Frequent Recurrence and Progression in Pilocytic

Astrocytoma in Adults
Carsten St uer, MD
1
Belinda Vilz
1
Michael Majores, MD
2
Albert Becker, MD
2
Johannes Schramm, MD
1
Matthias Simon, MD
1
1
Department of Neurosurgery, University of Bonn
Medical Center, Bonn, Germany.
2
Institute of Neuropathology, University of Bonn
Medical Center, Bonn, Germany.
BACKGROUND. Most pilocytic astrocytomas (piloA) are benign growths (World
Health Organization [WHO] grade 1) of the deep midline structures, the brain-
stem, and the cerebellum. To the authors knowledge, the literature contains only
scarce data regarding piloA in adults.
METHODS. Between 1995 and 2005, 44 patients (26 women and 18 men) under-
went surgery for a primary or recurrent piloA at the authors institution. All
patients were aged > 16 years (mean standard deviation: 31 14 years) at the
time of their first surgery. The histopathologic diagnoses were reviewed, and rele-
vant clinical information was obtained through a chart review and telephone
interviews. The mean follow-up was 76 59 months (range, 1227 months).
RESULTS. There were 20 patients (45%) with supratentorial lobar piloA (including
10 temporal/temporomesial tumors, 5 parietal tumors, 3 insular tumors, 1 frontal
tumor, and 1 occipital tumors), 12 patients with cerebellar piloA, 7 patients with
brainstem piloA, 2 patients with opticochiasmatic PiloA, 1 patient with intramed-
ullary piloA, and 2 patients with piloA of the basal ganglia. All but 1 patient with
a lobar tumor presented with epilepsy. In 6 of 44 patients (14%), increased prolif-
erative activity was revealed. WHO grade 3 primary anaplastic piloA was diag-
nosed in 2 patients (5%), and WHO grade 3 secondary anaplastic piloA was
diagnosed in 4 patients (9%). Tumor recurrence or disease progression was
observed in 13 of 44 patients (30%). Eight of 44 patients (18%) died from their
disease. Histologic grading and extent of surgical resection proved to be impor-
tant predictors of survival.
CONCLUSIONS. PiloA in adult patients, surprisingly, often was not a benign dis-
ease. The degree of surgical resection was found to be of major importance for
the patients further clinical course; therefore, an aggressive surgical resection
should be attempted whenever possible. Cancer 2007;110:2799808. 2007
American Cancer Society.
KEYWORDS: pilocytic astrocytoma, brain tumor, adult, microsurgery, World Health
Organization classification.
P
ilocytic astroytomas (piloA) account for up to 25% of all brain
tumors encountered in pediatric neurosurgical practice. Conver-
sely, piloA are diagnosed only rarely in adults. Overall, 2.3% of all
brain tumors are classified as piloA in large neuropathologic series.
1
The incidence rate (age-adjusted to the World Standard Population)
was calculated as 4.8 per 1 million per year in 2003.
2
PiloA most often arise from the deep cerebral midline struc-
tures, the brainstem, and the cerebellum. Lobar growth is observed
less often.
2,3
Whereas most PiloA only demonstrate a low prolifera-
tive activity, some specimens may exhibit increased proliferative ac-
tivity that is not caused by inflammatory changes. Nonetheless,
increased proliferative activity is not a definitive sign of malignancy.
Address for reprints: Matthias Simon, MD, Neuro-
chirurgische Universitatsklinik, Sigmund-Freud-
Strasse 25, 53105 Bonn, Germany; Fax: (011) 49
228 287 4758; E-mail: matthias.simon@ukb.
uni-bonn.de
Supported in part by funding from the Deutsche
Krebshilfe (German Glioma Network, 70-3163-Wi
3).
Carsten St uers current address: Neurochirur-
gische Universit atsklinik, Klinikum Rechts der
Isar der TU M unchen, Munich, Germany.
Part of this work forms the doctoral thesis of
Belinda Vilz.
We thank the Division of Neuroradiology of the
University of Bonn Medical Center (head: Prof. H.
Urbach) for some of the magnetic resonance
images and M. von Lehe, MD (Department of
Neurosurgery, University of Bonn Medical Center)
for some epileptologic follow-up data.
Received May 29, 2007; accepted July 30, 2007.
2007 American Cancer Society
DOI 10.1002/cncr.23148
Published online 31 October 2007 in Wiley InterScience (www.interscience.wiley.com).
2799
Most piloA are assigned to the World Health Organi-
zation (WHO) grade 1. Histologically anaplastic
growths are classified as WHO grade 3 tumors.
4
Pilo-
myxoid astrocytomas recently have been distin-
guished from piloA by certain histologic features and
their considerably worse prognosis.
5,6
PiloA appears to be curable by a complete surgi-
cal resection. A 75% progression-free survival (PFS)
rate at 5 years and overall survival rates of 95.8% at
10 years and 82% at 20 years have been reported af-
ter macroscopic total resection.
2,7
The clinical course
after subtotal removal appears to be less predictable.
Long-term stable disease and even partial involution,
but also tumor progression, have been described.
814
Radiation and chemotherapy have been used suc-
cessfully to control progressive and recurrent pediat-
ric tumors and are included, for example, in the
Society of Pediatric Oncology Low-grade Glioma
(SIOP-LGG) 2004 protocol for tumors in children and
adolescents, an international, multicenter, rando-
mized controlled trial for therapy optimization in pe-
diatric patients with low-grade gliomas.
15
The majority of the available data on piloA stem
from pediatric studies. To our knowledge, only very few
studies to date have focused on piloA in adults. Several
authors have reported favorable outcomes and an excel-
lent prognosis after surgery.
3,16,17
In contrast, we
recently observed several patients who had an adverse
clinical course at our institution. This prompted us to
review our experience with piloA in a series of 44 adult
patients. Our results suggest a far higher incidence of
tumor recurrence and malignant progression and, con-
sequently, adverse prognosis than previously described.
Our data also confirm the prognostic impact of a com-
plete tumor resection in these patients.
MATERIALS AND METHODS
Patients and Follow-up
We searched the archives of the Department of Neuro-
surgery and the Institute of Neuropathology of the Uni-
versity of Bonn Medical Center for patients who
underwent surgery between 1995 and 2005 and who
were diagnosed with a piloA. For the purposes of this
investigation, we considered all patients adults if they
were aged > 16 years at the time of their first surgery
(including patients who underwent surgery at outside
institutions), because age < 16 years is a common inclu-
sion criteria of large, prospective pediatric trials such as
the SIOP-LGG1/Gesellschaft fur Padiatrische Onkologie,
Society for Pediatric Oncology (GPOH) 1996 study and
the current SIOP-LGG 2004 study.
15
Relevant clinical in-
formation was obtained through a chart review. Between
1 and 5 magnetic resonance imaging (MRI) studies and
radiologic reports per patient could be made available
for review (including preoperative neuroimaging studies
or radiology notes from42 patients).
Tumors were operated on using standard supraten-
torial approaches, suboccipital craniotomies, and
stereotactic techniques as indicated. The principal
objective of the surgical resections was to achieve a tu-
mor removal as complete as possible without inflicting
a relevant neurologic deficit. In the majority of patients,
the degree of surgical resection was assessed by MRI.
Assessment relied on computed tomography imaging
after 3 surgeries and on the surgeons intraoperative
impression in another 3 of 51 tumor resections.
Routine follow-up included a first postoperative
outpatient visit 3 months after surgery. Thereafter,
patients were seen and MRI scans were obtained ev-
ery 3 to 12 months, depending on the histologic
grade, the degree of tumor resection, and the time
elapsed since the initial surgery. Every attempt was
made to contact patients or their families by tele-
phone if they had not been seen in our outpatient
clinic within the last 6 months. Four patients from
abroad elected to be followed in their home countries
and could not be reached for a recent follow-up.
Neuropathologic Review
Histopathologic findings of the respective patients pri-
mary tumors (n 5 42 patients; 2 patients underwent
surgery at foreign institutions and therefore appropriate
slides could not be made available) and of 9 tumor
recurrences were reviewed at the Institute for Neuro-
pathology/German Brain Tumor Reference Center at
the University of Bonn. PiloA are characterized by vari-
ably textured glioma matrix with Rosenthal fibers, eo-
sinophilic granular bodies, and usually low mitotic
activity. Tumor cells stain positive with antibodies
against glial fibrillary acid protein and microtubule-
associated protein 2 (MAP2). Vascular proliferation and
necrosis are not indicative of a high-grade glioma in
piloA. However, increased proliferation of the glial tu-
mor component is observed in a subset of piloA (ie, Ki-
67-labeling in > 5% of tumor cell nuclei in the absence
of significant CD45-positive lymphocytes or CD68-
expressing macrophages, or activated microglia). We
refer to these tumors as piloA with increased prolifera-
tive activity. PiloA may undergo malignant transforma-
tion. These lesions often feature multiple mitoses per
single high-power field, microvascular proliferation, and
palisading necrosis. In these patients, the designation
anaplastic (malignant) piloA is preferred.
4
Statistical Analysis
A commercially available software program was used
for statistical analysis (SSPS software, version 14.0;
2800 CANCER December 15, 2007 / Volume 110 / Number 12
SSPS Inc., Chicago, Ill). For survival analysis, Kaplan-
Meier estimates and log-rank tests were used. The
level of significance was set at P <.05.
RESULTS
Patient Characteristics, Tumor Sites, Clinical
Presentation, and Epilepsy Outcomes
Our series included 26 women (59%) and 18 men
(41%). The mean age at first surgery was 31 14 years
(range, 1668 years) (Fig. 1). One patient who had a
piloA with increased proliferative activity of the upper
cervical spinal cord suffered from neurofibromatosis
type I. One patient with a secondary, WHO grade 3
anaplastic piloA developed subarachnoid seeding late
in the course of his disease (Fig. 2). Infiltration of the
wall of the carotid artery resulting in intraoperative
rupture of the vessel was observed in 1 patient who
had a left temporomesial piloA. The mean ( the
standard deviation) follow-up for the overall patient
series was 76 59 months (range 1227 months), and
the median follow-up was 55 months.
Tumor sites and the individual clinical presenta-
tions of the patients are shown in Table 1. Note that lo-
bar tumors accounted for 44% (20 of 44 tumors) of all
piloA, in contrast to pediatric neurosurgical experience.
However, the mean age at first surgery did not differ
between patients who had lobar, cerebellar, or brain-
stem tumors. All but 1 of the 20 patients who had lobar
tumors presented with epilepsy, including 4 patients
who were referred specifically to our department for
surgical treatment of their epilepsy and their tumor.
Radiologic Findings
The following MRI patterns were observed: tumors
with large cysts and contrast-enhancing, mural
nodules (21%); solid tumors (50%) with variable con-
trast enhancement ranging from strong (13%) to no
or only minimal (13%); and neoplasms with mixed
cystic and solid areas and variable degrees of con-
trast enhancement (29%). Examples are provided in
Figures 2 and 3. Intratumoral bleeds were observed
in 4 patients. Growth patterns and radiologic findings
did not appear to be correlated with tumor site, his-
tologic grade, or tumor recurrence.
Neuropathologic Review
Tissue samples of 42 primary tumors and 9 recurrent
tumors from 44 patients were available for
FIGURE 1. Age distribution of patients with adult pilocytic astrocytoma.
The mean age was 31 14 years.
FIGURE 2. World Health Organization (WHO) grade 1 pilocytic astrocytoma
(piloA) of the left parietal subcortical white matter in a man aged 62 years
(A). Tumor recurrence was diagnosed 17 months after the first surgery. The
recurrent tumor showed signs of anaplasia and was assigned to WHO grade
3. Despite a repeat resection, radiotherapy, and chemotherapy, the patient
developed subarachnoid seeding (B) and died of his disease 32 months after
the initial diagnosis. (A) From left to right: T1-weighted, axial magnetic reso-
nance imaging (MRI) scans before and after the administration of contrast
medium depict a strongly contrast enhancing, largely solid tumor. The T2-
weighted image shows substantial perifocal edema. The tumor appears to
infiltrate the corpus callosum, which is appreciated best on the coronal scan
(far right, T1-weighted image after administration of contrast medium). (B)
From left to right: A T1-weighted, sagittal MRI scan (after application of con-
trast medium) of the craniocervical junction and cervical spine shows subar-
achnoid spread of the tumor in the fourth ventricle (arrowhead),
intramedullary tumor growth in the upper cervical cord possibly along the
central canal (small arrow), and a leptomeningeal metastasis at the level of
the first thoracic segment (T1) (large arrow). The intramedullary tumor and
edema also are well delineated on the T2-weighted, sagittal image. A T1-
weighted MRI scan (after administration of contrast medium) of the lumbar
spine reveals a second leptomeningeal metastasis at the level of the first
and second lumbar segments (large arrow). Subependymal growth in the
fourth ventricle (arrowhead; far right, top) and intramedullary tumor (small
arrow; far right, bottom) also can be seen on axial, T1-weighted images after
the application of contrast.
Adult Pilocytic Astrocytoma/St uer et al. 2801
histopathologic reevaluation. Thirty-four primary
tumors were assigned to WHO grade 1. Six primary
tumors (14%) were diagnosed as piloA with increased
proliferative activity of the glial tumor component.
Primary WHO grade 3 anaplastic piloA was diag-
nosed in 2 patients (5%) (Fig. 4). No tumor was
reclassified as a pilomyxoid astrocytoma.
Samples of 9 recurrent tumors from 8 patients
were available for review. In 4 patients, the primary
and recurrent tumors were assigned to WHO grade 1.
Five recurrences from 4 patients with WHO grade 1
tumors (n 5 3 patients) and piloA with increased
proliferative activity (n 5 1 patient) had histologic
signs of anaplasia and were assigned to WHO grade
3, ie, a 50% rate (4 of 8 patients) of malignant pro-
gression was observed in patients who underwent
repeat surgery for recurrent tumors. The mean age
( standard deviation) of patients who were diag-
nosed with primary or secondary WHO grade 3
tumors was 40 19 years. For comparison, the mean
age of patients who had benign tumors and no re-
currence during follow-up was 31 16 years (ie,
anaplastic tumors tended to occur in somewhat
older patients). The relatively small sample size pre-
cluded any statistical significance of this finding. His-
tologic grade did not correlate with tumor site.
Surgical Management
Primary surgery (including 4 surgeries performed at
outside institutions) resulted in a complete tumor
resection in 25 patients (57%) and in subtotal re-
moval in 12 patients (27%). Seven patients under-
went a biopsy only (16%). Approximately 66% of
lobar and cerebellar tumors could be resected com-
pletely compared with approximately 33% of brain-
stem piloA. We performed 9 tumor resections and 1
biopsy for residual tumors (2 patients) or recurrent/
progressive tumors. Permanent ventriculoperitoneal
shunts were required in 4 patients (all with posterior
fossa tumors; 21% of posterior fossa tumors) and
insertion of a ventriculoventricular shunt was
required in 1 patient. There was 1 epidural bleed
that required surgical evacuation. There was no sur-
gical mortality.
TABLE 1
Synopsis of Presenting Signs and Symptoms Versus Tumor Sites
Tumor site No. of patients (%) Signs and symptoms, no. of patients
Lobar 20 (44) Headache, 4; seizures, 19; increased intracranial pressure, 2
Temporal/temporomesial 10
Parietal 5
Insular 3
Frontal 1
Occipital 1
Cerebellar 12 (28) Headache, 8; nausea/vomitus/vertigo/nystagmus, 9; ataxia/other cerebellar signs, 9; long tract
signs, 2; diplopia/other cranial nerve signs, 2; hydrocephalus, 3
Hemisphere 9
Vermis 3
Opticochiasmatic 2 (5) Headache, 1; decreased vision, 2
Brainstem 7 (16) Headache, 4; nausea/vomitus/vertigo/nystagmus, 4; ataxia/other cerebellar signs, 2; long tract
signs,4; diplopia/other cranial nerve signs, 4; hydrocephalus, 5
Basal ganglia 2 (5) Headache, 1; hemiparesis, 2; cognitive disturbances, 1
Intramedullary (C0C3) 1 (2) Tetraplegia, 1
C0C3 indicates cervical segments 0 through 3.
FIGURE 3. Magnetic resonance imaging features of temporomesial (A and B) and cerebellar (C and D) pilocytic astrocytoma (piloA) of different World Health
Organization (WHO) grades. (A) From left to right: Axial, sagittal, and coronal T1-weighted images after the application of contrast depict a left temporomesial,
WHO grade 1 piloA with a large tumor cyst and contrast-enhancing mural tumor nodules. (B) From left to right: Axial T1-weighted scans (before and after the
administration of contrast medium) and a coronal T2-weighted image of a temporomesial, WHO grade 1 tumor with solid and cystic areas and variable contrast
enhancement. The T2-weighted image also shows prominent hemosiderin deposits, indicating previous hemorrhages. It is worth noting that this patient experi-
enced a symptomatic intratumorous and intraventricular hemorrhage 4 years earlier but declined surgery at that time. (C) From left to right: Small piloA with
increased proliferative activity of the vermis cerebelli. Axial T1-weighteed (after application of contrast) and T2-weighted images and a coronal fluid-attenuated
inversion recovery image (FLAIR) scan show a small, nonenhancing, nodular tumor. (D) From left to right: Anaplastic, WHO grade 3 piloA of the left cerebellar
hemisphere. Axial T1-weighted (with contrast) and FLAIR images and a coronal T1-weighted scan after application of contrast medium show a tumor with small
cysts, solid areas, and strong contrast enhancement.
"
2802 CANCER December 15, 2007 / Volume 110 / Number 12
FIGURE 3
Adult Pilocytic Astrocytoma/St uer et al. 2803
The mean (standard deviation) preoperative
Karnofsky performance index (KPI) was 80 12
overall, 80 14 at the time of discharge from the
hospital, and 90 32 at the 1-year follow-up visit.
There were 11 new or worsened neurologic deficits
(including 4 new visual field cuts in patients with
temporal and temporomesial tumors and 3 patients
with minimal loss of fine motor skills/dyspraxia).
One patient experienced pure word deafness after
undergoing resection of a tumor of the tectal plate.
18
There were 4 local infectious complications, includ-
ing a superficial wound and 1 shunt infection.
Adjuvant Therapies
All patients who had primary (n 5 2 patients) and
secondary (n 5 4 patients) anaplastic piloA received
FIGURE 4. Histopathologic features of World Health Organization (WHO) grade 1 pilocytic astrocytoma (piloA) (A-D), piloA with increased proliferative activity
(E), and WHO grade 3 anaplastic piloA (F). (A) Hematoxylin and eosin (H & E) stains of typical piloA show a glioma matrix with a variable texture ranging from
solid to loose with occasional Rosenthal fibers (arrowheads) (H & E stain, original magnification 3 200). Eosinophilic granular bodies also are identified fre-
quently (inset: periodic acid-Schiff stain; original magnification, 3 200). Note the incidental mitotic activity (asterisk). (B) Prominent vascular proliferations with
occasional hyalinization and necrosis reflect regressive changes and should not be misinterpreted as signs of a high-grade glioma (H & E stain, original magnifi-
cation 3 200). (C) Immunohistochemical analyses frequently reveal abundant expression of glial fibrillary acid protein (original magnification 3 200). Note the
microcalcifications, which are indicative of a slowly growing process (arrowhead). Slow tumor growth may result in large tumor-associated macrocalcifications.
(D) Proliferation of vascular cells often is picked up by antibodies against Ki67 (MIB-1, original magnification 3 200) and must be distinguished properly from
an increased proliferative activity of the glial cell component. (E) Only increased proliferation of the latter allows for the diagnoses of a piloA with increased pro-
liferative activity (Ki-67/MIB-1, original magnification 3 200). (F) Brisk mitotic and proliferative activity escorted by palisading necrosis (N) is a sign of malig-
nancy. Arrowheads indicate mitoses (Ki-67/MIB-1, original magnification 3 100).
2804 CANCER December 15, 2007 / Volume 110 / Number 12
postoperative, fractionated, involved-field radiother-
apy (at a dose of 5458 Gray). Four patients received
concomitant chemotherapy or chemotherapy for fur-
ther tumor progression (procarbazine, lomustine,
and vincristine [PCV] [n 5 1 patient]; PCV and temo-
zolomide [n 5 1 patient]; and an unknown regimen
[n 5 2 patients]).
No routine adjuvant (radio)therapy was pre-
scribed after a subtotal or macroscopic total resec-
tion of a WHO grade 1 tumor or piloA with increased
proliferative activity (including surgeries for recurrent
tumors). Two patients with WHO grade 1 brainstem
piloA underwent interstitial radiosurgery and fractio-
nated radiotherapy after biopsy. Two patients who
had grade 1 opticochiasmatic gliomas received con-
ventional radiotherapy directly after surgery and for
progressive disease, respectively. Tumor progression
after a biopsy and a subtotal surgical resection in a
patient with a piloA of the basal ganglia was treated
with radiotherapy. Chemotherapy with nimustine
and temozolomide was administered for further tu-
mor growth.
FIGURE 5. Kaplan-Meier estimates of overall survival and progression-free survival (PFS) for adult patients with pilocytic astrocytoma (piloA). (A) Overall sur-
vival (5 years, 87 5%; 10 years, 77 8%) and (B) PFS (5 years, 72 7%; 10 years, 67 8%) for all patients. Note that no recurrences or deaths
occurred after 8 years of uneventful follow-up. (C) World Health Organization (WHO) grade influenced overall survival (P 5.02), and (D) there was a similar
trend for PFS. The extent of resection also was found to be correlated with (E) overall survival and (F) with PFS (P 5.006). Biopsies were performed in very dif-
ferent clinical settings, which may explain the intermediate outcome of this patient group.
Adult Pilocytic Astrocytoma/St uer et al. 2805
Outcome and Prognostic Factors
Tumor recurrence or progression was observed in 13
of 44 patients (30%). Eight of 44 patients (18%) even-
tually died from their disease. The 5-year and 10-
year overall survival rates were 87% 5% and
77% 8%, respectively. The PFS rates at 5 years and
10 years were 72% 7% and 67% 8%, respectively
(Fig. 5A,B). No recurrences or deaths were reported
after 8 years of uneventful follow-up.
WHO tumor grade was correlated with overall
survival (Kaplan-Meier estimates, log-rank test;
P 5.02). A similar trend toward an association of
WHO tumor grade with PFS failed to reach statistical
significance (Fig. 5C,D). The recurrence rate per
patient follow-up year was 4% for 36 patients who
had primary WHO grade 1 tumors, 18% for 6 patients
who had piloA with increased proliferative activity,
and 22% for 6 patients who had primary or second-
ary WHO grade 3 tumors. The extent of resection at
first surgery was found to be strongly predictive of
PFS (Kaplan-Meier estimates, log-rank test; P 5.006),
and a trend was observed for a correlation with over-
all survival (Fig. 5E,F). The recurrence rate per
patient follow-up year was 2% after a macroscopic
total resection and 16% after a subtotal resection.
Biopsies were performed under very different cir-
cumstances (eg, for indolent opticochiasmatic gli-
oma, at outside institutions before resection of a
lobar piloA, and for patients with large diffuse
tumors in a dismal clinical state), which may explain
the intermediate survival outcome of this patient
group. There was no correlation between WHO grade
and extent of resection.
The preoperative KPI and postoperative KPI also
were found to be strongly predictive of PFS (P 5.013
and P 5.001, respectively) and overall survival
(P 5.006 and P 5 <.001, respectively). However,
this was mainly because of the differences in survival
between patients with a KPI 70 versus patients with
a KPI < 70. There was an inverse correlation between
age > 40 years and age < 40 years and overall sur-
vival (Kaplan-Meier estimates, log-rank test; P 5.03),
and there was a trend toward a similar correlation
with PFS. Because radiotherapy and chemotherapy
were administered routinely only for (histologically)
malignant disease, these parameters were not ana-
lyzed as prognostic factors. It is interesting to note
that tumor site did nor correlate with PFS or overall
survival.
DISCUSSION
PiloA frequently are encountered in pediatric neuro-
surgical practice; thus, a considerable number of
reports deal with pediatric piloA.
1923
Conversely, to
our knowledge only scant data are available regard-
ing piloA occurring in adult patients.
3,16,17
In the cur-
rent study, we report on a single institutions
experience with piloA diagnosed in patients aged
> 16 years. Some recent molecular biologic data
have suggested site-specific pathogenic differences.
24
We have analyzed tumors that occurred in different
locations (ie, lobar, cerebellar, optic pathway piloA,
etc) under the same subheading because of their
similar histologic appearance, which may indicate a
common pathobiology. Our results confirm that su-
pratentorial lobar tumors are more frequent after
childhood and tend to grow in the temporal or parie-
tal region.
7,16,17,25,26
Several authors have suggested that piloA in
adults take a relatively benign course.
3,16,17
Brown
et al. reported a prospective study of 20 adults
(mean age of 32 years compared with 31 years in the
current series) with a median follow-up of 10 years.
In their series, 11 patients underwent a complete
surgical resection, 6 patients underwent a subtotal
tumor resection, and 3 patients underwent a biopsy,
including 1 patient who had tumor progression and
1 patient who died of an unknown cause.
16
Bell et al.
described 10 patients who were treated over a 6-year
period (mean age, 49.5 years), including 8 patients
who underwent a complete tumor resection and
2 patients who underwent a biopsy. There were no
tumor-related deaths in that series.
17
In contrast, we observed 13 patients (30%) with
tumor recurrence or progression and 8 deaths (18%)
after a median follow-up of 55 months. The overall
small numbers of patients analyzed (20 patients in
the series by Brown et al.,
16
10 patients in the series
by Bell et al.,
17
and 44 patients in the current series)
may account in part for these differences. Malignant
(anaplastic) piloA may have been diagnosed as ordi-
nary, WHO grade 3 anaplastic astrocytomas by previ-
ous investigators and therefore were excluded from
their respective series. A less than benign clinical
course also has been reported in the larger pediatric
piloA series in a sizable number of patients. Some
studies have reported tumor progression in nearly
50% of patients after incomplete resection of pediat-
ric piloA.
27,28
It is worth noting that we observed increased
proliferative activity and anaplastic histologic fea-
tures in a surprisingly large number of patients.
Overall, 6 of 44 patients (14%) in our current series
were diagnosed with primary or secondary, WHO
grade 3 anaplastic piloA, and the tumors in 6 of 44
patients (14%; including a patient who had anaplas-
tic changes in a recurrent tumor) initially revealed
2806 CANCER December 15, 2007 / Volume 110 / Number 12
increased proliferative activity. Secondary malignant
progression was observed in 4 of 8 patients (50%)
who underwent repeat surgery for recurrent tumors.
Although the overall numbers are rather small, the
percentage is the same as that reported for diffuse,
low-grade gliomas in adults.
29
Histopathologic grade proved to be an important
prognostic factor in the current series. We must
emphasize that these clinical data support the use of
the neuropathologic diagnostic labels (including the
recognition of an intermediate prognostic category of
piloA [ie, with increased proliferative activity]) and
the histopathologic criteria that we used in this se-
ries. An adverse clinical course was observed fre-
quently in patients who had piloA with increased
proliferative activity. Admittedly, analyses of much
larger patient numbers will be required before the
use of an intermediate histologic category for piloA
(eg, atypical piloA WHO grade 2) can be recom-
mended formally. Several immunohistochemical and
molecular genetic markers already have been studied
for a possible association with the patients progno-
sis
3032
and may serve well as adjuncts to more con-
ventional histopathologic techniques in the future.
Analysis of the current series allowed us to reach
some conclusions with respect to the clinical man-
agement of patients with piloA, although the patient
numbers were limited. Routine follow-up beyond
10 years for asymptomatic patients most likely is
unnecessary. No tumor recurrences or deaths were
noted after 8 years of uneventful follow-up in this se-
ries. Patient age and KPI proved to be powerful prog-
nostic factors; however, these parameters cannot be
modified (or at least not to a major degree) through
therapeutic intervention. Most of the impact of the
KPI on prognosis stemmed from the difference
between patients with a KPI 70 and patients with
a KPI < 70. Therefore, so long as the patients low
performance status is not caused by hydrocephalus
or another readily curable condition, and so long as
no major deficit is incurred during surgery, treatment
will not have a major influence on the patients KPI.
In contrast, the degree of surgical resection was
found to be correlated with PFS and overall survival
and was a primary therapeutic variable. A complete
surgical resection most likely is curative for the ma-
jority of WHO grade 1 piloA cases. We have observed
only 1 recurrence in a total of 19 macroscopic total
resections performed for benign piloA. The recurrent
tumor was diagnosed as an anaplastic piloA WHO
grade 3. Even among patients who have piloA with
increased proliferative activity and primary anaplas-
tic tumors, recurrent disease was observed in only
1 of 6 patients, whereas all patients who underwent
a biopsy or a subtotal resection experienced tumor
recurrence or progression. Therefore, a complete
resection should be the neurosurgeons primary
objective whenever possible. The surgical morbidity
encountered in this series appeared to reflect primar-
ily the location of the tumors and compared well
with recent population-based studies.
33
There were 7 recurrences and 4 deaths (all
occurring in patients who had tumors of the brain-
stem and basal ganglia) after subtotal resection of
WHO grade 1 piloA. These data may support the use
of adjuvant therapy (ie, radiotherapy, because to our
knowledge no chemotherapy has been proven effec-
tive) for patients in whom progressive disease most
likely will result in new neurologic deficits and repeat
surgery most likely will not be an appropriate option.
Our experience suggests the need for a similar and
possibly even more aggressive approach for patients
who have piloA with increased proliferative activity.
The usually adverse prognosis for patients who are
diagnosed with WHO grade 3 tumors appears to
clearly justify aggressive adjuvant treatments (ie,
radiotherapy with the possible addition of chemo-
therapy).
In summary, the results of the current study sug-
gest that, in contrast to previous investigations, piloA
in adult patients surprisingly often is not a benign
disease. Recurrence rates are considerable, and tu-
mor-related death is not infrequent. Instances of
piloA with increased proliferative activity and ana-
plastic tumors are not uncommon. However, recur-
rences after 10 years of uneventful follow-up most
likely are very rare, and routine follow-up beyond 10
years for asymptomatic patients most likely is unne-
cessary. Tumor grading provides valuable prognostic
information. The degree of resection is of major im-
portance for the patients further clinical course;
therefore, an aggressive surgical resection should be
attempted whenever possible. Further studies refin-
ing the histologic grading and the indications for ad-
juvant treatment appear to be warranted.
REFERENCES
1. Central Brain Tumor Registry U.S. Statistical Report: Pri-
mary Brain Tumors in the United States, 19982002. Chi-
cago, Ill: Central Brain Tumor Registry of the United States;
2005.
2. Burkhard C, Di Patre PL, Schuler D, et al. A population-
based study of the incidence and survival rates in patients
with pilocytic astrocytoma. J Neurosurg. 2003;98:1170
1174.
3. Afra D, Muller W, Slowik F, Firsching R. Supratentorial lo-
bar pilocytic astrocytomas: report of 45 operated cases,
including 9 recurrences. Acta Neurochir (Wien). 1986;81:
9093.
Adult Pilocytic Astrocytoma/St uer et al. 2807
4. Burger PC, Scheithauer BW, Paulus W. Pilocytic astrocyto-
mas. In: Kleihues P, Cavenee W, editors. Pathology and
Genetics of Tumours of the Nervous System. Lyon, France:
IARC Press; 2000:4551.
5. Komotar RJ, Burger PC, Carson BS, et al. Pilocytic and pilo-
myxoid hypothalamic/chiasmatic astrocytomas. Neurosur-
gery. 2004;54:7279.
6. Tihan T, Fisher PG, Kepner JL, et al. Pediatric astrocytomas
with monomorphous pilomyxoid features and a less favor-
able outcome. J Neuropathol Exp Neurol. 1999;58:10611068.
7. Forsyth PA, Shaw EG, Scheithauer BW, OFallon JR, Layton
DD Jr, Katzmann JA. Supratentorial pilocytic astrocytomas.
A clinicopathologic, prognostic, and flow cytometric study
of 51 patients. Cancer. 1993;72:13351342.
8. van der Wal EJ, Azzarelli B, Edwards-Brown M. Malignant
transformation of a chiasmatic pilocytic astrocytoma in a
patient with diencephalic syndrome. Pediatr Radiol. 2003;
33:207210.
9. Krieger MD, Gonzalez-Gomez I, Levy ML, McComb JG. Re-
currence patterns and anaplastic change in a long-termstudy
of pilocytic astrocytomas. Pediatr Neurosurg. 1997;27:111.
10. Dirven CM, Koudstaal J, Mooij JA, Molenaar WM. AgNOR
staining may reflect the growth potential of pilocytic astro-
cytomas. Childs Nerv Syst. 1999;15:384388.
11. Balkhoyor KB, Bernstein M. Involution of diencephalic
pilocytic astrocytoma after partial resection. Report of two
cases in adults. J Neurosurg. 2000;93:484486.
12. Alkhani AM, Boop FA, Rutka JT. Involution of enhancing
intrinsic tectal tumors after endoscopic third ventriculost-
omy. Report of two cases. J Neurosurg. 1999;91:863866.
13. Hadjipanayis CG, Kondziolka D, Gardner P, et al. Stereotac-
tic radiosurgery for pilocytic astrocytomas when multimo-
dal therapy is necessary. J Neurosurg. 2002;97:5664.
14. Tomlinson FH, Scheithauer BW, Hayostek CJ, et al. The sig-
nificance of atypia and histologic malignancy in pilocytic
astrocytoma of the cerebellum: a clinicopathologic and
flow cytometric study. J Child Neurol. 1994;9:301310.
15. Gnekow AK, Kortmann RD, Pietsch T, Emser A. Low grade
chiasmatic-hypothalamic glioma: carboplatin and vincris-
tine chemotherapy effectively defers radiotherapy within a
comprehensive treatment strategyreport from the Multi-
center Treatment Study for Children and Adolescents with
Low Grade GliomaHIT-LGG 1996of the Society of Pedi-
atric Oncology and Hematology (GPOH). Klin Padiatr.
2004;216:331342.
16. Brown PD, Buckner JC, OFallon JR, et al. Adult patients
with supratentorial pilocytic astrocytomas: a prospective
multicenter clinical trial. Int J Radiat Oncol Biol Phys.
2004;58:11531160.
17. Bell D, Chitnavis BP, Al-Sarraj S, Connor S, Sharr MM, Gullan
RW. Pilocytic astrocytoma of the adultclinical features,
radiological features and management. Br J Neurosurg.
2004;18:613616.
18. Meyer B, Kral T, Zentner J. Pure word deafness after resec-
tion of a tectal plate glioma with preservation of wave V of
brain stem auditory evoked potentials. J Neurol Neurosurg
Psychiatry. 1996;61:423424.
19. Fiks T, Jesionek-Kupnicka D, Zakrzewski K, Polis L, Liberski
PP. Clinicopathological analysis of pilocytic astrocytomas
and gangliogliomas in children. Folia Neuropathol. 1999;
37:152156.
20. Gunny RS, Hayward RD, Phipps KP, Harding BN, Saunders
DE. Spontaneous regression of residual low-grade cerebel-
lar pilocytic astrocytomas in children. Pediatr Radiol. 2005;
35:10861091.
21. Hojer C, Hildebrandt G, Lanfermann H, Schroder R, Haupt
WF. Pilocytic astrocytomas of the posterior fossa. A follow-
up study in 33 patients. Acta Neurochir (Wien). 1994;129:
131139.
22. Palma L, Guidetti B. Cystic pilocytic astrocytomas of the
cerebral hemispheres. Surgical experience with 51 cases
and long-term results. J Neurosurg. 1985;62:811815.
23. Vinchon M, Assaker R, Soto-Ares G, Ruchoux MM, Dhel-
lemmes P. Cerebellar pilocytic astrocytomas in children.
Report of 72 cases. Neurochirurgie. 2001;47:8391.
24. Sharma MK, Mansur DB, Reifenberger G, et al. Distinct
genetic signatures among pilocytic astrocytomas relate to
their brain region origin. Cancer Res. 2007;67:890900.
25. Hayostek CJ, Shaw EG, Scheithauer B, et al. Astrocytomas
of the cerebellum. A comparative clinicopathologic study
of pilocytic and diffuse astrocytomas. Cancer. 1993;72:856
869.
26. Katsetos CD, Krishna L. Lobar pilocytic astrocytomas of
the cerebral hemispheres: I. Diagnosis and nosology. Clin
Neuropathol. 1994;13:295305.
27. Sutton LN, Cnaan A, Klatt L, et al. Postoperative surveil-
lance imaging in children with cerebellar astrocytomas.
J Neurosurg. 1996;84:721725.
28. Dirven CM, Mooij JJ, Molenaar WM. Cerebellar pilocytic
astrocytoma: a treatment protocol based upon analysis of
73 cases and a review of the literature. Childs Nerv Syst.
1997;13:1723.
29. Schmidt MH, Berger MS, Lamborn KR, et al. Repeated
operations for infiltrative low-grade gliomas without inter-
vening therapy. J Neurosurg. 2003;98:11651169.
30. Bowers DC, Gargan L, Kapur P, et al. Study of the MIB-1
labeling index as a predictor of tumor progression in pilo-
cytic astrocytomas in children and adolescents. J Clin
Oncol. 2003;21:29682973.
31. Sharma MK, Watson MA, Lyman M, et al. Matrilin-2
expression distinguishes clinically relevant subsets of pilo-
cytic astrocytoma. Neurology. 2006;66:127130.
32. Wong KK, Chang YM, Tsang YT, et al. Expression analysis
of juvenile pilocytic astrocytomas by oligonucleotide
microarray reveals two potential subgroups. Cancer Res.
2005;65:7684.
33. Chang SM, Parney IF, McDermott M, et al. Perioperative
complications and neurological outcomes of first and sec-
ond craniotomies among patients enrolled in the Glioma
Outcome Project. J Neurosurg. 2003;98:11751181.
2808 CANCER December 15, 2007 / Volume 110 / Number 12