1598 / 1708
reviewed in 17 children. In 31 children, the tumor site was
infratentorial. No patient with spinal ependymoma was
included in the study.
In 23 patients,CSF cytologic studies for evaluation of
leptomeningeal dissemination were available at presen-
tation. Three children had positive CSF cytological findings
(13%); in 11 children, CSF-samples were not available, but all
children underwent craniospinal imaging. Solid metastases
to the CNS were found in two children (5.9%). Twenty-nine
children (85.3%) did not present with any metastases
(
Table 1).
Treatment
Surgery
All children underwent surgery. Extent of resection was
assessed by postoperative CT/MRI, and was considered to be
macroscopically complete in 18 children.
Chemotherapy
In HIT-SKK 87 trial, 15 children and, in HIT-SKK 92, 19
children were treated. All children received adjuvant
chemotherapy.
Radiotherapy
Thirteen children did not receive any radiotherapy
(38.2%). Ten children were irradiated at the primary tumor
site only. Eleven children received craniospinal irradiation
with an additional boost to the tumor. The median
cumulative total dose to the tumor was 54.0 Gy (range,
20.8–56.4 Gy). The median total dose to the neuraxis was
35.2 Gy (range, 24.0–39.6 Gy). The median dose per fraction
was 1.8 Gy (range, 1.4–2.2 Gy). In 12 children, radiotherapy
was given immediately after completion of chemotherapy
without any sign of recurrence or progression of disease. In
nine children, radiotherapy was delayed and, administered
only in case of recurrence or progression as salvage therapy.
Median time interval between surgery and start of
irradiation was 11 months (range, 4–34 months).
Survival
Follow-up for all patients ranged from 7 to 146 months. In
survivors, the median time of follow-up was 76.5 months
(range, 53–146). For all patients, the 3-year estimated
overall survival rate and 3-year PFS rate were 55.9%
(confidence interval (
Z
CI) 39.2–72.6) and 27.3% (CI 12.1–
42.5), respectively (
Fig. 3). For histologically reviewed
children (
n
Z
17), PFS rates were estimated separately, but
no difference could be detected (3-year overall survival and
PFS of 58.8% (CI 35.4–82.2) and 25% (CI 3.8–46.2),
respectively). Twenty-one children died of recurrent
disease. One chemotherapy-related death occurred. There
were no other causes of death. For patients who failed,
median time to progression was 8 months and median time to
death was 29.5 months (range, 7–95 months).
Patterns of failure
At last follow-up, nine children were free of disease, and
25 children showed progression (73.5%). Nineteen children
(76.0%) failed at the tumor site only. Six children (24%)
developed dissemination within the CNS (four of them
intracranial, one spinal, and one both intracranial and
spinal); all of those six patients had local failures also.
Late effects
At last follow-up, in five survivors information about late
toxicity was available. In two children, growth retardation,
pituitary insufficiency and need for hormonal replacement
were reported. In one other child, retardation in language
and mental development was described. One more child had
motor deficits in the left upper arm, and one child suffered
from cerebellar ataxia.
Table 1
Characteristics of the 34 children with ependymoma
Characteristics
No. of patients
(%)
Age (months)
Median
20.5
Range
1.0–33.0
Sex
Male
18
52.9
Female
16
47.1
Site
Infratentorial
31
91.2
Supratentorial
3
8.8
Spinal
0
0
Metastases
M0/M
x
29
85.3
M1
3
8.8
M2/3
2
5.9
Resection
Complete
18
52.9
Incomplete
16
47.1
Chemotherapy
SKK 87
15
44.1
SKK 92
19
55.9
Radiotherapy
CSI
C
boost
11
32.4
Local field
10
29.4
None
13
38.2
Radiotherapy
Preventive
12
35.3
Salvage
9
26.5
None
13
38.2
0
20
40
60
80
Survial (months)
0
0.2
0.4
0.6
0.8
1
Probability
3-yrs.-pfs = 27.3 %
3-yrs.-os = 55.9 %
Overall survival
Progression free survival
Fig. 3. Kaplan–Meier plots of overall survival and PFS.
B. Timmermann et al. / Radiotherapy and Oncology 77 (2005) 278–285
281