Leukemia (2005) 19, 2030–2042
& 2005 Nature Publishing Group All rights reserved 0887-6924/05 $30.00
www.nature.com/leu
Treatment strategies and long-term results in paediatric patients treated in four
consecutive AML-BFM trials
U Creutzig1, M Zimmermann2, J Ritter1, D Reinhardt1, J Hermann3, G Henze4, H Jürgens1, H Kabisch5, A Reiter6, H Riehm2,
H Gadner7 and G Schellong1, for the AML-BFM Study Group
1
Department of Haematology, Oncology, University Children’s Hospital, Münster, Germany; 2Department of Haematology,
Oncology, University Children’s Hospital, Hannover, Germany; 3Department of Haematology, Oncology, University Children’s
Hospital, Jena, Germany; 4Department of Haematology, Oncology, University Children’s Hospital, Berlin, Germany;
5
Department of Haematology, Oncology, University Children’s Hospital, Hamburg, Germany; 6Department of Haematology,
Oncology, University Children’s Hospital, Giessen, Germany; and 7Children’s Cancer Research Institute and St Anna Kinderspital,
Vienna, Austria
A total of 1111 children with acute myeloblastic leukaemia
(AML) were treated in four consecutive Berlin–Frankfurt–
Münster (BFM) studies from 1978 to 1998. The first cooperative
trial AML-BFM 78 established intensive chemotherapy with
seven drugs, CNS irradiation and 2-year maintenance, achieving a long-term survival (overall survival (OS)) of 40%. Induction
intensification in AML-BFM 83 resulted in significant improvement of disease-free survival (DFS). The risk of haemorrhage,
especially in children with hyperleukocytosis, proved the high
relevance of supportive care. In AML-BFM 87, the benefit of
CNS irradiation in preventing CNS/systemic relapses was
demonstrated. In AML-BFM 93, the introduction of idarubicin
during first induction followed by intensification with HAM
increased the 5-year EFS, DFS and OS to 5072, 6173 and
5772%, respectively. Stem cell transplantation (SCT), as
applied in high-risk patients with a matched related donor, did
not significantly improve the outcome compared to chemotherapy alone. In spite of treatment intensification, the therapyrelated death rate decreased from trial to trial, mainly during
induction. The future aim is to reduce long-term sequelae,
especially cardiotoxicity, by administration of less cardiotoxic
drugs, and toxicity of SCT by risk-adapted indications. The
AML-BFM studies performed in three European countries with
470 cooperating centres have significantly improved the
outcome in AML children; nevertheless, increasing experience
with these intensive treatment regimens is of fundamental
importance to reduce fatal complications.
Leukemia (2005) 19, 2030–2042. doi:10.1038/sj.leu.2403920
Keywords: AML therapy; children; long-term results
Introduction
The Berlin–Frankfurt–Münster (BFM) group started their cooperation in 1976 for paediatric acute lymphoblastic leukaemia
(ALL) and extended the cooperation in 1978 to acute
myeloblastic leukaemia (AML). Since that time, four consecutive
AML trials have been performed within the AML-BFM study
group.
Table 1 presents some details of the four trials, including the
number of centres involved, their average number of patients
and the number of patients per trial.
Correspondence: Professor U Creutzig, Klinik und Poliklinik für Kinderheilkunde, Pädiatrische Hämatologie/Onkologie, Albert-Schweitzer-Str.
33, D-48129 Münster, Germany; Fax: þ 49 251 83 56489;
E-mail: [email protected]
This paper is dedicated to Christa Lausch, our valuable coworker in
the AML-BFM Trial Centre from 1982-2003.
Received 14 December 2004; accepted 10 May 2005
Background and treatment strategy of the AML-BFM
trials
The treatment regimens of the four trials are shown schematically in Figure 1, and the treatment elements of the AML-BFM
protocols including drug dosages are summarised in Table 2.
Each trial was based on the experience of the previous studies
and of other AML trials in children or adults. The intensity of
treatment was increased from trial to trial, with the exception of
AML-BFM 87.
Before 1978, children with AML had been treated in Germany
either according to ALL strategies or adult AML regimens, and
the outcome was accordingly poor. Between 7/1973 and
10/1978, 23 children with AML were treated at the University
Children’s Hospitals in Münster and Berlin with a similar drug
combination as used in the ‘West-Berlin Pilot Study’ for children
with ALL,1 however, including more AML specific drugs like
cytarabine and anthracyclines, according to a schedule which
was also introduced by Riehm et al. Based on the fact that the
outcome of this group was better than expected, with a 4-year
survival probability of 44%,2 the first cooperative AML-BFM trial
was initiated in 1978.
AML-BFM 78: The AML-BFM 78 treatment protocol used a
seven-drug, 8-week induction/consolidation (see Consolidation,
Table 2), which was similar to the initial ‘West-Berlin’ ALL
therapy regimen,1 combined in its second phase with preventive CNS irradiation (standard dose ¼ age-dependent: 0–o1
year ¼ no irradiation, 1–o3 years ¼ 15 Gy, X3 years ¼ 18 Gy)
and followed by 2 years of maintenance with daily thioguanine
40 mg/m2 orally and cytarabine 40 mg/m2 s.c. 4 days monthly
for a total duration of 24 months. Doxorubicin was given every
8 weeks during the first year of maintenance. The probability
of event-free survival (EFS) at 5 years was 3874%,3 which was
a good result at that time. Therefore, this therapy strategy was
used as backbone for the following trials.
AML-BFM 83: At the beginning of the 80s, AML studies for
adults had demonstrated that cytarabine administration for 7
days was superior to a 5-day treatment with the same agent
(Cancer and Leukemia Group B (CALGB) Study in adults4), and
that high remission rates could be achieved with thioguanine,
cytarabine, daunorubicin (TAD) induction in the AMLCG
Study.5 Consequently, in AML-BFM 83, therapy started with
an 8-day induction cycle consisting of cytarabine and daunorubicin (3 60 mg/m2) as in the adult AMLCG study, however,
as thioguanine was already included as daily medication
during maintenance, it was replaced by etoposide. This ADE
Improved survival in paediatric AML, BFM strategy
U Creutzig et al
2031
Table 1
Patient accrual and follow-up
Trial
AML-BFM
AML-BFM
AML-BFM
AML-BFM
78
83
87
93
Accrual of
patients: time
period
Total number of
patients (n)
Centres (n)
12/78–09/82
12/82–09/86
12/86–09/92
1/93–06/98
151
182
307
471
30
30
37
68
Figure 1
Flow diagram of trials AML-BFM 78–93. RT ¼ CNS
radiotherapy.
( ¼ cytarabine, daunorubicin and etoposide) induction was
followed by an 8-week induction/consolidation therapy as
administered in AML-BFM 78. After this treatment element,
maintenance was given as in the previous trial.
Compared to the previous trial, AML-BFM 83 resulted in a
major improvement in long-term results, with a 5-year EFS of
4774%.6
AML-BFM 87: The main question of this trial was the
importance of CNS irradiation for CNS prophylaxis. Irradiation
was not generally included in therapy protocols for AML in
adults and children. Most investigators agreed that cranial
irradiation could prevent CNS relapses, but its effect on overall
remission duration was unknown.7 Results of ALL trials in
children8 and of the AML study 72–75 reported by Dahl et al
indicated that cranial irradiation had an impact on the number
of CNS relapses, but not on overall survival (OS).9 Most AML
studies in children and adults used intrathecal methotrexate or
cytarabine or a combination of these agents with hydrocortisone
as CNS-directed therapy. In the AML-BFM studies, CNS
irradiation was generally included. The AML-BFM 87 study
tested prospectively if CNS irradiation could be replaced by late
intensification therapy with high-dose cytarabine and etoposide.
The premature decision to stop irradiation for all children,
however, resulted in a significant increase of CNS and systemic
relapses. Detailed analyses of the study cohort and the
randomised and nonrandomised patients clearly demonstrated
the need of prophylactic irradiation within the AML-BFM
setting. As this conclusion was based on nonrandomised
patients too, these findings still have to be confirmed by other
paediatric or adult AML trials.
ADE induction therapy remained the same as in the previous
study. Duration of consolidation was reduced to 6 weeks,
because considerable variations with many interruptions due to
aplasia or toxicity had occurred in AML-BFM 83 during the
Number of
patients/centre
(median/range)
4
5
6
5
(1–24)
(1–22)
(1–33)
(1–41)
Number of
patients/year
41
48
53
86
Follow-up
(median, range,
years)
15
11
9
5
(2–19)
(4–17)
(3–14)
(1–9)
second phase of the 8-week consolidation.10 After consolidation, two cycles of late intensification with high-dose cytarabine
(3 g/m2) and etoposide (HAE) were added. CNS irradiation was
scheduled at the beginning of maintenance, and patients
without CNS involvement were randomly assigned to receive
an irradiation of 18 Gy or no irradiation. Duration of maintenance was reduced to approximately 12 months (total therapy
duration: 18 months), and doxorubicin was removed from
maintenance therapy since AML-BFM 87.
Allogeneic stem cell transplantation (SCT) in first CR (after the
second treatment course) was recommended in children of the
high-risk group only if a sibling donor was available.
Based on the results of studies AML-BFM 83 and 87, two
different risk groups (standard and high) could be defined, see
‘risk group definition’ page 8.
AML-BFM 93: Patients were randomly assigned to receive
the 8-day induction with either daunorubicin (ADE) or
idarubicin (AIE) in order to compare the efficacy and toxicity
of these two drugs (Figure 2). The rationale for using idarubicin
was that in vitro and preclinical studies had suggested a possibly
higher clinical benefit of idarubicin, showing a faster cellular
uptake, increased retention and lower susceptibility to multidrug resistance.12,13 Additional advantages seemed to arise from
the long plasma half-life of 54 h of its main metabolite
idarubicinol, which also showed antileukaemic activity in the
cerebral spinal fluid.14 Furthermore, idarubicin showed less
cardiotoxic side-effects in several animal models12 – an
important finding regarding the risk of an increased anthracycline-induced cardiomyopathy in children.15
After induction, patients were treated depending on risk
groups (see Figure 3).11 High-risk patients were randomised to
receive either high-dose cytarabine and mitoxantrone (HAM)
followed by consolidation ( ¼ arm early HAM) or 6-week
consolidation followed by HAM ( ¼ arm late HAM). Standardrisk patients received consolidation only, without HAM.
Subsequently, all patients were treated with one intensification
cycle of HAE instead of two HAE cycles after consolidation as in
AML-BFM 87. Recommendations for SCT in first CR were the
same as in the previous trial.
The rationale for this schedule with HAM was the intention to
improve even more the unsatisfactory outcome in children with
high-risk AML.
Another aim of AML-BFM 93 was to evaluate whether placing
HAM as the first or second post-induction treatment cycle had a
different impact and prognosis. This question was based on the
experience in adults with AML showing that the dose intensity
during the first two treatment courses was of prognostic
significance.16 In view of reports showing that this intensification regimen was associated with increased toxicity and
consequently required rigorous supportive care, HAM was
restricted to high-risk patients.
Leukemia
2032
Leukemia
Table 2
Treatment elements in the four trials AML-BFM 78, 83, 87 and 93
AML-BFM 78
AML-BFM 83
AML-BFM 87
ADE induction
introduced in trial 83
Not given
ADE: cytarabine and etoposide as before ADE: cytarabine, daunorubicina
and etoposide as before
daunorubicin 30 mg/m2 30 min infusion
every 12 h on days 3–5
or with idarubicin (AIE): idarubicin
12 mg/m2 30 min infusiona every 24 h,
days 3–5 (instead of daunorubicin),
cytarabine and etoposide as before
Consolidation
introduced in trial 78
8-week induction/consolidation:
Phase 1: 6-thioguanine 60 mg/m2/day,
days 1–43 orally;
prednisone 60 mg/m2/day, days
1–28 orally;
vincristine 1.5 mg/m2/day, days 1,
8, 15, 22;
doxorubicin 25 mg/m2/day, days 1,
8, 15, 22;
cytarabine 75 mg/m2/day, days 3–6,
10–13, 17–20, 24–27,
Phase 2: cytarabine 75 mg/m2/day, days
31–34, 38–41, 45–48, 52–55;
doxorubicin 25 mg/m2/day, days 36, 50;
cyclophosphamide 500 mg/m2/day, days
29, 43, 57
Not given
ADE:
cytarabine 100 mg/m2/day continuous
infusion on days 1 and 2 followed by
30 min infusion every 12 h on days 3–8,
daunorubicin 60 mg/m2 30 min infusion
per day on days 3, 4, 5
etoposide 150 mg/m2 60 min infusion
on days 6–8
As in trial 78, except:
Not given
Not given
Not given
HAE
introduced in trial 87
HAM
introduced in trial 93
Doxorubicin 30 mg/m2/day, days 1,
8, 15, 22;
Phase 2: no doxorubicin
cyclophosphamide 500 mg/m2/day,
days 29, 43
18 Gy ( ¼ standard dose in children
As in trial 78
X3 years),
o1 year ¼ 12 Gy, 1–2 years ¼ 15 Gy
(during the second phase of consolidation)
Intrathecal therapy
Methotrexate 12.5 mg days 31; 38, 45; 52 Cytarabine days 31; 38, 45; 52 (during
(during consolidation)
consolidation) adjusted to age, children
o1 year ¼ 20 mg, 1–2 years ¼ 26 mg,
2–3 years 34 mg, X3 years 40 mg
2
As in trial 78
Maintenance
Daily thioguanine 40 mg/m orally,
cytarabine 40 mg/m2 s.c. 4 days
introduced in trial 78
monthly for a total duration of 24 months;
doxorubicin 25 mg/m2 every 8 weeks
during the first year
Stem cell transplantation Not given
Not recommended
CNS irradiation
introduced in trial 78
a
Since October 1996, 4 h infusion of daunorubicin and idarubicin.
6-week consolidation:
6-thioguanine 60 mg/m2/day, days
1–43 orally;
prednisone 40 mg/m2/day, days
1–28 orally;
vincristine 1.5 mg/m2/day, days 1,
8, 15, 22;
doxorubicin 30 mg/m2/day, days 1,
8, 15, 22;
cytarabine 75 mg/m2/day, days 3–6,
10–13, 17–20, 24–27, 31–34, 38–41;
cyclophosphamide 500 mg/m2/day,
days 29, 43
AML-BFM 93
As in trial 87
Two cycles of high-dose cytarabine 3 g/m2 One cycle HAE
3-h infusion every 12 h for 3 days and
etopside 125 mg/m2 on days 2–5
Not given
High-dose cytarabine 3 g/m2 3-h infusion
every 12 h for 3 days, mitoxantrone
10 mg/m2 on days 4 and 5
High-risk patients only
As in trial 78
As in trial 78 except,
compared with no irradiation
o1 year ¼ no irradiation (at the beginning
(at the beginning of maintenance)
of maintenance)
Cytarabine days 1, 15, 29, 43 (during
consolidation and concomitantly with
CNS irradiation days 1, 8, 15, 21)
adjusted to age
Without doxorubicin total duration 18
months (maintenance 12 months)
As in trial 87
in addition at day 1 during induction
Matched related in high-risk patients
As in trial 87
As in trial 87
Improved survival in paediatric AML, BFM strategy
U Creutzig et al
New element
Improved survival in paediatric AML, BFM strategy
U Creutzig et al
2033
SR Consolidation
ADE
Cumulative doses
HAM
R
AIE
Consolidation
HAE
Maintenance1 year
HR
Consolidation
HAM
alloSCT, if MSD
available
200 mg
180
mg/m2
10 g
120
mg/m2
100
mg/m2
18 g/m2
Cytarabine
41.1 g/m2
Anthracyclines
400 mg/m2
18 g/m2
1.9 g/m2
1.4 g/m2
1.8 g/m2
Figure 2
Overview of the AML-BFM 93 regimen and cumulative doses of cytarabine and anthracyclines. HAM ¼ high-dose cytarabine/
mitoxantrone; SR ¼ standard risk; HR ¼ high-risk; RT ¼ CNS radiotherapy; ADE ¼ cytarabine/daunorubicin/etoposide; AIE ¼ cytarabine/idarubicin/
etoposide; HAE ¼ high-dose cytarabine/etoposide.
M0 and M7 subtypes always required immunological confirmation.18,19 Day 15 bone marrow aspirates were also reviewed
centrally.
Definitions and statistics
Figure 3
Risk classification by morphology, cytogenetics and
response on day 15 – according to the current AML-BFM trials (until
trial 93 information about cytogenetics were not included).
Patients and methods
Eligibility (see editorial)
The entry criteria for studies AML-BFM 78–93 were: Newly
diagnosed AML (diagnosis confirmed by the reference laboratory), patient aged between 0 and 17 years, and written
informed consent signed by the patient or parent. Patients with
myelosarcoma, secondary AML, myelodysplastic syndrome or
Down’s syndrome as well as patients with a pretreatment for
more than 14 days were registered and treated with the standard
study therapy – partially with reduced dosages – but analysed
separately from the study patients (see paragraph ‘Down
syndrome’).
Diagnosis
The initial diagnosis of AML and its subtypes was established
according to the FAB classification.17–19 Initial smears were
centrally reviewed at the University Children’s Hospital in
Münster and in addition reviewed by a regular panel, including an external investigator (H Löffler). The diagnoses of
Complete remission (CR) was defined according to the CALGB
criteria20 with minor deviations as described below, and had to
be achieved at the end of intensification treatment.
Definition of CR: p5% leukaemic blasts in the bone marrow
with signs of normal haematopoiesis in the bone marrow and
with clear signs of regeneration of normal blood cell production
in the peripheral blood (platelets 480 109/l without transfusions, neutrophils 41.0 109/l), and no leukaemic cells in the
peripheral blood or anywhere else (the required platelet and
neutrophil number is lower than originally requested, because
many patients have slow regeneration especially of platelets
after the intensive therapy used today).
Early death (ED) patients were those dying before or within
the first 6 weeks of treatment. Response after induction was
evaluated on day 15 by the blast cell count in the bone marrow
p/45%.
Since the main cause of death is different in each specific
phase of initial treatment, ED was subclassified into (a) ED
before starting treatment, (b) ED during and after the first therapy
course (p14 days of treatment), (c) ED in aplasia between day
15–42 before remission with bone marrow regeneration is
generally seen (after the second induction course). This
classification reflects the ED rate due to initial problems
(hyperleukocytosis, leukostasis) or due to aplasia after induction
therapy.
Nonresponders (NR): All patients not achieving remission
(CR) and surviving the first 6 weeks of treatment were classified
as NR. Patients with CR criteria lasting less than 4 weeks
were also reported as NR (type V failure according to CALGB
criteria).
Toxicity was assessed according to the NCI common toxicity
criteria.21
Leukemia
Improved survival in paediatric AML, BFM strategy
U Creutzig et al
2034
EFS was calculated from the date of diagnosis to the last
follow-up or first event (failure to achieve remission, resistant
leukaemia, relapse, second malignancy or death of any
cause). Patients who did not attain a complete remission
were considered failures at time zero. Survival was calculated
from the date of diagnosis to death of any cause or last
follow-up, disease-free survival (DFS) from the date of remission
to the first event (relapse, second malignancy, death of any
cause).
Univariate analysis was conducted by the Wilcoxon test for
quantitative variables and Fisher’s exact test for qualitative
variables. When frequencies were sufficiently large, w2 statistic
was used. Probabilities of survival were estimated using the
Kaplan–Meier method with standard errors according to Greenwood, and were compared with the log-rank test. Cumulative
incidence functions of relapse and death in CCR were
constructed by the method of Kalbfleisch and Prentice.
Functions were compared with Gray’s test. Analysis of efficacy
data was performed according to the intent-to-treat principle.
Toxicity data were evaluated for treatment groups. Computations were performed using Statistical Analysis System (SAS
Version 6.12, SAS Institute Inc, Cary, NC, USA).
The database for all analyses was ‘frozen’ on March 1, 2002.
Patient characteristics
From January 1978 to June 1998, 1111 patients were enrolled in
studies AML-BFM 78–93. A comparison of initial patient data
(see Table 3) of AML-BFM 78, 83, 87 and 93 did not reveal
clinically important differences, except for the shift from FAB
type M1 to M2 seen from studies 78/83 to 87/93 due to
more precise definitions of these types and the introduction of
the FAB-types M0 (1991)19 and M7 (1985).18
Results
Overall outcome in AML-BFM 78, 83 and 87, 93
The results of the four trials are given in Table 4. In addition, the
results in patients o15 years of age are presented separately in
order to facilitate comparisons with other studies (Table 4a).
Overall, the probabilities of survival, EFS and DFS have
improved in the four consecutive studies – with the exception of
AML-BFM 87 (see Figures 4 and 5).
The ED rate (before therapy or during the first 42 days)
decreased from 12.6% in AML-BFM 78 to 7.4% in AML-BFM 93
(P-trend ¼ 0.03). The decrease was seen mainly during the first
phase of treatment (before day 15), see Table 4b. Death rates
between day 15 and 150 remained similar in all studies
(Table 4b). The percentage of nonresponse did not change
significantly (12.1% in AML-BFM 83, 10.4% in AML-BFM 93,
Table 4). The main reason for a better outcome was a reduction
in relapse rates.
Results of AML-BFM 78: Results of the first AML-BFM trial
were considerable, with a 5-year survival of 4274%. This was
achieved by an 8-week induction consolidation therapy,
including CNS irradiation, followed by continuous maintenance
treatment over 2 years. These treatment elements were the basis
for the following trials.
Results of AML-BFM 83: The main difference between the
first two AML-BFM studies, 78 and 83, was the addition of an
Leukemia
8-day ADE induction treatment in the second trial. Due to this
intensification, the relapse rate was reduced but not the rate
of induction failures. The probability of a 5-year EFS increased
from 3874% in AML-BFM 78 to 4774% in AML-BFM 83,
Plogrank 0.13 (P-survival to 5274%, Plogrank 0.07). The improvement of the 5-year DFS was significant in the second trial (from
4875 to 6274%, Plogrank 0.03). According to the results in
AML-BFM 83, two different risk groups could be identified by
combinations of predominantly pretherapeutic parameters (see
the section ‘Risk groups’).
Results of AML-BFM 87: AML-BFM 87 compared prospectively whether CNS irradiation could be abandoned by adding
two cycles of intensification with high-dose cytarabine and
etoposide after consolidation, and furthermore whether this
could improve prognosis compared to AML-BMF 83.
Out of 307, 230 (75%) children achieved complete remission.
Kaplan–Meier estimates for survival, EFS and DFS of 5 years
were: 4973, 4173 and 5573%.
Preventive cranial irradiation was either given or not in
children without initial CNS involvement after achieving
remission. It remained mandatory for patients with leukocyte
count 470 000/ml and/or initial CNS involvement. Results
excluding the latter group as well as patients with events before
the scheduled time of irradiation showed that the cumulative
incidence (CI) of relapses was significantly lower in irradiated
than in nonirradiated patients (5-year CI 2975% vs 5076%,
P(Gray) ¼ 0.001; Figure 6). This comparison included randomised and non-randomised patients; similar incidences were
found in randomised patients only (5-year CI 33714 vs
56712%, P(Gray) ¼ 0.12). Relapses in nonirradiated children
occurred mainly in the bone marrow and less often in the CNS.
Results of AML-BFM 93: In AML-BFM 93, 387 of 471
(82%) patients achieved remission. The estimated probabilities
for 5-year survival, EFS and DFS were 5772, 5072 and
6173%, respectively. Idarubicin induction resulted in a
significantly better blast cell reduction in the bone marrow on
day 15 (17% of patients with 45% blasts compared to 31% of
patients in the daunorubicin arm, Pw2 ¼ 0.01).22 However,
probabilities of 5-year EFS (IDA 5174% vs ADE 5074%,
Plogrank 0.72) and survival were similar (IDA 6074% vs ADE
5774%, Plogrank 0.55).22,23
Estimated survival and pEFS were higher in AML-BFM 93 than
in AML-BFM 87 (Plogrank 0.01; Table 4, Figures 4 and 5). This
improvement, however, concerned only the 310 high-risk
patients (remission rate in AML-BFM 93 vs AML-BFM 87: 78
vs 68%, P ¼ 0.007, and 5-year pEFS 42 vs 31%, Plogrank 0.03;
Figure 7). There was also a significant difference for high-risk
patients in terms of survival (5073 vs 3973%, Plogrank 0.03),
but not of DFS (5473 vs 4674%, Plogrank 0.46). As to the 161
standard-risk patients, 5-year survival, pEFS and DFS were
similar to those of AML-BFM 87: 7174 vs 7175%, Plogrank
0.88; 6574 vs 6375%, Plogrank 0.5; 7374 vs 7075%, Plogrank
0.42, respectively. Outcome of the 28 high-risk patients
receiving an allogeneic matched related donor SCT in first CR
was in the same range as of the non-SCT high-risk patients (DFS
6479%). Figure 8 shows the estimated survival for high-risk
patients in AML-BFM 87 and 93 with and without SCT.
Results by randomisation of early HAM vs late HAM in
high-risk patients: In all, 196 patients were randomised to
either early HAM (n ¼ 98) or late HAM (n ¼ 98). Overall results
regarding response and relapse rates were similar in both arms
Improved survival in paediatric AML, BFM strategy
U Creutzig et al
2035
Table 3
Initial patient data of trials AML-BFM 78, 83, 87 and 93
AML-BFM 78
AML-BFM 83
AML-BFM 87
AML-BFM 93
n
%
n
%
n
%
n
%
Number of eligible patients
Gender (male)
151
81
53.6
182
99
54.4
307
166
54.1
471
255
54.1
Age o2 years
2–9 years
X10 years
23
55
73
15.2
36.4
48.3
33
69
80
18.1
37.9
44.0
67
123
117
21.8
40.1
38.1
112
174
185
23.8
36.9
39.3
Leukocytes ( 103/mm3)
o20
20–o100
X100
71
47
33
47.0
31.1
21.9
84
60
38
46.2
33.0
20.9
131
118
58
42.7
38.4
18.9
241
143
87
51.2
30.4
18.5
CNS leukemia (yes)
(ND or questionable)
13
(3)
8.8a
10
(13)
5.9a
36
(13)
12.2a
45
(36)
10.3a
42
(39)
29.4a
85
(45)
32.4a
96
(89)
25.1a
0b
37b
(18)
37
(23)
5
45
(17)
48
7
3b
0
0
20.3
(9.9)
20.3
(12.6)
2.7
24.7
(9.3)
26.4
3.8
1.6
0
17b
30b
(16)
84
(57)
15
66
(38)
69
8
17b
1
5.5
9.8
(5.2)
27.3
(18.6)
4.9
21.5
(12.4)
22.5
2.6
5.5
0.3
25
55
(35)
125
(94)
23
91
(36)
101
16
31
4
5.3
11.7
(7.4)
26.5
(20.0)
4.9
19.3
(7.6)
21.4
3.4
6.6
0.8
123
12
8
1
3
20.3a
13.6
1.7
5.1
161
42
25
12
5
28.8a
17.1
8.2
3.4
185
82
41
25
16
28.7a
14.3
8.7
5.6
21
26
4
35.6
44.1
6.7
29
75
17
19.9
51.4
11.6
58
146
31
20.3
51
10.8
Blasts day 15 45%
(ND)
FAB types
M0
M1
(M1 with Auer)a
M2
(M2 with Auer)a
M3
M4
(M4eo)
M5
M6
M7
Other (basophilic leukaemia)
Karyotypes ND
Cytogenetic favourablec
t(8;21)
inv(16)
t(15;17)
Normal
Other
t(9;11)
0
36
(17)
34
(26)
6
40
(12)
32
3
0
151
0
24
(11)
23
(17)
4
27
(8)
21
2
0
ND ¼ no data.
Percentage of patients with data, no information about Auer rods in eight patients of study 83 and four patients of study 87.
b
P(w2) o0.05.
c
Definition of favourable cytogenetics: t(8;21), t(15;17), inv(16).
a
(5-year survival, EFS and DFS in early HAM compared with
late HAM patients were: 5775 vs 5475%; 4975 vs 4175%;
and 5676 vs 4976%). The pEFS was slightly higher in
patients initially treated with daunorubicin who received
early HAM instead of late HAM, whereas results with early or
late HAM were similar in patients initially treated with
idarubicin.24
Toxicity in AML-BFM 93 during intensive therapy
elements and according to randomisation: Nonhaematological toxicities of CTC-grade 3/4 were mainly infections
occurring in 17–18% of patients during induction and HAM,
and in 11–12% of patients during consolidation and late
intensification HAE. Other nonhaematological toxicities were
rarely seen (o3% of patients) during the intensive therapy
elements of AML-BFM 93. The infection rate in the AIE group
was slightly higher than in the ADE group (P-trend ¼ 0.016), and
duration of aplasia (until neutrophil recovery to 500/ml) was also
2 days longer in the AIE patients. Toxicities were similar in the
arms of early HAM and late HAM.
Toxicity in trials AML-BFM 78–93
To compare acute toxicities in different studies, fatal events
are listed in Table 4b according to time periods. EDs, mainly
due to haemorrhage or leukostasis, occurring before or
during the first 15 days of treatment, could be reduced from
trial to trial. The number of deaths in aplasia or early after
achieving CR (day 15–150) did not change significantly during
the studies.
Prophylactic CNS irradiation was used generally in the BFM
trials. Only in AML-BFM 87 a significant group of children did
not receive irradiation. In order to evaluate late sequelae, longterm survivors were tested for neuropsychological functions
Leukemia
Improved survival in paediatric AML, BFM strategy
U Creutzig et al
2036
Table 4
Results in trials AML-BFM 78, 83, 87 and 93 (total groups): (a) results in trials AML-BFM 78, 83, 87 and 93 (only patients o 15 years at
diagnosis); (b) death before or during induction and intensification in trials AML-BFM 78, 83, 87 and 93 (total groups)
AML-BFM 78
n
Number of patients
Median follow-up of pts. in CCR (years, range)
a
AML-BFM 83
% (s.e.)
151
15.2
n
182
(2.1–19.4)
11.1
AML-BFM 87
AML-BFM 93
% (s.e.)
n
% (s.e.)
n
% (s.e.)
100
307
100
471
100
(4.4–16.6)
9.4
(2.6–13.8)
5.1
(1.1–9.0)
19
13
119
12.6
8.6
78.8
21
22
139
11.5
12.1
76.4
28
49
230
9.1
16.0
74.9
35
49
387
7.4
10.4
82.2
Death in CCR (cumulative incidence)
Relapse (cumulative incidence)
(with CNS involvement)
10
57
9
5.3 (3)
35.2 (6)
5.3 (3.2)
7
50
7
3.3 (2)
25.8 (5)
3.8 (2.3)
9
98
15
2.9 (2)
30.6 (4)
4.9 (2.2)
18
131
23
3.8 (1)
28.1 (3)
4.9 (1.3)
Lfu in CCRb
12
11
6.0
18
5.9
Early deaths (total)
Nonresponse
CR achieved
7.9
9
1.9
p survival
At 5 years
At 10 years
42 (4)
39 (4)
52 (4)
50 (4)
49 (3)
48 (3)
57 (2)
pEFS
At 5 years
At 10 years
38 (4)
36 (4)
47 (4)
46 (4)
41 (3)
40 (3)
50 (2)
pDFS
At 5 years
At 10 years
48 (5)
44 (5)
62 (4)
60 (4)
55 (3)
53 (3)
61 (3)
Allogeneic BMT in first CR
(a) only patients o15 years at diagnosis
Number of patients
Median follow-up of pts. in CCR (years, range)
Early deathsa
Nonresponse
CR achieved
4
2.2
17
5.5
42
8.9
146
15.0
18
12
116
100.0
(2.1–15.0)
12.3
8.2
79.5
161
11.3
18
19
124
100.0
(4.4–15.0)
11.2
11.8
77.0
283
9.4
25
47
211
100.0
(3.7–13.8)
8.8
16.6
74.6
427
7.5
31
42
354
100.0
(1.1–9.0)
7.3
9.8
82.9
Death in CCR (cumulative incidence)
Relapse (cumulative incidence)
(with CNS involvement)
Lfu in CCRb
9
56
9
12
4.8 (3)
35.7 (6)
5.5 (3.3)
8.2
6
44
6
10
3.1 (2)
26.1 (5)
3.7 (2.4)
6.2
6
89
14
18
2.1 (1)
30.0 (4)
4.9 (2.3)
6.4
15
118
21
7
3.5 (1)
27.8 (3)
4.9 (1.4)
1.6
p survival
At 5 years
At 10 years
61
54
42 (4)
40 (4)
82
56
52 (4)
51 (4)
137
49
50 (3)
48 (3)
137
58 (2)
pEFS
At 5 years
At 10 years
56
49
39 (4)
36 (4)
75
51
48 (4)
47 (4)
117
38
42 (3)
41 (3)
124
51 (3)
pDFS
At 5 years
At 10 years
55
49
48 (5)
45 (5)
75
51
62 (4)
60 (4)
117
36
57 (3)
54 (3)
104
62 (3)
100
471
4
21
35
100
(b) death before or during induction and intensification in trials AML-BFM 78, 83, 87 and 93 (total groups)
Patients No.
151
100
182
100
307
ED before therapy
2
9
4
9.3
9.8
ED oday 15
12
9
14
15
9.9
13
7.1
26
Death day 15 –o150 (during intensification courses)c
5.9
8.5
5.3
7.5
CR, complete remission; CCR, continuous complete remission; s.e., standard error; lfu, lost to follow-up.
a
Early deaths are defined as death until day 42 and reported in detail (a) before starting therapy (b) during the first 14 days of treatment in Table 4b.
b
Lfu after 0.7–11.5, median 8.7 years (all patients).
c
Irrespective of response or nonresponse.
such as concentration, intelligence quotient and academic
performance. Although no statistically different results regarding
intelligence quotient and concentration or academic perforLeukemia
mance could be shown, questionnaires to patients and parents
revealed some problems in social behaviour and school
attendance in children who had been irradiated prior to 5 years
Improved survival in paediatric AML, BFM strategy
U Creutzig et al
2037
Figure 4
Estimated probability of survival in patients of trials
AML-BFM 78–93, 5-year data given. Slash indicates the patient alive
with the shortest follow-up.
Figure 7
Comparison between the estimated probability of eventfree survival in standard and high-risk patients in trial AML-BFM 93
and trial AML-BFM 87, 5-year data given (slash: see Figure 4).
Figure 5
Estimated probability of event-free survival in patients of
trials AML-BFM 78–93, 5-year data given (slash: see Figure 4).
Figure 8
Estimated probability of survival for high-risk patients
with or without matched related (MRD) SCT in first CR (survival 443
years, median time to SCT) in trials AML-BFM 87 and 93, 5-year data
given (slash: see Figure 4).
in the four AML-BFM trials so far, including two brain tumors
and two thyroid carcinomas).
Risk factors
Table 5 details the results for different risk parameters, showing
an improvement especially for children with a lower risk (first
standard-risk definition was slightly different from the currently
used definition, see next paragraph) in AML-BFM 83 compared
to AML-BFM 78 and in high-risk patients in AML-BFM 93
compared to the previous studies. The better outcome concerned patients with a blast count 45% on day 15, as well as
patients with hyperleukocytosis.
Figure 6
Cumulative incidence of relapses in trial AML-BFM 87 in
patients with and without CNS irradiation, 5-year data given (slash: see
Figure 4).
of age.25 An analysis of growth and endocrine system did not
show differences between irradiated and nonirradiated children.
Finally, the overall risk of secondary malignancies was low (16
Risk group definition: The main risk factors had been
defined according to the results of AML-BFM 836 and were
reanalysed, including the data of AML-BFM 87,11 leading to the
current risk definition (Figure 3).
In AML-BFM 83, among the different FAB types, the
difference in prognosis was significant in patients with (a) M1
Leukemia
Improved survival in paediatric AML, BFM strategy
U Creutzig et al
2038
Table 5
Results according to different risk parameters in trials AML-BFM 78–93 (5-year pEFS (%), only for subgroups nX10)
Presenting feature
AML-BFM 78
AML-BFM 83
AML-BFM 87
AML-BFM 93
Total
number of
patients
pEFS
(s.e.)
Total
number of
patients
pEFS
(s.e.)
Total
number of
patients
EFS (s.e.)
Total
number of
patients
EFS (s.e.)
SR
HR
61a
90a
43 (6)
36 (5)
51
131
78 (6)
35 (4)
99
208
63 (5)
31 (3)
161
310
65
42 (3)
Cytogenetics favourableb
t(8;21)
inv(16)
t(15;17)
Cytogenetics normal
Cytogenetics other
11q23 (excl. t(9;11))
t(9;11)
ND
12
8
1
3
21
26
ND
4
83 (11)
ND
ND
ND
43 (11)
39 (10)
ND
ND
42
25
12
5
29
75
ND
17
55 (8)
64 (10)
33 (14)
82
41
25
16
58
146
20
31
70
63
80
69
51
41
50
48
(5)
(8)
(8)
(12)
(7)
(4)
(11)
(9)
ND
37
37
5
45
17
48
7
3
62 (8)
54 (8)
ND
49 (8)
77 (10)
27 (6)
ND
ND
17
30
84
15
66
38
69
8
17
18 (9)
25
55
125
23
91
36
101
16
31
40
48
60
65
55
80
37
31
52
(10)
(7)
(5)
(10)
(5)
(7)
(5)
(12)
(9)
101
42
60 (5)
36 (7)
177
85
54 (4)
27 (5)
286
96
57 (3)
44 (5)
FAB
FAB
FAB
FAB
FAB
FAB
FAB
FAB
FAB
M0
M1
M2
M3
M4 (excl. M4eo)
M4eo
M5
M6
M7
ND
ND
ND
36
34
6
40
12
32
3
ND
44 (8)
41 (8)
ND
32 (7)
42 (14)
41 (9)
45 (9)
28 (5)
ND
35 (12)
29
47
52
53
42
53
32
(11)
(9)
(5)
(13)
(6)
(8)
(6)
Blasts day 15 p5%
Blasts day 15 45%
ND
ND
Leukocytes o 100 000/ml
Leukocytes 4100 000/ml
118
33
44 (5)
18 (7)
144
38
54 (4)
21 (7)
249
58
45 (3)
24 (6)
384
87
52 (3)
39 (5)
Age o2 years
Age X2 years
23
128
44 (10)
37 (4)
33
149
42 (9)
48 (4)
67
240
22 (5)
47 (3)
112
359
41 (5)
53 (3)
CNS positive
CNS neg.
13
135
23 (12)
41 (4)
10
159
30 (15)
49 (4)
36
260
36 (8)
42 (3)
45
390
42 (8)
53 (3)
ND ¼ no data.
a
Risk definition in study 78 according to morphological parameters only.
b
Definition of favorable cytogenetics: t(8;21), t(15;17), inv(16).
with or without Auer rods, (b) in patients with M2 with a white
blood cell count lower or higher than 20 000/ml and (c) in M4
patients with or without atypical eosinophils (P each o0.07).6
According to these results, two different risk groups could be
identified by combinations of predominantly pretherapeutic
parameters: The low-risk group, comprising 32% of the patients,
included the FAB types with granulocytic differentiation and
specific additional features: FAB M1 with Auer rods, FAB M2
with WBC o20 000 mm3, FAB M3 (all patients) and FAB M4
with eosinophils. The 5-year Kaplan–Meier estimation of DFS
for this group was 9074%, compared to 4775% in the highrisk group (all other patients).6
Based on a higher number of patients and the results of AMLBFM 83 and 87, the risk group definition (standard and high-risk)
was slightly modified and could be improved mainly by
including the blast cell reduction in bone marrow on day 1511
in addition to the initial morphological parameters (Figure 3).
The standard-risk group included: FAB M1/M2 with Auer rods,
FAB M4eo (if these patients presented with X5% blasts in the
bone marrow on day 15, they were shifted to the high-risk
group) and FAB M3 (all patients regardless of the blast count on
day 15); high-risk group: ) all others. The risk group definition
was prospectively used in AML-BFM 93.
Leukemia
FAB M3
With the introduction of all-trans retinoid acid (ATRA) in the
year 1994, less EDs due to haemorrhage and/or severe infections
occurred in FAB M3 patients in AML-BFM 93. Seven of 22
patients within the patient group treated without ATRA (AMLBFM 87 and during the year 1993) died early, but only one of
22 patients treated with ATRA.26 The 5-year OS and EFS of the
children receiving ATRA and chemotherapy were significantly
better than those in conventionally treated children (survival:
8779 vs 45711%, Plogrank ¼ 0.003; EFS: 76711 vs 43711%,
Plogrank ¼ 0.02).
Down syndrome
Children with Down syndrome were not included in the patient
group with de novo AML presented above.27 Compared to the
non-Down children, these patients had been treated in the 80s
mostly with a less intensive therapy, and up to 48% of them had
not been treated with specific AML therapy because their
therapy tolerance was thought to be low and high toxicity was
expected.
Improved survival in paediatric AML, BFM strategy
U Creutzig et al
Since AML-BFM 93, treatment recommendations were
given (ie less intensive treatment like in non-Down children of
the standard-risk group and no CNS irradiation). In AML-BFM
93, only 14% of children with Down syndrome got no
treatment.27
Before 1993, in the total Down syndrome group (including nontreated patients), outcome was poor (5-year survival
2577%, in patients treated with chemotherapy 55714%). In
AML-BFM 93, the survival probability increased to 5377%
(in patients treated with chemotherapy 6778%).
Discussion
General results
The results of the four AML-BFM trials presented above indicate
that a stepwise increase in long-term survival could be achieved
by more intensive chemotherapy applied during induction and
early intensification. Compared to the previous study AML-BFM
78, the major improvement was seen after the introduction of
the ADE induction in AML-BFM 83. The relapse rate was
reduced mainly for the group of children with low-risk factors.
Later on, in AML-BFM 93, also the relapse rate in high-risk
patients decreased as a result of a more intensive induction and
intensification therapy.
Relapse remains the main reason for failure in AML therapy.
However, prevention of any other event is also of great
importance in AML considering the relatively high number of
deaths in aplasia and fatalities in CCR. In AML-BFM 83, the
main risk factors for ED by leukostasis and/or haemorrhage were
identified (initial hyperleukocytosis and especially FAB M5 type
and concomitant hyperleukocytosis and/or extramedullary
organ involvement).28 Consequently, in the next trials, recommendations for early exchange transfusion and slow blast cell
reduction before starting induction treatment were made to
avoid these complications.
Death in aplasia during or after intensive treatment occurred
in about 4–6% of patients. Some of them were also categorised
as NR because they died during the period when achievement
of CR would have been possible. Some of these patients were
NR with resistant disease (remaining blasts 45 to 410%).
However, most patients were aplastic or hypoplastic, with
insufficient regeneration of haematopoieses; this is why these
cases should be assigned to the category ‘aplastic death’ and not
to the group of NR.
It might be possible to reduce the number of patients dying in
aplasia by improving supportive care and treatment conditions
like in stem cell transplant units. Under these circumstances,
these patients may have a chance to achieve a late remission
even after 42 up to 150 days.
The number of treatment-related deaths in CCR has remained
constant at a range of 3–4% over all studies since 1978, which is
remarkable when keeping in mind that the intensity of treatment
has increased a lot.
Nearly all of these patients – referred to as failures in CCR –
died of severe infections during intensive treatment or shortly
after the end of intensification.29,30 Transplant-related deaths in
first CR occurred in 10% of this particular patient group.
In conclusion, the possibility of improving results by further
intensification of conventional chemotherapy may be limited
because acute and late toxicity is already on threshold value.
However, there may be options to further reduce fatalities which
are not caused by progression of leukaemia.
AML-BFM-specific therapy elements
2039
Compared to other AML studies in children and adults, there are
some particular therapy elements in the AML-BFM studies.
Maintenance therapy was included since AML-BFM 78: The
results of this first trial using 2 years of maintenance after an
8-week induction/consolidation regimen combined with CNS
irradiation showed a long-term survival of 3974% after 10
years, which was much better than that expected in the 70s. The
duration of maintenance therapy was reduced to 1.5 years in
AML-BFM 87 without inferior results for patients surviving at
least 1.5 years. Studies in adults showed that maintenance
benefited patients treated with less intensive induction consolidation regimens.15 Paediatric studies, by contrast, showed
that maintenance was not beneficial.31,32 As outcome with or
without maintenance depends on initial therapy, its significance
has to be clarified in the BFM setting by randomisation in a
future trial.
A specific characteristic of the BFM trials is the general use of
prophylactic CNS irradiation. Only AML-BFM 87 compared
patients with and without CNS irradiation. Results indicated that
CNS irradiation is an essential therapy element of the BFM
protocols. Outcome was inferior in the nonirradiated groups,
especially due to bone marrow recurrences, but also due to
higher incidence of isolated or combined CNS relapses. These
results suggested that chemotherapy alone could destroy
leukaemic blasts in the bone marrow, but with less success in
the CNS. Residual blasts from the CNS may reseed the bone
marrow and lead to bone marrow relapses. Therefore, CNS
irradiation has remained a therapy element in the AML-BFM
studies. However, in the ongoing AML-BFM trial 98, a dose
reduction is tested prospectively (randomisation of CNS irradiation with 12 vs 18 Gy) in order to reduce late sequelae.
SCT in first remission from a matched related donor has been
recommended in high-risk patients since AML-BFM 87. However, an advantage of SCT could not be proven in our patients so
far.33 SCT in first CR has remained an option in our trials,
because further improvement is expected.
Results by intensification
The main difference between AML-BFM 93 and the previous
trial 87 consisted rather in the introduction of the HAM
combination and its scheduling than in the application of
high-dose cytarabine, which in the 87 protocol had been given
as late intensification after consolidation in combination with
etoposide. Furthermore, all patients of AML-BFM 87 received
daunorubicin for induction.
The randomisation of idarubicin and daunorubicin during
induction showed a significantly better blast cell reduction in
the bone marrow on day 15 in the idarubicin group, whereas
long-term outcome was similar in both treatment arms22 and
was also comparable to that of AML-BFM 87 in standard-risk
patients.
The second randomisation comparing early and late HAM
was restricted to high-risk patients only. By the early application
of HAM, an attempt was made to enhance cytotoxic activity and
consequently to achieve higher efficacy. There was evidence to
suggest that this approach – as compared to a treatment course
with lower dose intensity – might overcome resistance by more
rapid blast cell clearance and cause a reduction in the rate of
minimal residual disease.
The randomisation of early vs late HAM was carried out to
evaluate whether or not a possibly improved blast cell reduction
Leukemia
Improved survival in paediatric AML, BFM strategy
U Creutzig et al
2040
in consequence of the early application might be counterbalanced by more toxicity after the course. HAM treatment,
however, was not scheduled for standard-risk patients because
of the expected higher rate of acute adverse events and possible
late cardiotoxicity associated with additional cardiotoxic drugs.
The cumulative dose of anthracyclines, including the anthracycline analogue mitoxantrone (assuming a dose ratio of 5:1 for
daunorubicin:mitoxantrone), was 300 mg/m2 in standard-risk
and 400 mg/m2 in high-risk patients.
Treatment with HAM has been shown to be an effective,
though toxic, therapy element in adults and children with
AML.16,34,35 The dose effect of cytarabine given either at a
standard (100 mg/m2), intermediate (400 mg/m2) or high dose
(3 g/m2) during post-remission treatment was first shown by the
CALGB Study.36 Arlin et al 34 reported a higher CR rate after a
single induction course of the mitoxantrone-based regimen
(3 12 mg/m2) than after the standard regimen using daunorubicin (3 45 mg/m2) in adult patients with de novo AML.
Büchner et al16,37 demonstrated in adults that HAM given as
second induction course benefited poor-risk patients.
With a 5-year estimated survival of 5772% and an EFS of
5072%, the results of AML-BFM 93 were significantly better for
the total group of patients than for those of the previous trial
AML-BFM 87 and similar to those of the Medical Research
Council (MRC) 10 trial in children38 and the Nordic Society for
Paediatric Haematology and Oncology (NOPHO) AML trial.39
This improvement can most probably be attributed to the
intensification with HAM in high-risk patients (two-thirds of our
patients).
The effect of dose scheduling and dose intensity during postremission treatment was demonstrated by the Children’s Cancer
Group (CCG) 213P Study. Two courses of high-dose cytarabine/
asparaginase administered at 7-day intervals resulted in better
survival rates compared to 28-day intervals.40 Furthermore, in
the CCG Study 2861, patients on intensive timing during
induction (second cycle 10 days after the first cycle) had a
significantly better DFS than patients on standard timing (14
days or later depending on bone marrow status).41 Another
study in adults, reported by Estey et al,42 demonstrated that time
to achieve remission was an important predictor of survival and
DFS. This supports the hypothesis that rapid blast clearance may
avoid the development of resistance.
Results of the randomised scheduling of HAM as the second
or third treatment course after induction did not reveal major
differences in outcome. However, the treatment given for
induction had to be considered as well. Induction with
idarubicin, as opposed to daunorubicin, was more effective in
reducing the blast cell count in the bone marrow on day 15.22
Patients who received the less-intensive daunorubicin treatment
during induction had a benefit from early HAM.24
Moreover, when comparing high-risk patients of the historical
control group of AML-BFM 87 to high-risk patients treated
initially with daunorubicin followed by late HAM, it became
obvious that results were similar (5-year pEFS: high-risk, AMLBFM 87 ¼ 3173% vs AML-BFM 93 late HAM after induction
with daunorubicin ¼ 3677%, P ¼ 0.54). This finding suggests
that there might be a cumulative effect of induction with
idarubicin and HAM in high-risk patients. These results are in
line with the German AMLCG trial in adults, showing that
mainly poor-risk patients benefit more from a two-course
induction combining TAD with HAM as the second course
than from two TAD courses.16
The incidence of therapy-related deaths and infections was
similar in early HAM and late HAM, indicating that this adult
therapy was feasible also in children with AML.
Leukemia
As improved treatment results in AML-BFM 93 in children
with high-risk AML were attributed mainly to the introduction of
HAM and as the rate of toxicity was tolerable, HAM was
introduced in AML-BFM 98 as second therapy course for the
vast majority of patients (excluding M3 and Down syndrome
children), aiming at a better survival rate for standard-risk
patients, too.
Risk groups
The risk definition used in the AML-BFM trials is based on results
of AML-BFM 83 and 87.11 It allows (a) a risk-adapted treatment
for standard-risk patients, reducing the risk of treatment
mortality and severe late toxicity. In AML-BFM 93, these
patients received a lower cumulative anthracycline dose and
no allogeneic SCT in first CR. (b) For high-risk patients, more
investigational therapy elements are justified during induction
and consolidation to increase the rate and quality of remission,
and in addition, allogeneic SCT in first remission from a
matched related donor is recommended. (c) In case of
hyperleukocytosis or extramedullary organ involvement, the
definition of a third, very-high-risk group is possible. However,
at present there are no curative treatment options available, and
the risk of ED, including death at diagnosis, is high for this
group.
In other AML studies in children and adults, mostly
cytogenetic results are used for defining risk groups. Most
investigators agree that the favourable karyotypes are t(8;21),
t(15;17) and inv16.31,38,43,44 These karyotypes are closely
related to favourable morphology.11
We could demonstrate that morphology in combination with
blast cell reduction on day 15 can be used as a prognostic
parameter identifying a standard-risk group. This group includes
the known favourable cytogenetic subtypes, but, in addition,
38% more patients with normal or other karyotypes who show
the same good prognosis (Figure 9). In children with high-risk
criteria and favourable karyotypes prognosis is excellent,
indicating that these children can be shifted to the standardrisk group.
Using this feature additionally in our studies, we can show
that only high-risk patients with unfavourable cytogenetics do
Figure 9
Estimated probability of event-free survival in patients
with standard and high risk combined with favourable (fav.) or unfavourable (unfav.) cytogenetics, 5-year data given (slash: see Figure 4).
Improved survival in paediatric AML, BFM strategy
U Creutzig et al
2041
worse compared to others (standard-risk patients with either
favourable or unfavourable cytogenetics and high-risk patients
with favourable cytogenetics; Figure 9). This led to the current
risk group definition, which includes cytogenetics (Figure 3).
However, it has been reported in recently published studies
including patients of AML-BFM 93 and 98 that FLT3 internal
tandem duplications (FLT3-ITDs) predict poor clinical outcome.45,46 This was especially noticeable in the group of
standard risk patients (as defined above); therefore, in the future,
FLT3-ITDs-positive patients from the standard risk group will be
reclassified to the high-risk group (FAB M3 excluded).
Promyelocytic leukaemia – FAB M3
Outcome could be improved especially in children with M3 due
to treatment with the new agent all-trans-retinoid acid (ATRA),
which was introduced for these patients in our trials in 1994. It
could be shown that ATRA combined with intensive chemotherapy can not only induce a stable, continuous remission but also
avoid ED in these children with considerable but manageable
toxic side effects.26 However, further improvements are
necessary; one example in this direction may be the use of
arsenic trioxide in combination with chemotherapy, which
recently showed similar remission rates and minimal toxicity.47
Perspective of further improvements
In the 80s, our data indicated that it was possible to improve
prognosis in standard-risk patients by intensification of initial
chemotherapy. In the last 10 years, prognosis could be
increased in high-risk patients by more intensive therapy, for
example, by high-dose cytarabine courses. This may indicate
that these patients might even benefit from further treatment
intensification. However, this has to be done without risking
therapy-related mortality. Before starting AML-BFM 98, this was
tried to increase in a pilot study with a dosage of 3 14 mg/m2
idarubicin during induction; however, results showed a relatively high toxicity.48 Reduction of the treatment-related
mortality which averages out at more than 10% in high-risk
patients may be of the same importance. For all patients, therapy
schedules with less toxic drugs (eg less cardiotoxic drugs) and
less toxic regimens (by limiting the indications for SCT in first
remission to the very high-risk patients) will spare more patients
late toxicities.
Acknowledgements
This work is supported by Deutsche Krebshilfe e.v. Principal
investigators of the AML-BFM trials in different countries are given
in Supplementary Information.
Supplementary Information
Supplementary Information accompanies the paper on the
Leukemia website (http://www.nature.com/leu).
References
1 Riehm H, Gadner H, Welte K. Die West-Berliner Studie zur
Behandlung der akuten lymphoblastischen Leukämie des Kindes –
Erfahrungsbericht nach 6 Jahren. Klin Pädiatr 1977; 189: 89–102.
2 Scheer U, Schellong G, Riehm H. Verbesserte Prognose der
akuten myeloischen Leukämien bei Kindern nach intensivierter
Anfangstherapie. Klin Pädiatr 1979; 191: 210–216.
3 Creutzig U, Ritter J, Riehm H, Langermann H-J, Henze G, Kabisch
H et al. Improved treatment results in childhood acute myelogenous leukemia: a report of the German cooperative study
AML-BFM-78. Blood 1985; 65: 298–304.
4 Rai KR, Holland JF, Glidewell OJ, Weinberg V, Brunner K, Obrecht
JP et al. Treatment of acute myelocytic leukemia: a study by
Cancer and Leukemia Group B. Blood 1981; 58: 1203–1212.
5 Büchner T, Urbanitz D, Hiddemann W, Rühl H, Ludwig WD,
Fischer J et al. Intensified induction and consolidation with or
without maintenance chemotherapy for acute myeloid leukemia
(AML): two multicenter studies of the German AML Cooperative
Group. J Clin Oncol 1985; 3: 1583–1589.
6 Creutzig U, Ritter J, Schellong G. Identification of two risk groups
in childhood acute myelogenous leukemia after therapy intensification in the study AML-BFM-83 as compared with study AMLBFM-78. Blood 1990; 75: 1932–1940.
7 Baehner RL, Bernstein ID, Sather H, Higgins G, McCreadie S,
Chard RL et al. Contrasting benefits of two maintenance programs
following identical induction in children with acute nonlymphocytic leukemia: a report from the Childrens Cancer Study Group.
Cancer Treat Rep 1984; 68: 1269–1272.
8 Pinkel D, Simone J, Hustu HO, Aur RJ. Nine years’ experience with
‘total therapy’ of childhood acute lymphocytic leukemia. Pediatrics 1972; 50: 246–251.
9 Dahl GV, Simone JV, Hustu HO, Mason C. Preventive central
nervous system irradiation in children with acute nonlymphocytic
leukemia. Cancer 1978; 42: 2187–2192.
10 Creutzig U, Hofmann J, Ritter J, Schellong G. Therapierealisierung
und Komplikationen in der Therapiestudie BFM-83 für die akute
myeloische Leukämie. Klin Pädiatr 1988; 200: 190–199.
11 Creutzig U, Zimmermann M, Ritter J, Henze G, Graf N, Löffler H
et al. Definition of a standard-risk group in children with AML. Br J
Haematol 1999; 104: 630–639.
12 Carella AM, Berman E, Maraone MP, Ganzina F. Idarubicin in the
treatment of acute leukemias. An overview of preclinical and
clinical studies. Haematologica 1990; 75: 159–169.
13 Berman E, McBride M. Comparative cellular pharmacology of
daunorubicin and idarubicin in human multidrug-resistant leukemia cells. Blood 1992; 79: 3267–3273.
14 Hollingshead LM, Faulds D. Idarubicin. A review of its pharmacodynamic and pharmacokinetic properties, and therapeutic
potential in the chemotherapy of cancer. Drugs 1991; 42: 690–719.
15 Kesavan S, Lincoff AM, Young JB. Anthracycline-induced cardiotoxicity. Ann Intern Med 1996; 125: 47–58.
16 Büchner T, Hiddemann W, Wörmann B, Löffler H, Gassmann W,
Haferlach T et al. Double induction strategy for acute myeloid
leukemia: the effect of high-dose cytarabine with mitoxantrone
instead of standard-dose cytarabine with daunorubicin and
6-thioguanine: a randomized trial by the German AML Cooperative Group. Blood 1999; 93: 4116–4124.
17 Bennett JM, Catovsky D, Daniel MT, Flandrin G, Galton DAG,
Gralnick HR et al. Proposed revised criteria for the classification of
acute myeloid leukemia. Ann Intern Med 1985; 103: 626–629.
18 Bennett JM, Catovsky D, Daniel MT, Flandrin G, Galton DAG,
Gralnick HR et al. Criteria for the diagnosis of acute leukemia of
megakaryocyte lineage (M7). A report of the French–American–
British Cooperative Group. Ann Intern Med 1985; 103: 460–462.
19 Bennett JM, Catovsky D, Daniel MT, Flandrin G, Galton DA,
Gralnick HR et al. Proposal for the recognition of minimally
differentiated acute myeloid leukaemia (AML-M0). Br J Haematol
1991; 78: 325–329.
20 Cheson BD, Cassileth PA, Head DR, Schiffer CA, Bennett JM,
Bloomfield CD et al. Report of the National Cancer Institutesponsored workshop on definitions and response in acute myeloid
leukemia. J Clin Oncol 1990; 8: 813–819.
21 The National Cancer Institute. Investigators Handbook, A Manual
for Participants in Clinical Trials of Investigational Agents
Sponsored by the Division of Cancer Treatment. The National
Cancer Institute, Washington, 1986.
22 Creutzig U, Ritter J, Zimmermann M, Hermann J, Gadner H,
Blütters-Sawatzki R, et al, for the AML-BFM Study Group.
Idarubicin improves blast cell clearance during induction therapy
Leukemia
Improved survival in paediatric AML, BFM strategy
U Creutzig et al
2042
23
24
25
26
27
28
29
30
31
32
33
34
Leukemia
in children with AML: results of study AML-BFM 93. Leukemia
2001; 15: 348–354.
Creutzig U, Berthold F, Boos J, Fleischhack G, Gadner H, Gnekow
A et al. Improved treatment results in children with AML: results of
study AML-BFM 93]. Klin Pädiatr 2001; 213: 175–185.
Creutzig U, Ritter J, Zimmermann M, Reinhardt D, Hermann J,
Berthold F, et al, for the AML-BFM Study Group. Improved
treatment results in high risk pediatric AML patients after
intensification with high dose Ara-C and mitoxantrone: results of
study AML-BFM 93. J Clin Oncol 2001; 19: 2705–2713.
Reinhardt D, Thiele C, Creutzig U. Neuropsychological sequelae
in children with AML treated with or without prophylactic
CNS-irradiation. Klin Pädiatr 2002; 214: 22–29.
Mann G, Reinhardt D, Ritter J, Hermann J, Schmitt K, Gadner H
et al. Treatment with all-trans retinoid acid in acute promyelocytic
leukemia reduces early deaths in children. Ann Hematol 2001; 80:
417–422.
Creutzig U, Reinhardt D, Diekamp S, Dworzak M, Stary J,
Zimmermann M. AML patients with Down syndrome have a high
cure rate with AML-BFM therapy with reduced dose intensity.
Leukemia 2005; 19: 1355–1360.
Creutzig U, Ritter J, Budde M, Sutor A, Schellong G. Early deaths
due to hemorrhage and leukostasis in childhood acute myelogenous leukemia: associations with hyperleukocytosis and acute
monocytic leukemia. Cancer 1987; 60: 3071–3079.
Creutzig U, Zimmermann M, Reinhardt D, Dworzak M, Stary J,
Lehrnbecher T. Early deaths and treatment-related mortality in
children undergoing therapy for acute myeloid leukemia: analysis
of the multicenter clinical trials AML-BFM 93 and AML-BFM 98.
J Clin Oncol 2004; 22: 4384–4393.
Lehrnbecher T, Varwig D, Kaiser J, Reinhardt D, Klingebiel T,
Creutzig U. Infectious complications in pediatric acute myeloid
leukemia: analysis of the prospective multi-institutional clinical
trial AML-BFM 93. Leukemia 2004; 18: 72–77.
Perel Y, Auvrignon A, Leblanc T, Vannier JP, Michel G, Nelken B,
et al., for the Group LAME of the French Society of Pediatric
Hematology and Immunology. Impact of addition of maintenance
therapy to intensive induction and consolidation chemotherapy
for childhood acute myeloblastic leukemia: results of a prospective randomized trial, LAME 89/91. J Clin Oncol 2002; 20:
2774–2782.
Steuber CP, Civin C, Krischer J, Culbert S, Ragab A, Ruymann FB
et al. A comparison of induction and maintenance therapy for
acute nonlymphocytic leukemia in childhood: Results of a
Pediatric Oncology Group Study. J Clin Oncol 1991; 9: 247–258.
Creutzig U, Reinhardt D. Current controversies: which patients
with acute myeloid leukaemia should receive a bone marrow
transplantation? A European view. Br J Haematol 2002; 118:
365–377.
Arlin Z, Case-DC J, Moore J, Wiernik P, Feldman E, Saletan S et al.
Randomized multicenter trial of cytosine arabinoside with
mitoxantrone or daunorubicin in previously untreated adult
patients with acute nonlymphocytic leukemia (ANLL). Lederle
Cooperative Group. Leukemia 1990; 4: 177–183.
35 Bishop JF, Matthews JP, Young GA, Szer J, Gillett A, Joshua D et al.
A randomized study of high-dose cytarabine in induction in acute
myeloid leukemia. Blood 1996; 87: 1710–1717.
36 Mayer RJ, Davis RB, Schifffer CA, Berg DT, Powell BL, Schulman P
et al. Intensive postremission chemotherapy in adults with acute
myeloid leukemia. N Engl J Med 1994; 331: 896–903.
37 Büchner T, Hiddemann W, Löffler G, Gassmann W, Maschmeyer
G, Heit W et al. Improved cure rate by very early intensification
combined with prolonged maintenance chemotherapy in patients
with acute myeloid leukemia: data from the AML Cooperative
Group. Semin Hematol 1991; 28: 76–79.
38 Stevens RF, Hann IM, Wheatley K, Gray RG, on behalf of the MRC
Childhood Leukaemia Working Party. Marked improvements in
outcome with chemotherapy alone in paediatric acute myeloid
leukaemia: results of the United Kingdom Medical Research
Council’s 10th AML trial. Br J Haematol 1998; 101: 130–140.
39 Lie SO, Abrahamsson J, Clausen N, Forestier E, Hasle H, Hovi L
et al. Treatment stratification based on initial in vivo response in
acute myeloid leukaemia in children without Down’s syndrome:
results of NOPHO-AML trials. Br J Haematol 2003; 122: 217–225.
40 Wells RJ, Woods WG, Lampkin BC, Nesbit ME, Lee JW, Buckley
JD et al. Impact of high-dose cytarabine and asparaginase
intensification on childhood acute myeloid leukemia: a report
from the Children’s Cancer Group. J Clin Oncol 1993; 11:
538–545.
41 Woods WG, Kobrinsky N, Buckley J, Neudorf S, Sanders J, Miller L
et al. Intensively timed induction therapy followed by autologous
or allogeneic bone marrow transplantation for children with acute
myeloid leukemia or myelodysplastic syndrome: a Children’s
Cancer Group pilot study. J Clin Oncol 1993; 11: 1448–1457.
42 Estey EH, Shen Y, Thall PF. Effect of time to complete remission on
subsequent survival and disease-free survival time in AML, RAEB-t,
and RAEB. Blood 2000; 95: 72–77.
43 Dastugue N, Payen C, Lafage Pochitaloff M, Bernard P, Leroux D,
Huguet Rigal F et al. Prognostic significance of karyotype in de
novo adult acute myeloid leukemia. The BGMT group. Leukemia
1995; 9: 1491–1498.
44 Haferlach T. More individual markers are necessary for patients
with acute myeloid leukemia (AML). Leukemia 1996; 10 (Suppl 3):
S5–S9.
45 Liang DC, Shih LY, Hung IJ, Yang CP, Chen SH, Jaing TH et al.
FLT3-TKD mutation in childhood acute myeloid leukemia.
Leukemia 2003; 17: 883–886.
46 Zwaan CM, Meshinchi S, Radich JP, Veerman AJ, Huismans DR,
Munske L et al. FLT3 internal tandem duplication in 234 children
with acute myeloid leukemia: prognostic significance and relation
to cellular drug resistance. Blood 2003; 102: 2387–2394.
47 George B, Mathews V, Poonkuzhali B, Shaji RV, Srivastava A,
Chandy M. Treatment of children with newly diagnosed acute
promyelocytic leukemia with arsenic trioxide: a single center
experience. Leukemia 2004; 18: 1587–1590.
48 Creutzig U, Körholz D, Niemeyer CM, Kabisch H, Graf N, Reiter A
et al. 3 14 mg idarubicin during induction. Results of a pilot
study in children with AML. Leukemia 2000; 14: 340–342.