Tumor recurrence and malignant progression of gangliogliomas

3355
Tumor Recurrence and Malignant Progression
of Gangliogliomas
Michael Majores, MD1,3
Marec von Lehe, MD2
Jana Fassunke, PhD1
Johannes Schramm, MD2
Albert J. Becker, MD1
Matthias Simon, MD2
1
Department of Neuropathology, University of
Bonn Medical Center, Bonn, Germany.
2
Department of Neurosurgery, University of Bonn
Medical Center, Bonn, Germany.
3
Department of Pathology, University of Bonn
Medical Center, Bonn, Germany.
BACKGROUND. Most gangliogliomas (GGs) are benign tumors, but tumor recurrence and malignant progression are observed in some patients.
METHODS. The authors analyzed their experience with 4 recurrent/progressive
GGs (World Health Organization [WHO] grade I), 21 tumors with atypical features (WHO grade II), and 5 tumors with anaplastic histologic features (WHO
grade III). Histopathologic findings (23 patients) were reviewed. The mean follow-up was 5.9 years (median, 4.5 years; range, 0.5-14.7 years).
RESULTS. The 5-year survival rates were only 79% for patients who had tumors
with atypical features and 53% for patients who had WHO grade III tumors. Secondary glioblastomas were diagnosed in 5 of 11 patients (45%) who underwent
surgery for tumor recurrence. Age at surgery <40 years (P 5 .007) was associated
significantly with better overall survival (OS), but it was not associated with better progression-free survival (PFS). Clinical presentation (drug-resistant epilepsy
vs all other patients with seizures vs no seizures) was associated significantly
with better OS (P 5 .005) and PFS (P < .001). Patients who had extratemporal
tumors had a significantly shorter PFS (P 5 .01) but not OS. A complete resection
was correlated strongly with both OS (P 5 .002) and PFS (P 5 .001). Neuropathologic examination revealed the presence of a gemistocytic cell component (PFS,
P 5 .025), a lack of protein droplets (OS, P 5 .04; PFS, P 5 .05), and focal tumor
cell-associated CD34 immunolabeling (OS, P 5 .03) as significant predictors of an
adverse clinical course.
CONCLUSIONS. The current data supported a 3-tiered GG histopathologic grading
system that included an intermediate diagnostic category (atypical GG, WHO
grade II). Careful attention to histopathologic findings and clinical parameters
usually will identify patients who are at risk for an adverse clinical course.
Cancer 2008;113:3355–63. 2008 American Cancer Society.
KEYWORDS: ganglioglioma, malignancy, recurrence, neuropathology.
Supported by the Deutsche Krebshilfe, the
German Glioma Network, the Deutsche Forschungsgemeinschaft (SFB TR3), the German–Israeli collaborative research program of the Bundesministerium
fur Bildung und Forschung, the Ministry of Science,
and Bonn Forschung.
The first and second authors contributed equally to
this article.
Address for reprints: Michael Majores, MD,
Department of Neuropathology, University of
Bonn Medical Center, Sigmund-Freud Strasse 25,
D-53105 Bonn, Germany; Fax: (011) 49-228287-14331; E-mail: [email protected]
Received April 16, 2008; revision received July
6, 2008; accepted July 25, 2008.
ª 2008 American Cancer Society
G
angliogliomas (GGs) are glioneuronal tumors composed of neoplastic glial and dysplastic neuronal elements. Both cell populations show marked heterogeneity, ranging from a predominantly
neuronal phenotype to variants with a prominent glial population.
GGs are characterized by a high epileptogenic potency. GGs are the
most common tumors encountered in young patients who have
chronic temporal lobe epilepsy.1-3 Histopathologic differential diagnoses comprise both high-grade and low-grade neoplasms, such as
diffuse astrocytomas, oligodendrogliomas, dysembryoplastic neuroepithelial tumors, pilocytic astrocytomas (PAs) and pleomorphic
xanthoastrocytomas (PXAs).4-8
The pathogenesis of GG is not very well understood. Individual
cases have been described in association with a family history of
neurofibromatosis,9 in a patient with Peutz-Jeghers syndrome,10 and
DOI 10.1002/cncr.23965
Published online 5 November 2008 in Wiley InterScience (www.interscience.wiley.com).
3356
CANCER
December 15, 2008 / Volume 113 / Number 12
in a patient with Turcot syndrome.11 An anaplastic
GG was detected in the Eker rat, a model for human
tuberous sclerosis carrying a mutation of the tuberous-sclerosis complex 2 (TSC2) gene.12 However,
patients with epilepsy who have GG usually lack the
typical clinical stigmata of tuberous sclerosis.13 TSC
gene mutations have been identified in a few GGs.14
It has been proposed that malignant GGs have a
monoclonal origin, suggestive of an initial transformation of a single neuroglial precursor cell with
subsequent malignant progression.15 Loss of p19
expression and p53 mutations also may be associated with malignant progression.15,16
Several studies have suggested a very benign
clinical course in most patients with GG. However,
tumor recurrence, malignant progression, and secondary glioblastoma multiforme (GBM) are observed
in some patients.17-20 Malignant transformation has
been associated with incomplete tumor resections18
and radiation therapy.21
Overall, only sparse information is available with
respect to patients who had tumors with an adverse
clinical course. Therefore, for the current study, we
analyzed our experience with recurrent and/or histologically atypical or anaplastic GG. Our results support a 3-tiered GG histopathologic grading system
that includes an intermediate diagnostic category
(atypical GG; World Health Organization [WHO]
grade II). Our data also indicate that an adverse clinical course in a patient with GG often can be predicted based on histopathologic findings and clinical
data, such as age, clinical presentation, tumor location, and degree of resection.
MATERIALS AND METHODS
Patients and Clinical Data
In a series of 203 patients with GG who underwent
surgery in the Department of Neurosurgery at the
University of Bonn from 1992 to 2006 (excluding
brainstem and spinal cord tumors), we identified 5
patients with anaplastic GG, WHO grade III (2.5%); 4
patients with recurrent/progressive GG, WHO grade I
(2%); and 21 patients (10.3%) who had tumors with
atypical histologic characteristics, such as increased
cellularity and increased mitotic activity, but that
lacked unequivocal anaplastic features.22 These latter
tumors were diagnosed as ‘atypical GG, WHO grade
II.’ Ten patients from the current series were
included in the large collection of mostly benign GGs
reported by Luyken and coworkers.19 Our series did
not include any patients with newly diagnosed GG
who had WHO grade IV (glioblastomatous) changes
in the glial component.23 We cannot exclude the pos-
sibility that some patients with recurrent WHO grade
I GG may have presented at another institution.
However, this does not seem very likely. A previous
report from our institution that covered the years
1988 through 2001 with a median follow-up of 8
years included only 1 recurrent tumor in a total of
171 WHO grade I GGs.19 Informed, written consent
was obtained from all patients.
The respective patients’ charts, operative notes,
preoperative and postoperative radiology reports,
and/or magnetic resonance images (MRIs) were
reviewed. Follow-up data were obtained through telephone interviews if necessary. The mean follow-up
was 5.9 years (median, 4.5 years; range, 0.5-14.7
years). Routine follow-up consisted of clinical and
MRI examinations every year for patients with WHO
grade I GG, twice a year for patients with WHO grade
II GG, and every 4 months for patients with WHO
grade III GG. Tumor recurrence/progression was
defined as radiographic evidence of tumor regrowth
with or without clinical signs.
Tumors were resected by using standard supratentorial approaches. In patients who had drugresistant epilepsy, the resection always included a 0.5
to 1 cm rim of adjacent cortical tissue; frequently
included an adjoining ictogenic cortical area; and, in
some patients who underwent temporal resection, it
included the hippocampus and the amygdala, as
indicated by the preoperative workup or intraoperative corticography. Multiple subpial transections were
used in 1 patient with a left temporo-occipital GG.24
Adjuvant therapy consisted of fractionated radiotherapy in 2 patients and radiotherapy and chemotherapy with combined procarbazine, lomustine, and
vincristine in 1 of 5 patients with WHO grade III
tumors. No adjuvant therapies were prescribed for
patients with WHO grade I or II GG. All but 3 of 13
patients with recurrent tumors underwent surgery.
Treatment at first tumor recurrence included radiotherapy in all patients who had histologically proven,
secondary GBM and in 1 patient who had a recurrent
WHO grade II tumor. Tumor recurrence after radiotherapy was treated with chemotherapy (temozolomide in 5 patients, nitrosourea-based protocols in 3
patients).
Neuropathology
Surgical specimens were fixed in formaldehyde overnight and embedded in paraffin. Specimens were
reviewed at the Institute for Neuropathology/German
Brain Tumor Reference Center at the University of
Bonn and were classified according to the WHO 2000
classification system for central nervous system
tumors.22 In total, 23 specimens from primary
Recurrence and Malignancy in GG/Majores et al
3357
FIGURE 1. Histopathologic characteristics of ganglioglioma (GG). (a) Typical GG specimens are characterized by a biphasic (ie, glial and neuronal) phenotype
(hematoxylin and eosin stain; original magnification, 3200). (b) The presence of dysplastic neuronal elements is highlighted by immunohistochemistry (IHC)
using antibodies against synaptophysin, frequently revealing a perisomatic accumulation (original magnification, 3200). (c) The lack of a dense reticulin fiber
network is helpful in the differential diagnosis between GG and pleomorphic xanthoastrocytomas (original magnification, 3200). (d) Perifocal occurrence of ramified CD34 immunoreactivity is highly suggestive of a GG (original magnification, 3200). (e) In some GG examples, CD34 immunoreactivity also was detected in
association with tumor cell processes (original magnification, 3400). (f) Malignant progression to a glioblastoma multiforme usually reveals a loss of GG features, including loss of CD34 immunolabeling (original magnification, 3200). Increased proliferative activity (g) (Ki67 [MIB-1]; original magnification, 3400) and
(h) a prominent gemistocytic tumor cell component (h) are atypical findings in GG (glial fibrillary acidic protein; original magnification, 3200). (i) Malignant progression to a glioblastoma is indicated by a brisk mitotic and proliferative activity as well as palisading necrosis in concert with loss of GG features (IHC; original magnification, 3100).
tumors and 12 recurrent samples (from 9 patients)
were available for a retrospective neuropathologic
workup. In addition to conventional hematoxylin
eosin stains, immunohistochemical analyses using
antibodies against glial and neuronal epitopes were
performed to reveal the biphasic nature of the
tumors. Typical histopathologic and immunohistochemical characteristics are summarized in Figure 1.
Statistical Analysis
Commercially available software was used for standard statistical analysis (SPSS version 14.0; SPSS Inc.,
Chicago, Ill). For survival analysis, Kaplan-Meier estimates and log-rank tests were used. The level of significance was set at P < .05 (2-sided).
RESULTS
Demographics and Clinical Presentation
The current study was comprised of 21 atypical
WHO grade II tumors, 5 anaplastic WHO grade III
tumors with or without recurrence, and 4 recurrent/
progressive GGs that initially were assigned WHO
grade I. Pertinent clinical data are summarized in
Table 1. The mean age at initial surgery was 19.8
3358
CANCER
December 15, 2008 / Volume 113 / Number 12
TABLE 1
Clinical Characteristics of Gangliogliomas (World Health Organization
[WHO] Grade I), Atypical Gangliogliomas (WHO Grade II), and
Anaplastic Gangliogliomas (WHO Grade III)
Gangliogliomas
Characteristic
No. of patients
Age, y
Mean
Median
Range
Aged 16 y, %
Men, %
Drug-resistant epilepsy, %
Tumor location, %
Temporal
Frontal
Other
Recurrence rate, % (no./total)
Malignant progression, % (no./total)
5-Y survival, %
5-Y PFS, %
WHO
Grade I
WHO
Grade II
WHO
Grade III
177
21
5
25
23
1-64
28
51
91
33.5
32
8-45
19
67
62
23.4
22
3-60
40
80
40
78
8
14
2 (4/177)
0.6 (1/177)
—
99
97
62
24
14
33 (7/21)
14 (3/21)
24 (5/21)*
79
70
80
20
—
60 (3/5)
20 (1/5)
40 (2/5)*
53
30
PFS indicates progression-free survival.
* Includes patients who had malignant progression diagnosed by magnetic resonance imaging only
(ie, without histologic confirmation).
years for patients with recurrent/progressive grade I
tumors (median age, 17.5 years; range, 10–34 years.).
Refractory epilepsy was observed in 3 patients (75%)
with recurrent/progressive WHO grade I tumors.
Clinical Prognostic Factors
Age at surgery <40 years (P 5 .007) was associated
significantly with better overall survival (OS), but not
with better progression-free survival (PFS). Clinical
presentation was associated significantly with OS
(P 5 .005) and PFS (P < .001). Patients with drug-resistant epilepsy fared best, the prognosis for all other
patients who presented with seizures was intermediate,
and survival was worst in patients without seizures.
Tumor Location and Radiologic Findings
Temporal lobe tumors were diagnosed in 17 patients,
including 10 temporomesial GGs. In addition, there
were 3 insular GGs with temporal lobe extensions,
and there was 1 temporo-occipital GG. Five frontal
lobe GGs were observed, and there were 4 tumors of
the parietal and/or occipital lobe. In 1 patient with a
WHO grade II GG, we observed an additional contrast-enhancing lesion in the fourth ventricle that
remained stable in size over a follow-up of 176
months. Patients with extratemporal tumors had a
significantly shorter PFS (P 5 .01), but OS was not
affected.
All available neuroimaging data from our
patients were reviewed. The preoperative differential
diagnosis included GG in 47% of patients. No specific
neuroimaging patterns that were suggestive of the
presence of WHO grade II or III GG or that pointed
to an adverse clinical course in general could be delineated. However, perifocal edema was observed in
26% of patients. Typical imaging findings (42% of
patients) included strongly contrast-enhancing tumor
nodules (or at least some contrast-enhancing areas)
often accompanied by intratumoral cysts within a
temporal (or frontal) tumor. The nonenhancing
tumor component involved the cerebral cortex as
well as white matter and was delineated best on
fluid-attenuated inversion recovery images. Only
19% of tumors showed no contrast enhancement.
Intratumoral cysts were observed in 50% of patients
(Fig. 2). Computed tomography scans were available
from 11 patients, and 72% of those scans showed
calcifications.
Surgical Therapy and Epileptologic Outcomes
Tumor growth in the basal ganglia and/or insular
region prevented radical surgery in some patients.
Complete resection was achieved at 80% of primary
surgeries. A complete resection proved to be a strong
predictor of both OS (P 5 .002) and recurrence-free
survival (P 5 .001).
Intraoperative findings that suggested a diagnosis
of GG, ie, the presence of a relatively firm, well delineated, and sometimes cystic and calcified mass of
yellow-brown or gray color, were recorded in only 11
of 21 patients (52%) with WHO grade II tumors and
in 1 of 5 patients (20%) with WHO grade III tumors.
Postoperatively, there was 1 new minor dysphasia, 1
patient with worsened hemiparesis, and 1 patient
with worsened hemiparesis and dysphasia. Visual
field defects (quadrantanopias) were observed after
temporomesial and/or occipital resection in 9
patients. There was no mortality.
Epileptologic outcomes were available for 15 of
18 patients who presented with drug-resistant epilepsy. Thirteen patients (87%) were seizure-free or
suffered only from isolated, nondebilitating seizures
at the 1-year follow-up (Engels Class I).25
Tumor Recurrence and Malignant Progression
Kaplan-Meier estimates of OS and PFS are shown in
Figure 3. The 5-year survival rates for patients with
atypical GG (WHO grade II) and grade III tumors
were 79% and 53%, respectively (see also Table 1).
The recurrence rates were 33% (7 of 21 patients) for
Recurrence and Malignancy in GG/Majores et al
atypical GG (median follow-up, 55 months; range, 6176 months) and 60% (3 of 5 patients) for WHO
grade III tumors (median follow-up, 38 months;
range, 15-118 months). The mean time interval
between the initial surgery and tumor recurrence
was 23 19 months (median, 18 months; range 6-72
months).
Malignant progression to glioblastoma was observed in 5 of 11 patients (45%; 4 men and 1 woman)
who underwent surgery for recurrent tumors (Table
1). Secondary glioblastoma was observed in 1 of 4
patients (25%) with recurrent WHO grade I GG, in 3
of 21 patients (14%) with atypical GG (Fig. 2C,D),
and in 1 of 5 patients (20%) with WHO grade III GG.
The mean age at initial surgery of patients who later
suffered from secondary GBM was 45.2 years (me-
3359
dian, 60 years; range, 22-60 years) as compared with
27 years (median, 29 years; range, 3–55 years) for
patients without malignant progression (P 5 .03).
The primary tumor was located in the temporal lobe
in 2 patients, in the frontal lobe in 2 other patients,
and in the parietal lobe in 1 patient. Only 2 patients
FIGURE 2. Magnetic resonance imaging (MRI) features of anaplastic
(World Health Organization [WHO] grade III) gangliogliomas (GG) (a) and atypical GG (WHO grade II) (b-e). (a) A man aged 45 years with a left temporomesial, anaplastic GG (WHO grade III) who presented with a single,
generalized seizure. From left to right: T1-weighted axial scan with a typical
modular contrast enhancing lesion. T2-weighted coronal image depicting the
tumor’s (mild) mass effect. A sagittal fluid-attenuated inversion recovery
(FLAIR) scan shows the true extent of the lesion. The MRI was read as typical GG. The tumor was completely resected, and the patient underwent adjuvant radiotherapy. Tumor recurrence consistent with glioblastoma multiforme
(GBM) (WHO grade IV) according to MRI criteria was diagnosed after 11
months. Despite chemotherapy with temozolomide, the patient died from tumor progression 4 months later. (b) A young woman aged 16 years with a
left temporopolar, atypical GG WHO grade II. The patient suffered from focal
seizures. From left to right: T1-weighted axial scan after administration of
contrast medium. Axial FLAIR image. T2-weighted axial scan. This MRI was
also read as typical GG. Note similarities to the patient shown in a. A complete resection was performed. The patient was followed for 6 months without evidence of tumor recurrence. (c) A man aged 60 years with a right
frontocentral, atypical GG (WHO grade II) and refractory epilepsy. From left to
right: T1-weighted axial scans before and after application of contrast medium. Axial FLAIR scan. The tumor could be completely resected. Tumor recurrence was diagnosed 20 months later. The patient died 5 months after
reoperation (histology: GBM, WHO grade IV) and radiotherapy from tumor
progression. (d) A man aged 60 years with a left temporoinsular, atypical GG
(WHO grade II) involving the basal ganglia and that presented with hemiparesis. From left to right: T1-weighted coronal scan before and sagittal scan after contrast application. Axial FLAIR image. The MRI depicts a large mass
that includes many small cysts and strong contrast enhancement. There is
not much perifocal edema. Tumor progression was noted only 6 months after
a subtotal (>70%) resection (residual tumor in the basal ganglia). The
patient died 5 months after reoperation (histology: GBM, WHO grade IV),
radiotherapy, and chemotherapy (temozolomide). (e) A man aged 37 years
with a left frontocentroinsular tumor involving the basal ganglia. The patient
presented with a single, generalized seizure and minimal dysphasia. From
left to right: MR scans obtained after an initial biopsy (histology: atypical GG,
WHO grade II). A T1-weighted coronal scan reveals a large tumor with a
considerable mass effect and focal contrast enhancement. Axial FLAIR
images depict tumor in the frontocentral region, the insula, and the basal
ganglia. There is surprisingly little perifocal edema. A subtotal resection was
performed (90% of the tumor). Symptomatic tumor progression 10 months
later (repeat biopsy: atypical GG, WHO grade II) was treated successfully
with radiotherapy and temozolomide chemotherapy. At the time of the current report (39 months after the first biopsy), the patient was evaluated for
salvage chemotherapy for further clinical progression.
3360
CANCER
December 15, 2008 / Volume 113 / Number 12
FIGURE 3. Kaplan-Meier estimates of overall survival (OS) and progression/recurrence-free survival (PFS). The World Health Organization (WHO) tumor grade
was correlated with OS (P < .001) (a) and PFS (P < .001) (b). For comparison, survival data are provided from the patients with WHO grade I tumors who were
treated during the study period. A complete resection resulted in longer OS (P 5 .002) (c) and longer PFS (P 5 .001) (d). Clinical presentation proved to be an
important prognostic parameter: OS (P 5 .005) (e) and PFS (d) were best in patients with refractory epilepsy and worst in patients without seizures.
initially presented with drug-resistant epilepsy. One
patient with a secondary glioblastoma who initially
was diagnosed with a GG WHO grade III of the right
temporal lobe developed leptomeningeal seeding
shortly before his death. Follow-up MRI scanning
diagnosed recurrent tumors consistent with glioblas-
toma in 2 additional patients with atypical GG and
in 1 patient with WHO grade III GG who did not
undergo repeat surgery. All patients with histologically proven or presumed secondary glioblastomas
died from their disease, accounting for all 8 tumorrelated deaths in this series.
Recurrence and Malignancy in GG/Majores et al
Histologic Analyses and Prognosis
Specific histopathologic features were tested as
possible predictors of an adverse clinical course.
Twenty-four percent of specimens from the current
series contained a substantial gemistocytic component, which was correlated significantly with shorter
PFS (P 5 .025). The presence of protein droplets predicted longer PFS (P 5 .05) and OS (P 5 .04). Expression of the stem cell marker CD34, which is
expressed transiently during early neurulation, was
assessed by immunohistochemistry.26 Seventy-seven
percent of the initial tumor specimens, but only 18%
of recurrent samples, demonstrated satellitosis-like,
perilesional CD34 expression. Intralesional CD34
labeling, ie, (focal) staining of the tumor cell processes, was observed in 52% of tumor samples and
was correlated with shorter OS in the current series
(P 5 .03). Somatic accumulation of synaptophysin or
neurofilament protein27 was observed in 39% and
54% of specimens, respectively. No correlations with
OS or PFS were observed.
DISCUSSION
In most patients, the prognosis after surgery for intracranial GG is good. Many patients can expect to
be cured from their tumors and also often from their
epilepsy.17-21,28 However, some patients will suffer
from recurrent tumor and/or malignant progression.
In 1 series of 58 GGs, 40 tumors were assigned to
histologic grade I, but 16 tumors were grade II, and 2
tumors were grade III. The event-free 5-year survival
rate in that cohort was 95% for GG of the cerebral
hemispheres.28 Rumana et al observed a surprisingly
high number of tumor-related deaths (n 5 10) in a
series of 42 patients with supratentorial GG.20
Blumcke and Wiestler reported 30 WHO grade II
tumors (9%) and 17 WHO grade III tumors (5%) in a
cohort of 326 patients with GG. Tumor recurrence
was observed in 25% of patients with grade II tumors
and in 38% of patients with grade III tumors.17 Similarly, we identified 21 WHO grade II tumors (10.3%)
and 5 WHO grade III tumors (2.5%) in a series of 203
patients with GG. The overall recurrence rate was 7%
(14 of 203 patients). Recurrence rates for patients
with WHO grade II and III tumors were 33% and
60%, respectively. Im and coworkers studied 34
patients with GG and observed tumor progression in
3 patients: Two of those patients had secondary glioblastomas identified at their second surgery.18 Secondary glioblastomas were diagnosed in 5 of 11
patients (45%) in our series who underwent repeat
surgery. GG differs vastly from diffuse low-grade glioma (LGG) with respect to the rate of recurrence.
3361
Nevertheless, a similar 50% rate of malignant progression was reported in patients who underwent
surgery for LGG recurrence.29 Similar to LGG
regrowth, GG recurrence is not a benign condition.
Histologic grading for GG is controversial. The
new WHO 2007 classification distinguishes between
benign GG WHO grade I, and anaplastic WHO grade
III tumors. An intermediate category (formerly
labeled atypical GG WHO grade II) is no longer
included in the WHO grading system.6,22 The diagnosis of an atypical GG, WHO grade II was applied to
tumors that had increased cellularity, nuclear pleomorphy, and/or increased proliferative activity in the
glial cell component but that lacked definitive criteria for anaplasia, such as palisading necrosis or a
brisk mitotic count. Seventy percent of the tumors analyzed in the current study belonged to this
category.
Our data support a 3-tiered rather than a 2-tiered
histopathologic grading system for GG. The 5-year
survival rate for atypical (WHO grade II) GG was 79%
in this series. The recurrence rate of these growths
was 33% after a median follow-up of 55 months
compared with only 2% after surgery for WHO grade
I tumors. Conversely, outcomes after surgery for
WHO grade III GG were considerably worse (recurrence rate, 60%; 5-year survival rate, 53%) than outcomes for atypical (WHO grade II) tumors despite
postoperative radiotherapy in 4 of 5 patients. On the
basis of these data, a distinction of WHO grade II
GG from both GG WHO grades I and III seems
warranted, because it provides valuable prognostic
information.
To arrive at a better histologic characterization of
atypical and anaplastic GG, we investigated specific
histopathologic features for possible correlations
with the patient outcomes in our series. A gemistocytic differentiation pattern in GG has been suggested
as a possible marker for atypia.19 Indeed, the presence of a gemistocytic cell component was correlated significantly with shorter PFS in the current
series. Conversely, patients who had tumors that
contained protein droplets had better OS and longer
PFS. This is not surprising, because protein droplets
are a frequent finding in typical benign GG and other
low-grade neoplasms, such as PA, in young patients.
Finally, tumor cell process-associated CD34 immunoreactivity, albeit only focal in most instances, was
correlated with shorter OS. In such patients, a PXA
has to be ruled out, especially for extratemporal
tumor locations. It is noteworthy that composite
PXA-GG lesions have been reported.8,30,31
Can an adverse clinical outcome also be predicted by clinical parameters? Most GG series include
3362
CANCER
December 15, 2008 / Volume 113 / Number 12
slightly more patients who are men.19 It has been
suggested that being a man rather than a woman
may be a negative prognostic parameter.20 Indeed,
men clearly were over-represented among the recurrent and WHO grade II and III tumors in the current
study. Age 40 years was correlated strongly with
shorter OS (but not PFS, although patients with
tumor recurrence after resection for GG WHO grade
II and III, on average, were older than those without
recurrence). There was a significant correlation
between age at initial surgery and the formation of a
secondary GBM. Older age also has been associated
with an adverse prognosis by other authors.20 It was
suggested that anaplastic GGs occur more often in
older patients compared with their nonmalignant
counterparts.5,17,19,32
Clinical presentation was correlated strongly
with both OS and PFS in this series. The best outcomes were observed in patients with drug-resistant
epilepsy. The prognosis was worst in patients without
seizures. Tumor site was correlated with PFS but not
with OS in the current analysis. All 5 frontal tumors
from this series recurred. The analysis of GG series
that included mostly benign tumors has provided
similar data, but some studies failed to identify these
associations.18-20
Finally, the degree of resection proved to be an
important prognostic parameter in our patients. A
strong correlation was observed with OS and PFS.
Similar results were obtained by Luyken and coworkers in their series, which included mostly benign
tumors.19 Im and coworkers reported tumor recurrences after 3 of 9 incomplete resections, and none
of their completely resected tumors relapsed.18 These
data support the role of aggressive surgery in the
management of higher grade GG. In contrast to all
other prognostic parameters discussed here, the
degree of resection is the only variable that can be
influenced, at least in part, by the treating physician.
We conclude from our data that the results of a
careful histopathologic analysis, together with certain
clinical parameters allow an individualized assessment of the risk of tumor recurrence and tumorrelated death in patients with GG. The typical patient
who is at risk for an adverse clinical course is 1) a
man, 2) aged >40 years, 3) who presents with signs
and symptoms other than (drug- resistant) epilepsy,
and 4) has undergone an incomplete resection for 5)
an extratemporal tumor. Histopathologic analysis will
reveal 6) some atypical features (such as the presence of a gemistocytic cell component) or some
overtly anaplastic features.
Early repeat surgery for resectable residual tumor
should be contemplated in such patients, and
patients should undergo regular surveillance imaging
for more than 5 years. Our data do not allow us to
make firm conclusions with respect to adjuvant
therapies. Radiotherapy and chemotherapy may have
some efficacy (Fig. 2).
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
Wolf HK, Muller MB, Spanle M, Zentner J, Schramm J,
Wiestler OD. Ganglioglioma: a detailed histopathological
and immunohistochemical analysis of 61 cases. Acta Neuropathol (Berl). 1994;88:166-173.
Wolf HK, Wiestler OD. Neuropathologic findings in chronic
epilepsy. Laryngorhinootologie. 1995;74:127-129.
Zentner J, Wolf HK, Ostertun B, et al. Gangliogliomas:
clinical, radiological, and histopathological findings in 51
patients. J Neurol Neurosurg Psychiatry. 1994;57:1497-1502.
Evans AJ, Fayaz I, Cusimano MD, Laperriere N, Bilbao JM.
Combined pleomorphic xanthoastrocytoma-ganglioglioma
of the cerebellum. Arch Pathol Lab Med. 2000;124:17071709.
Hirose T, Scheithauer BW, Lopes MB, Gerber HA, Altermatt
HJ, VandenBerg SR. Ganglioglioma: an ultrastructural and
immunohistochemical study. Cancer. 1997;79:989-1003.
Louis DN, Ohgaki H, Wiestler OD, et al. The 2007 WHO
classification of tumours of the central nervous system.
Acta Neuropathol. 2007;114:97-109.
Prayson RA. Composite ganglioglioma and dysembryoplastic neuroepithelial tumor. Arch Pathol Lab Med. 1999;123:
247-250.
Yeh DJ, Hessler RB, Stevens EA, Lee MR. Composite pleomorphic xanthoastrocytoma-ganglioglioma presenting as a
suprasellar mass: case report. Neurosurgery. 2003;52:14651469.
Mekni A, Chelly I, Haouet S, Zitouna M, Kchir N. [Malignant cerebellar ganglioglioma. A case report and review of
the literature]. Neurochirurgie. 2006;52(2-3 pt 1):119-122.
Resta N, Lauriola L, Puca A, et al. Ganglioglioma arising in
a Peutz-Jeghers patient: a case report with molecular
implications. Acta Neuropathol. 2006;112:106-111.
Tamiya T, Hamazaki S, Ono Y, et al. Ganglioglioma in a
patient with Turcot syndrome. Case report. J Neurosurg.
2000;92:170-175.
Mizuguchi M, Takashima S, Yamanouchi H, Nakazato Y,
Mitani H, Hino O. Novel cerebral lesions in the Eker rat
model of tuberous sclerosis: cortical tuber and anaplastic
ganglioglioma. J Neuropathol Exp Neurol. 2000;59:188-196.
Roach ES, Gomez MR, Northrup H. Tuberous sclerosis
complex consensus conference: revised clinical diagnostic
criteria. J Child Neurol. 1998;13:624-628.
Becker AJ, Lobach M, Klein H, et al. Mutational analysis of
TSC1 and TSC2 genes in gangliogliomas. Neuropathol Appl
Neurobiol. 2001;27:105-114.
Pandita A, Balasubramaniam A, Perrin R, Shannon P, Guha
A. Malignant and benign ganglioglioma: a pathological and
molecular study. Neuro Oncol. 2007;9:124-134.
Fukushima T, Katayama Y, Watanabe T, Yoshino A, Komine
C, Yokoyama T. Aberrant TP53 protein accumulation in the
neuronal component of ganglioglioma. J Neurooncol. 2005;
72:103-106.
Blumcke I, Wiestler OD. Gangliogliomas: an intriguing tumor entity associated with focal epilepsies. J Neuropathol
Exp Neurol. 2002;61:575-584.
Recurrence and Malignancy in GG/Majores et al
18. Im SH, Chung CK, Cho BK, et al. Intracranial ganglioglioma: preoperative characteristics and oncologic outcome after surgery. J Neurooncol. 2002;59:173-183.
19. Luyken C, Blumcke I, Fimmers R, Urbach H, Wiestler OD,
Schramm J. Supratentorial gangliogliomas: histopathologic
grading and tumor recurrence in 184 patients with a median follow-up of 8 years. Cancer. 2004;101:146-155.
20. Rumana CS, Valadka AB, Contant CF. Prognostic factors
in supratentorial ganglioglioma. Acta Neurochir (Wien).
1999;141:63-69.
21. Rumana CS, Valadka AB. Radiation therapy and malignant
degeneration of benign supratentorial gangliogliomas. Neurosurgery. 1998;42:1038-1043.
22. Kleihues P, Cavenee WK. Pathology and Genetics of
Tumours of the Nervous System. Lyon, France: IARC; 2000.
23. Dash RC, Provenzale JM, McComb RD, Perry DA, Longee
DC, McLendon RE. Malignant supratentorial ganglioglioma
(ganglion cell-giant cell glioblastoma): a case report and
review of the literature. Arch Pathol Lab Med. 1999;123:
342-345.
24. Schramm J, Clusmann H. The surgery of epilepsy. Neurosurgery. 2008;62(suppl 2)463-481; discussion 481.
25. Engel J Jr. Outcome with respect to epileptic seizures. In:
Engel J Jr, ed. Surgical Treatment of the Epilepsies. New
York, NY: Raven Press; 1993:609-621.
3363
26. Blumcke I, Giencke K, Wardelmann E, et al. The CD34 epitope is expressed in neoplastic and malformative lesions
associated with chronic, focal epilepsies. Acta Neuropathol.
1999;97:481-490.
27. Wierzba-Bobrowicz T, Schmidt-Sidor B, Gwiazda E, Bertrand E. The significance of immunocytochemical markers,
synaptophysin and neurofilaments in diagnosis of ganglioglioma. Folia Neuropathol. 1999;37:157-161.
28. Lang FF, Epstein FJ, Ransohoff J, et al. Central nervous system gangliogliomas. Part 2: clinical outcome. J Neurosurg.
1993;79:867-873.
29. Schmidt MH, Berger MS, Lamborn KR, et al. Repeated
operations for infiltrative low-grade gliomas without intervening therapy. J Neurosurg. 2003;98:1165-1169.
30. Furuta A, Takahashi H, Ikuta F, Onda K, Takeda N, Tanaka
R. Temporal lobe tumor demonstrating ganglioglioma and
pleomorphic xanthoastrocytoma components. Case report.
J Neurosurg. 1992;77:143-147.
31. Perry A, Giannini C, Scheithauer BW, et al. Composite
pleomorphic xanthoastrocytoma and ganglioglioma: report
of 4 cases and review of the literature. Am J Surg Pathol.
1997;21:763-771.
32. Brainer-Lima PT, Brainer-Lima AM, Azevedo-Filho HR.
Ganglioglioma: comparison with other low-grade brain
tumors. Arq Neuropsiquiatr. 2006;64(3A):613-618.