1583 / 1708
heterogeneous mix of tumour parameters and prognostic
factors, including extent of resection and grade.
DOSE-RESPONSE RELATIONSHIP
When interpreting retrospective data it is difficult to
rule out selection for lower RT dose based on adverse
prognostic features e.g. age and tumour size, with lower
dose RT being employed for younger patients and those
with larger target volumes. It is also difficult in some
studies to analyse the impact of dose on response when a
dose-fractionation regimen has been uniformly applied, or
when there is a high local relapse rate.
This review includes 11 series of patients reported since
the early 1990s [4–14] and comprises 526 patients,
involving treatment over time periods between 14 and
38 years. In these series the mean number of patients
treated per institution per year was 1.8. This reflects the
low incidence of EP. Overall 5-year survival varied from
40 to 79%. Seven of the 11 series demonstrated that
the outcome was related to grade [4,7,9,11–14] and 7
series demonstrated the outcome related to the extent of
surgical resection [4–7,9,13,14]. For incompletely
resected tumours 5-year OS varied from 22 to 64.1%,
whereas for completely resected tumours this was from
61 to 80%.
DOSE-RESPONSE DATA
Table I shows outcome data from 11 series reported
since the early 1990s, in which information on dose-
response has been given. Although data is inconclusive,
there is some evidence of a dose-response effect, either for
>
45 Gy versus
<
45 Gy or
>
50 Gy versus
<
50 Gy.
ROLE OF HYPERFRACTIONATED
RADIOTHERAPY (HFRT)
There is no data on the radiobiology of EP. Thus
consideration of the potential benefit for HFRT relies on
the empirical analysis of series of patients treated by
HFRT. In the PediatricOncologyGroup (POG) 9132 study,
in 15 patients who had incomplete resection a HFRT dose
of 69.6 Gy given in 58 twice daily fractions resulted in a
3-year EFS of 52% [15]. This compared favourably with a
similar group treated in an earlier study with conventional
fractionation, who had a 5-year EFS of 27%. Several other
studies have explored the role of HFRT in ependymoma
and results are awaited.
ROLE OF DURATION OF RT
In one study [16] the impact on outcome of prolonga-
tion of the duration of RT has been examined. In this
study in patients for whom the RT treatment duration was
<
50 days, the 5-year OS was 85.5% compared with 45.5%
for 50 days or greater (
P
¼
0.01). The 5-year local control
rate for patients whose treatment duration was
<
50 days
was 70.6% compared with 45.5% for 50 days or greater
(
P
¼
0.05). In this type of analysis it is important to rule out
an impact of other prognostic factors. However, this study
is of interest and for future analyses of outcome of RT for
TABLE I. Influence of Radiotherapy (RT) Dose on Outcome
Author [reference]
Institution
Dates
No. of
patients
RT dose
<
45 Gy
45 Gy
<
50 Gy
50 Gy
Goldwein et al., 1990 [6]
Philadelphia
1970–1988
51
18% 5Y OS 51% 5Y OS
0% 5Y PFS 32% 5Y PFS
Vanuytsel et al., 1992 [7]
Royal Marsden 1952–1988
93
53% 5Y OS
(
<
¼
50 Gy)
55% 5Y OS
(
>
50 Gy)
Chiu et al., 1992 [8]
MD Anderson 1955–1986
25
33% 5Y OS 58% 5Y OS
Rousseau et al., 1994 [4]
IGR, Paris
1975–1989
65
51% 5Y OS 69% 5Y OS
Carrie et al., 1995 [9]
Lyon
1974–1993
37
6/12 (50%)
relapsed
(
<
50 Gy)
6/16 (37.5%)
relapsed
(
>
50 Gy)
Pollack et al., 1995 [5]
Pittsburgh
1975–1993
37
Routinely applied to a dose
>
¼
50 Gy)
Stuben et al., 1997 [10]
Essen
1963–1995
41
36% 5Y PFS
( 45 Gy)
45% 5Y PFS
(
>
45 Gy)
Schild et al., 1998 [11]
Mayo clinic
1963–1994
45
‘No dose response’
Mc Laughlin et al., 1998 [12] Gainesville
1966–1989
32
N/A (high loc rec rate)
Paulino et al., 2002 [13]
Iowa
1965–1997
52
GTR
þ
>
45
Gy LC 76.9%
Oya et al., 2002 [14]
Kyoto
1961–1999
48
Uniformly applied, modified
according to tumour size,
no association (
<
55 Gy vs.
>
¼
55 Gy)
PFS, progression-free survival; OS, overall survival; GTR, gross total resection; LC, local control.
458
Taylor