central nervous system: a comparative study with
primitive neuroectodermal tumor/medulloblastoma. Acta
Neuropathol (Berl). 2000;99:482-488.
7. Bergmann M, Spaar HJ, Ebhard G, et al. Primary malignant
rhabdoid tumors of the central nervous system: an
immunohistochemical and ultrastructural study. Acta
Neurochir (Wien). 1997;139:961-969.
8. Oka H, Scheithauer BW. Clinicopathological characteristics
of atypical teratoid/rhabdoid tumor. Neurol Med Chir (Tokyo).
1999;39:510-518.
9. Bhattacharjee M, Hicks J. Central nervous system atypical
teratoid/rhabdoid tumor of infancy and childhood. Ultrastruct
Pathol. 1997;21:369-378.
10. Behring B, Bruck W, Goebel HH, et al.
Immunohistochemistry of primary central nervous system
malignant rhabdoid tumors: report of five cases and review of
the literature. Acta Neuropathol (Berl). 1996;91:578-586.
11. Pomeroy SL, Tamayo P, Gassenbeek M, et al. Prediction of
central nervous system embryonal tumour outcome based on
gene expression. Nature. 2002;415:436-442.
12. Chan JYH, Chen LK, Chang JF, et al. Differential gene
expression in a DNA double-strand-break repair mutant XRS5 defective in Ku80: analysis by cDNA microarray. J Radiat Res
(Tokyo). 2001;42:371-385.
13. Furger KA, Menon RK, Tuckl AB, et al. The functional and
clinical roles of osteopontin in cancer and metastasis. Curr
Mol Med. 2001;1:621-632.
14. Hirama M, Takahashi F, Takahashi K, et al. Osteopontin
overproduced by tumor cells acts as a potent angiogenic factor
contributing to tumor growth. Cancer Lett. 2003;198:107-117.
15. Shijubo N, Uede T, Kon S, et al. Vascular endothelial growth
factor and osteopontin in tumor biology. Crit Rev Oncog.
2000;11:135-146.
16. Senger DR, Asch BB, Smith BD, et al. A secreted
phosphoprotein marker for neoplastic transformation of both
epithelial and fibroblastic cells. Nature. 1983;302:714-715.
17. Craig AM, Bowden GT, Chambers AF, et al. Secreted
phosphoprotein mRNA is induced during multi-stage
carcinogenesis in mouse skin and correlates with the
metastatic potential of murine fibroblasts. Int J Cancer.
1990;46:133-137.
18. Oates AJ, Barraclough R, Rudland PS. The identification of
osteopontin as a metastasis-related gene product in a rodent
mammary tumor model. Oncogene. 1996;13:97-104.
19. Kim JH, Skates SJ, Uede T, et al. Osteopontin as a potential
diagnostic biomarker for ovarian cancer. JAMA.
2002;287:1671-1678.
20. Agrawal D, Chen T, Irby R, et al. Osteopontin identified as
lead marker of colon cancer progression, using pooled sample
expression profiling. J Natl Cancer Inst. 2002;94:513-521.
304
304
Am J Clin Pathol 2005;123:297-304
DOI: 10.1309/0FTKBKVNK4T5P1L1
21. Fedarko NS, Jain A, Karadag A, et al. Elevated serum bone
sialoprotein and osteopontin in colon, breast, prostate, and
lung cancer. Clin Cancer Res. 2001;7:4060-4066.
22. Takahashi F, Akutagawa S, Fukumoto H, et al. Osteopontin
induces angiogenesis of murine neuroblastoma cells in mice.
Int J Cancer. 2002;98:707-712.
23. Agnihotri R, Crawford HC, Haro H, et al. Osteopontin, a
novel substrate for matrix metalloproteinase-3 and matrix
metalloproteinase-7. J Biol Chem. 2001;276:28261-28267.
24. Philip S, Kundu GC. Osteopontin induces nuclear factor
kappa B–mediated promatrix metalloproteinase-2 activation
through I kappa B alpha /IKK signaling pathways, and
curcumin down-regulates these pathways. J Biol Chem.
2003;278:14487-14497.
25. Chabas D, Baranzini SE, Mitchell D, et al. The influence of
the proinflammatory cytokine, osteopontin, on autoimmune
demyelinating disease. Science. 2001;294:1731-1735.
26. Gao C, Guo H, Wei J, et al. Osteopontin inhibits expression
of cytochrome c oxidase in RAW 264.7 murine macrophages.
Biochem Biophys Res Commun. 2003;309:120-125.
27. Ding Q, Stewart J Jr, Prince CW, et al. Promotion of
malignant astrocytoma cell migration by osteopontin
expressed in the normal brain: differences in integrin signaling
during cell adhesion to osteopontin versus vitronectin. Cancer
Res. 2002;62:5336-5343.
28. Nemoto H, Rittling SR, Yoshitake H, et al. Osteopontin
deficiency reduces experimental tumor cell metastasis to bone
and soft tissues. J Bone Miner Res. 2001;16:652-659.
29. Crawford HC, Matrisian LM, Liaw L. Distinct roles of
osteopontin in host defense activity and tumor survival during
squamous cell carcinoma progression in vivo. Cancer Res.
1998;58:5206-5215.
30. Rudlland PS, Platt-Higgins A, El-Tanani M, et al. Prognostic
significance of the metastasis-associated protein osteopontin in
human breast cancer. Cancer Res. 2002;62:3417-3427.
31. Pan HW, Ou YH, Peng SY, et al. Overexpression of
osteopontin is associated with intrahepatic metastasis, early
recurrence, and poorer prognosis of surgically resected
hepatocellular carcinoma. Cancer. 2003;98:119-127.
32. Le QT, Sutphin PD, Raychaudhuri S, et al. Identification of
osteopontin as a prognostic plasma marker for head and neck
squamous cell carcinomas. Clin Cancer Res. 2003;9:59-67.
33. Saitoh Y, Kuratsu JI, Takeshima H, et al. Expression of
osteopontin in human glioma. Lab Invest. 1995;72:55-63.
34. Horre SJ, Wingquist EW, Stitt L, et al. Plasma osteopontin
association with survival and metastasis to bone in men with
hormone-refractory prostate carcinoma. Cancer. 2002;95:506512.
35. Rittling SR, Chen Y, Feng F, et al. Tumor-derived osteopontin
is soluble, not matrix associated. J Biol Chem. 2002;277:91759182.
36. Coppola D, Szabo M, Boulware D, et al. Correlation of
osteopontin protein expression and pathological stage across a
wide variety of tumor histologies. Clin Cancer Res.
2004;10:184-190.
© American Society for Clinical Pathology