1450 / 1708
Articles
http://oncology.thelancet.comVol 8 August 2007
699
proportion of the cumulative dose of the chemotherapy
regimen actually received relative to that defined in the
protocol (table 1). These proportions were then averaged
over all patients receiving the regimen. The relative dose
intensity (RDIChemo) adjusted the individual patient
SRDChemo by the ratio of the time the regimen took to
be given, divided by the corresponding protocol-defined
time. These are then averaged for all patients to give the
regimen RDIChemo.
After surgery, patients were classified into four groups:
those who (A) continue to be alive without progression
and without radiotherapy; (B) have disease progression,
or die without progression first being documented, but
who do not receive radiotherapy; (C) receive elective
radiotherapy without documentation of progression; or
(D) receive radiotherapy after progression.
Of the corresponding survival times in these groups,
those in A were censored for all events, whereas those for
B, C, and D represent times to competing events.
A competing event is one which, if it occurs, prevents
another event from being observed. In the presence of
competing risks, the cumulative incidence method
19
estimates the cumulative probability for each cause of
failure for B, C, or D in the presence of all risks to the
patients concerned. If the radiotherapy-free survival
(RADFS) rate is calculated, which adds all the event types
together, then (1–RADFS) is the sum of the three separate
cumulative incidences.
Event-free survival was defined as the time from date of
surgery to the date of the first event—ie, a recurrence or
death. When death followed recurrence, the event was the
recurrence. Overall survival was calculated as the time
from the date of surgery to death from any cause. Patients
still alive were censored at the date last seen. Survival
probabilities were calculated using the Kaplan-Meier
method. The hazard ratios and 95% CIs for comparing
metastatic and non-metastatic patients were estimated
using the Cox proportional hazards model. The potential
influences of age, sex, histology, tumour site, and extent
of resection on the hazard ratio were also investigated
with the Cox model.
20
The effect of the RDIChemo was assessed by calculating
the residual overall survival time from 1-year after the
start of chemotherapy.
21
The protocol of this trial is currently under review to
ensure it complies with good clinical practice, and the
revised protocol will be registered in a publicly accessible
database upon completion.
Role of the funding source
The sponsor of the study had no role in the study design,
conduct of the study, data collection, data management,
data analysis, data interpretation, preparation of the
report, review of the report, or approval of the report.
RGG, SAW, CLW, KR, JI, TC, WKC, NT, DWE, and DM
had full access to all of the raw data and RGG had final
responsibility to submit for publication.
Results
89 patients with a diagnosis of intracranial ependymoma
from 21 participating centres were registered to the
study between Dec 1, 1992, and April 31, 2003. Of these,
80 (90%) presented without metastatic disease and nine
(10%) with imaging evidence of primary dissemination.
Of the children with metastatic disease, three had
nodular spinal leptomeningeal dissemination (M3),
three had cerebral nodules (M2), and three had cranial
and spinal disease (M3) present on pre-operative MRI
images. Table 2 shows patients’ characteristics. The
children were predominantly male (58 [66%]). 76
(85%) had infratentorial tumours. Pre-chemotherapy
assessment was done consistent with a Lansky play scale
of 70–90%.
Number of relapses
(n=59)
5-year overall survival from
the date of relapse (95% CI)
Relapse site
Local relapse site
47
26 (13-41)
Local and metastatic relapse
6
Too early to tell
Metastatic relapse only
4
Too early to tell
Unknown
1
0
Perioperative death
1
0
WHO grade
WHO primary tumour grade II
37
30 (13-48)
WHO primary tumour grade III
22
25 (8-47)
Surgery
No surgery after relapse
28
24 (8-44)
Surgery after relapse
30
31 (13-51)
Perioperative death
1
0
Radiotherapy
No radiotherapy after relapse
18
Too early to tell
Radiotherapy without surgery
17
20 (5-43)
Radiotherapy with surgery
23
32 (11-55)
Perioperative death
1
0
Table 4
: Outcome in patients with a first relapse after primary treatment
0·2
0·4
0·6
0
2
4
6
8
10
··
Numbers at risk
of competing
events
79
51
33
20
12
9
Elective radiotherapy but no progression (C)
Radiotherapy after progression (D)
No radiotherapy despite progression (B)
0·0
Event-free survival (years)
Event-free survival (%)
12
Figure 2:
Competing risks analysis for patients with non-metastatic ependymoma