doi: 10.1136/jnnp.2007.121582
August 31, 2007
2008 79: 574-580 originally published onlineJ Neurol Neurosurg Psychiatry
S-Y Yang, C-K Park, S-H Park, et al.
features
prognostic implications of clinicopathological
Atypical and anaplastic meningiomas:
http://jnnp.bmj.com/content/79/5/574.full.html
Updated information and services can be found at:
These include:
References
http://jnnp.bmj.com/content/79/5/574.full.html#ref-list-1
This article cites 34 articles, 4 of which can be accessed free at:
service
Email alerting
box at the top right corner of the online article.
Receive free email alerts when new articles cite this article. Sign up in the
Topic collections
(6064 articles)Surgical oncology
(14701 articles)Radiology
(654 articles)CNS cancer
(1036 articles)Neurooncology
(43508 articles)Immunology (including allergy)
Articles on similar topics can be found in the following collections
Notes
http://jnnp.bmj.com/cgi/reprintform
To order reprints of this article go to:
http://jnnp.bmj.com/subscriptions
go to: Journal of Neurology, Neurosurgery & PsychiatryTo subscribe to
group.bmj.com on December 23, 2009 - Published by jnnp.bmj.comDownloaded from
Atypical and anaplastic meningiomas: prognostic
implications of clinicopathological features
S-Y Yang,
1,2
C-K Park,
2,3
S-H Park,
4
D G Kim,
2,3
Y S Chung,
2,3
H-W Jung
2,3
1
Department of Neurosurgery,
DongGuk University International
Hospital, Goyang, Korea;
2
Clinical Research Institute,
Seoul National University
Hospital, Seoul, Korea;
3
Department of Neurosurgery,
Seoul National University
College of Medicine, Seoul,
Korea;
4
Department of
Pathology, Seoul National
University College of Medicine,
Seoul, Korea
Correspondence to:
Young Seob Chung, MD, PhD,
Department of Neurosurgery,
Seoul National University
College of Medicine, 28
Yeongeon-dong, Jongno-gu,
Seoul 110–744, Republic of
Korea; yschung@snu.ac.kr
Received 27 March 2007
Revised 16 August 2007
Accepted 18 August 2007
Published Online First
31 August 2007
ABSTRACT
Objectives: To evaluate patient outcome and investigate
the prognostic factors of high-grade meningiomas by
adopting the 2000 World Health Organization (WHO)
classification system.
Methods: Between 1986 and 2004, 74 patients were
diagnosed with high-grade meningioma: 33 with atypical
and 41 with anaplastic meningioma. The mean follow-up
was 58.5 months. We reclassified all surgical specimens,
according to the 2000 WHO classification system, using
two expert neuropathologists.
Results: Forty of 74 meningiomas were reclassified as
atypical meningioma and 24 as anaplastic meningioma.
Overall and recurrence-free survivals were significantly
longer in patients with atypical than in those with
anaplastic meningioma: 142.5 versus 39.8 months and
138.5 versus 32.2 months, respectively (p,0.001). In
patients with atypical meningiomas, brain invasion and
adjuvant radiotherapy were not associated with survival;
however, in the brain invasion subgroup, adjuvant
radiotherapy improved patients’ survival. In patients with
anaplastic meningioma, the prognostic factors were brain
invasion, adjuvant radiotherapy, malignant progression,
p53 overexpression and extent of resection. The p53
overexpression was the only factor associated with
malignant progression (p = 0.009).
Conclusions: The 2000 WHO classification has identified
the truly aggressive meningiomas better than did the
previous criteria. A precise meningioma grading system
may help to avoid over-treatment of patients with an
atypical meningioma as, once the tumour has ‘‘declared
itself’’ by recurrence and histological features, it becomes
a tumour that is poorly amenable to current therapies.
Since malignant meningioma was first recognised
by Cushing and Eisenhardt in 1938,
1
there have
been diverse criteria for histopathologically grading
atypical and anaplastic meningiomas. To improve
this situation, the 2000 WHO classification recom-
mends much more stringent and objective criteria.
2
In the 2000 WHO classification, some important
diagnostic variables were amended, particularly
proliferation index, brain invasion and mitotic
activity.
Most tumours tend to become more aggressive
in clinical behaviour and more ‘‘malignant’’ in their
characteristics over time, although this time course
may be variable. Recent studies into the cytoge-
netic alteration of meningioma have provided tools
for understanding the mechanisms underlying
malignant progression.
34
These advances may be
useful in improving our ability to predict clinical
outcome and develop therapeutic strategies to
improve outcomes in patients with high-grade
meningiomas.
This study was primarily motivated by growing
concerns about the validity of treatment based on
both histological grading and cytogenetic altera-
tions of meningiomas. To address this issue, by
adopting the 2000 WHO criteria, we reclassified
previous atypical and anaplastic meningiomas,
reanalysed their treatment outcomes and re-eval-
uated prognostic factors by clinicopathological
aspects.
MATERIALS AND METHODS
Patients
Between 1986 and 2004, 1098 patients with an
intracranial meningioma were treated surgically at
Seoul National University Hospital (SNUH). Of
these 1098 patients, 79 were previously diagnosed
as having atypical or anaplastic meningiomas. The
rhabdoid or papillary variants were initially
excluded. Five of these 79 patients were also
excluded from our study: 1 patient had neurofi-
bromatosis type 2; 1 was younger than 15 years;
and biopsy tissue from 3 patients was not available
for review. Of the 74 included patients, 33 were
originally diagnosed with atypical meningioma and
41 with anaplastic meningioma. Using the 2000
WHO classification system, we reviewed biopsy
tissues of these 74 patients.
The clinical records of all patients were reviewed
to obtain data on overall survival (time from
diagnosis of atypical or anaplastic meningioma to
death), recurrence-free survival (time from diag-
nosis of atypical or anaplastic meningioma to first
evidence of recurrence or disease progression either
clinically or radiologically), and time for malignant
progression (time interval from low-grade menin-
gioma to high-grade meningioma). We also col-
lected the information on both cause of death and
date of death through the database of the Korea
National Statistical Office and the Korean National
Heath Insurance Corporation. The extent of
resection was deduced from the operation records
or postoperative scans when available, and assessed
according to the Simpson grading system.
5
The
institutional review board of SNUH approved this
study. However, it did not require informed
consent from these patients for their inclusion in
this study because the study depended only on
information obtained as a part of routine clinical
care and patient medical records.
Pathological review
We reviewed all surgical specimens of the 74
patients, including their previous samples that
had been diagnosed as benign meningioma either
from SNUH or other hospitals. Two neuropathol-
ogists reviewed all available slides according to the
Research paper
574 J Neurol Neurosurg Psychiatry 2008;79:574–580. doi:10.1136/jnnp.2007.121582
group.bmj.com on December 23, 2009 - Published by jnnp.bmj.comDownloaded from
2000 WHO classification
2
without prior knowledge of the
original grading or patients’ outcome. Where differences
between the neuropathologists’ opinions were encountered,
the sections were re-examined by both neuropathologists and
results were recorded after an agreement was reached. Brain
invasion was recorded as present, absent or unassessable (no
adjacent brain parenchyma).
For immunohistochemical studies, the avidin-biotin perox-
idase indirect technique with diaminobenzadine as the chroma-
gen was applied using antibody-directed p53 protein
(DakoCytomation, Carpinteria, CA, USA; monoclonal, dilution
1:1200) and the proliferation marker MIB-1 (DakoCytomation;
monoclonal, dilution 1: 80). Immunostained sections were
interpreted as positive if they showed either a focal or diffuse
cytoplasmic reaction product, with nuclear expression in the
case of p53 and MIB-1. To quantify the percentage of nuclei
staining with MIB-1, a nuclear antigen, at least 500 tumour cells
were manually counted three times in the region of tumour
with the highest nuclear staining, and an average percentage of
positively stained nuclei calculated.
Statistical methods
Continuous variables and categorical variables with more than
three categories were dichotomised on the basis of their median
values or the variables’ clinical significance—that is, age (,65
years, >65 years), MIB-1 labelling index ((5, .5), extent of
tumour removal (Simpson grade I, II and >III) and tumour
location (convexity, non-convexity). All the data are expressed
with standard errors. Survival was estimated by the Kaplan–
Meier method, and the log-rank test was used for comparisons.
A Firth’s Cox proportional hazards model was used to analyse
possible prognostic factors for the risk of recurrence or death. To
assess the effect of adjuvant radiotherapy, the patients were
divided into two groups: radiotherapy group and non-radio-
therapy group. A two-sided probability level of 0.05 was chosen
for statistical significance.
Results
On application of the 2000 WHO classification system, 10 of 74
meningiomas were reclassified as benign, 40 as atypical and 24
as anaplastic meningioma (table 1). Reassessment of the
meningiomas resected before 2001 (meningiomas originally
graded by using the 1993 WHO classification system) revealed
that 43.8% (25/57) of high-grade meningiomas were reclassified.
Reassessment of the meningiomas resected after January 2001
(meningiomas graded by using the 2000 WHO classification
system after January 2001) revealed that 5.8% (1/17) of
meningiomas originally classified as grade II were reclassified
as grade I.
Brain invasion was evaluated only in 81% (60/74) of cases
where it was assessable. Dissection of the tumour was
performed between two membranes of the tumour capsule.
Complete resection of tumour could be carried out with ease if a
clear arachnoid plane was preserved between the tumour and
adjacent brain parenchyma. However, when a clear arachnoid
plane disappeared, adjacent brain parenchyma as well as
tumour was intentionally removed to assess the brain invasion.
The remaining 14 cases without adjacent brain parenchyma
were unassessable to brain invasion because the existence of a
clear arachnoid plane helped the tumour to be completely
resected. These 14 cases did not show tumour recurrence in the
last follow-up neuroimaging studies. Of these 60 cases wherein
brain parenchyma was identified, 46 (77%) showed brain
invasion. The presence of brain invasion was associated with
decreased recurrence-free survival (p = 0.013) but had no
influence on overall survival (p.0.268).
The rate of p53 overexpression was observed in 10% (1/10) of
benign meningiomas, 25% (10/40) of atypical meningiomas, and
79% (19/24) of anaplastic meningiomas, and it became
significantly higher as histological malignancy increased
(p,0.001). The mean MIB-1 labelling indexes for p53 over-
expression-negative and p53 overexpression meningiomas were
3.3¡5.7% and 8.9¡10.5%, respectively (p = 0.010).
Overexpression of p53 was detected in 40.5% (30/74) of
meningiomas (15/51 nonrecurrent and 15/23 recurrent). There
was a significant correlation between p53 overexpression and
tumour recurrence (p = 0.004).
As shown in figure 1, overall and recurrence-free survivals
were all better for patients with benign compared with atypical
meningiomas (p,0.001) and for patients with atypical com-
pared with anaplastic meningioma (p,0.001). The mean MIB-1
labelling indexes for benign, atypical and anaplastic meningio-
mas were 0.8%, 3.2% and 12.0%, respectively (p,0.001) (fig. 2).
The 64 patients reclassified as having either atypical or
anaplastic meningioma were included in the present analysis.
Clinicopathological characteristics of these 64 patients are
summarised in table 2.
The female to male ratio was 2.1:1 for atypical meningioma
and 0.5:1 for anaplastic meningioma. This lack of a female
predominance in anaplastic meningiomas was significant
(p = 0.008).
Atypical meningioma
The mean clinical follow-up period was 63.6 months (range,
0.6–154.5). The mean overall and mean recurrence-free survivals
were 142.5¡6.0 months and 138.5¡7.0 months, respectively
(fig. 1). Ten-year overall and recurrence-free survival rates were
89.6% and 87.1%, respectively. In univariate testing, the extent
of resection (> Simpson grade III) and malignant progression
Table 1 Cross-tabulation of grade change on the basis of histopathological classification system
Initial grade
Grade according to the WHO 2000 classification system
I (Benign) II (Atypical) III (Anaplastic) Total
II 5* 26 2
{
33
III 5
{
14
1
22 41
Total 10 40 24 74
*These were previously considered to be atypical meningioma only because of a raised MIB-1 labelling index.
{
These tumours were reclassified based on a mitotic index of more than 20 mitotic figures per 10 high-power fields.
{
These five patients were previously diagnosed on the basis of the presence of brain invasion rather than obviously malignant
cytology.
1
Thirteen patients were previously diagnosed on the basis of the presence of brain invasion and 1 patient did not fulfill the
pathological criteria for anaplastic meningioma.
Research paper
J Neurol Neurosurg Psychiatry 2008;79:574–580. doi:10.1136/jnnp.2007.121582 575
group.bmj.com on December 23, 2009 - Published by jnnp.bmj.comDownloaded from
were significant unfavourable prognostic factors for recurrence-
free survival. In multivariate analysis, no factor was associated
with increased overall survival, except the extent of resection,
which had a borderline significance (p = 0.057). The extent of
resection was significantly associated with reduced recurrence-
free survival (OR 2.406; 95% CI 1.092 to 5.457; p = 0.012). Brain
invasion and adjuvant radiotherapy were not associated with
patient survival. Moreover, differences between the radio-
therapy subgroup and non-radiotherapy subgroup were not
significant for overall survival and recurrence-free survival.
However, in the radiotherapy subgroup, multivariate analysis
identified brain invasion as an independent unfavourable
prognostic factor for overall (OR 3.245; 95% CI 1.896 to
7.065; p = 0.025) and recurrence-free survival (OR 4.125; 95% CI
2.145 to 9.568; p = 0.007), whereas it had no effect on survival
in the non-radiotherapy subgroup. Age, MIB-1 labelling index,
tumour location, malignant progression and p53 overexpression
had no significant effect on either recurrence-free survival or
overall survival.
Of the four patients showing tumour recurrence, three were
treated using gamma knife radiosurgery and the other under-
went the second operation. At last follow-up, three patients had
died as a result of tumour progression.
Anaplastic meningioma
The mean clinical follow-up period of the 24 patients with
anaplastic meningioma was 42.5 months (range, 0.3–122.6). Of
these 24 patients, 17 died and 7 remained alive: 14 patients died
from tumour progression, 1 from postoperative bacterial
ventriculoencephalitis, 1 from a traffic accident and 1 from
acute myocardial infarction. The median overall survival was
39.8¡7.8 months, and the 3- and 5-year survival rates were
55% and 35%, respectively (fig 1A). Tumour recurrence occurred
in 75% (18/24) of patients, and 58% (14/24) of patients died
from tumour progression at a median of 5.1 months, with
spinal metastases developing in 3 of the 18 patients. The median
recurrence-free survival was 32.2¡8.0 months. The 3- and 5-
year recurrence-free survival rates were 50% and 29%, respec-
tively (fig 1B). Multivariate analysis identified brain invasion,
adjuvant radiotherapy, malignant progression, p53 overexpres-
sion and extent of resection as independent prognostic factors
for survival (table 3).
Of the 18 patients showing tumour recurrence, 5 patients
with 6 recurrent anaplastic meningiomas were treated using
gamma knife radiosurgery. The median tumour volume was
6.3 ml (range, 0.6–36.8) and the median margin dose was 13 Gy
(range, 10–21). The median neuroradiological follow-up period
was 14 months (range, 8–34). Four of 6 tumours showed a local
failure and 80% (4/5) of patients died from tumour recurrence at
the last follow-ups. One patient with tumour recurrence
received five courses of chemotherapy with carmustine, but he
died of tumour progression. Of the three patients with spinal
metastases, two underwent surgical resection followed by
radiotherapy and the other received radiotherapy only. These
3 patients did not show tumour recurrence in the last follow-up
spinal neuroimaging studies.
Malignant progression
Among the 64 patients with atypical or anaplastic meningio-
mas, 20 showed a histopathological progression towards a
higher grade that was associated with an aggressive clinical
course. Among these 20 meningiomas, 7 atypical and 10
anaplastic meningiomas were transformed from benign menin-
giomas, and 3 anaplastic meningiomas from atypical meningio-
mas. The risk of death was significantly higher for patients with
malignant progression than for patients without malignant
progression (OR 13.507; 95% CI 2.015 to 90.556; p = 0.007)
(fig 3A). The risk of recurrence was also higher for the patients
Figure 1 Kaplan–Meier survival curves
were significantly better for patients with
benign meningioma compared with
atypical meningioma (p,0.001), and for
patients with atypical meningioma
compared with anaplastic meningioma
(p,0.001). (A) Overall survival; (B)
Recurrence-free survival.
Figure 2 The proliferative marker, MIB-1, is directly correlated with the
histological grade of meningiomas (p,0.001). Mean labelling indexes
are 0.8% in benign, 3.2% in atypical and 12.0% in anaplastic
meningiomas.
Research paper
576 J Neurol Neurosurg Psychiatry 2008;79:574–580. doi:10.1136/jnnp.2007.121582
group.bmj.com on December 23, 2009 - Published by jnnp.bmj.comDownloaded from
with malignant progression (OR 4.793; 95% CI 1.614 to 14.239;
p = 0.005) (fig 3B). The mean period for malignant progression
was 70.0 months (range 10.7 to 118.3) from benign to atypical
meningiomas, 89.7 months (range 10.8 to 152.2) from benign to
anaplastic meningiomas, and 39.8 months (range 13.5 to 62.5)
from atypical to anaplastic meningiomas. Multivariate analysis
of these 20 tumours showed that p53 overexpression was the
only independent prognostic factor for malignant progression
(OR 5.753; 95% CI 1.551 to 21.329; p = 0.009).
The patients with malignant progression (75%) (52%, 80%
and 100% in benign to atypical, benign to anaplastic, and
atypical to anaplastic, respectively) had a significantly higher
percentage of p53 overexpression than the patients without
malignant progression (32%) (p = 0.001). The mean MIB-1
labelling indexes in the progression group and the non-
progression group were 8.0¡10.1% and 5.9¡8.1%, respectively,
but lacked statistical power. Tumours with malignant progres-
sion resulted in 35% (7/20) of atypical meningiomas and 65%
(13/20) of anaplastic meningiomas. On the other hand, tumours
without malignant progression resulted in 75% (33/44) of
atypical and 25% (11/44) of anaplastic patients. The proportion
of anaplastic meningioma among the tumours with malignant
progression was significantly higher than that among tumours
without malignant progression (p,0.001). Tumour recurrence
was observed in 70% (14/20) of the patients with malignant
progression and 18% (8/44) of the patients without malignant
progression. There was a significant correlation between
malignant progression and tumour recurrence (p,0.001).
Table 2 Clinicopathological characteristics of patients with atypical or anaplastic meningioma
Characteristics
Value
Atypical Anaplastic
Patients (number) 40 24
Follow-up period 62.8¡37.5 (range, 0.6–153.5) 42.2¡33.4 (range, 0.3–121.6)
Age at diagnosis (years) 49.0¡13.3 (range, 17–75) 51.1¡16.0 (range, 17–77)
Male/female ratio 13/27 16/8
Karnofsky performance score 80 (range, 40–100) 80 (range, 60–100)
Tumor location
Convexity 13 6
Parasagittal 10 6
Sphenoid ridge 5 3
Falx 3 4
Tuberculum sellae 2 1
Intraventricular 2 1
Tentorial 2 1
Olfactory groove 1
Foramen magnum 1 1
Cerebellopontine angle 1
Petroclival 1
Extent of resection (Simpson grading system
5
)
Grade I 16 6
Grade II 14 2
Grade III 4 8
Grade IV 6 4
Grade V 0 4
Adjuvant radiotherapy 23 17
Adjuvant chemotherapy 0 1
Palliative gamma knife radiosurgery 3 5
Extracranial metastasis 0 3 (spine)
MIB-1 labelling index (%) 3.2¡3.0 (range, 0.5–10.0) 12.0¡11.6 (range, 0.5–38.0)
Brain invasion 20 (50%) 21 (88%)
p53 overexpression 10 (25%) 19 (79%)
Values are expressed as mean¡SD.
Table 3 Potential prognostic factors for survival in anaplastic meningioma
Survival factors
Overall survival Recurrence-free survival
p Value Odds ratio (95% CI) p Value Odds ratio (95% CI)
Brain invasion 0.002 8.226 (2.251–30.026) ,0.001 9.157 (2.699–31.061)
Adjuvant radiotherapy 0.002 0.006 (0.000–0.150) ,0.001 0.004 (0.000–0.085)
Malignant progression 0.021 10.262 (2.753–52.144) 0.019 8.140 (1.981–33.481)
Extent of resection (I, II/III, IV, V) 0.001 2.529 (1.205–5.309) 0.018 2.121 (1.140–3.949)
p53 overexpression 0.034 3.019 (1.725–8.652) 0.026 2.878 (1.549–7.818)
Sex (male/female) 0.926 0.848 (0.027–27.005) 0.098 0.904 (0.587–2.233)
Age (>65/,65 years) 0.654 0.967 (0.836–1.119) 0.300 0.967 (0.907–1.031)
Tumour location (non-convexity/convexity) 0.090 0.514 (0.332–5.907) 0.210 0.733 (0.155–3.460)
MIB-1 labelling index ((5/.5) 0.541 1.033 (0.201–8.269) 0.062 1.486 (0.981–2.250)
Research paper
J Neurol Neurosurg Psychiatry 2008;79:574–580. doi:10.1136/jnnp.2007.121582 577
group.bmj.com on December 23, 2009 - Published by jnnp.bmj.comDownloaded from
