Leukemia (2012) 26, 654–661

& 2012 Macmillan Publishers Limited All rights reserved 0887-6924/12

www.nature.com/leu

ORIGINAL ARTICLE

Favorable outcome in infants with AML after intensive first- and second-line treatment:
an AML-BFM study group report
U Creutzig1,2, M Zimmermann2, J-P Bourquin3, MN Dworzak4, B Kremens5, T Lehrnbecher6, C von Neuhoff2, A Sander2,
A von Stackelberg7, I Schmid8, J Starý 9, D Steinbach10, J Vormoor11 and D Reinhardt2
1
Klinik und Poliklinik für Kinderheilkunde, Pediatric Hematology and Oncology, University Hospital Münster, Münster,
Germany; 2Pediatric Hematology, Oncology, Hannover Medical School, Hannover, Germany; 3Pediatric Hematology, Oncology,
University of Zürich, Zürich, Switzerland; 4Children’s Cancer Research Institute and St Anna Children’s Hospital, Vienna, Austria;
5
Pediatric Hematology, Oncology, University of Essen, Essen, Germany; 6Pediatric Hematology and Oncology, University of
Frankfurt, Frankfurt, Germany; 7Pediatric Oncology, Hematology, Charité-Universitätsmedizin Berlin, Berlin, Germany;
8
Pediatric Hematology, Oncology, Dr von Haunersches Kinderspital, München, Germany; 9Pediatric Hematology and Oncology,
Charles University Prague, Prague and Czech Pediatric Hematology Working Group (CPH), Prague, Czech Republic;
10
Pediatric Oncology, University Hospital Ulm, Ulm, Germany and 11The Newcastle Cancer Center at the Northern Institute
for Cancer Research, Newcastle University, Newcastle, UK

Infants o1 year of age have a high prevalence of prognostically
unfavorable leukemias and a presumed susceptibility to treat-
ment-related toxicities. A total of 125 infants with acute myeloid
leukemia (AML) were treated in studies AML-BFM-98 (n ¼ 59)
and -2004 (n ¼ 66). Treatment regimens of both studies were
comparable, consisting of intensive induction followed by four
courses (mainly high-dose cytarabine and anthracyclines).
Allogeneic-hematopoietic stem-cell-transplantation (allo-HSCT)
in 1st remission was optional for high-risk (HR) patients. Most
infants (120/125 ¼ 96%) were HR patients according to morpho-
logical, cytogenetic/molecular genetic and response criteria.
Five-year overall survival was 66±4%, and improved from
61±6% in study-98 to 75±6% in study-2004 (Plogrank 0.14) and
event-free survival rates were 44±6% and 51±6% (Plogrank
0.66), respectively. Results in HR infants were similar to those
of older HR children (1–o2- or 2–o10-year olds, Plogrank 0.90 for
survival). Survival rates of HSCT in 1st remission, initial partial
response and after relapse were high (13/14, 2/8 and 20/30
patients, respectively). The latter contributes to excellent 5-year
survival after relapse (50±8%). Despite more severe infections
and pulmonary toxicities in infants, treatment-related death rate
was identical to that of older children (3%). Our data indicate
that intensive frontline and relapse AML treatment is feasible in
infants, toxicities are manageable, and outcome is favorable.
Leukemia (2012) 26, 654–661; doi:10.1038/leu.2011.267;
published online 4 October 2011
Keywords: acute myeloid leukemia; infants; children; treatment
Introduction
Infants o1 year of age with acute myeloid leukemia (AML)
generally show features of intermediate and high-risk (HR)
AML.1,2 Rearrangements of chromosome 11q23Frespectively
of the mixed lineage leukemia (MLL) geneFare the most
common abnormalities found in infants with AML or acute
lymphoblastic leukemia (ALL), consistent with a prenatal origin
of these leukemias.3 Intensive treatment is required to achieve
Correspondence: Professor Dr U Creutzig, Klinik und Poliklinik für
Kinderheilkunde, Pediatric Hematology and Oncology, University
Children’s Hospital Münster, Albert-Schweitzer-Campus 1, Münster
D-48149, Germany.
E-mail: ursula@creutzig.de
Received 20 April 2011; revised 20 August 2011; accepted 24 August
2011; published online 4 October 2011
complete remission (CR) and long-term survival. Very young
children are considered particularly vulnerable patients who
may not tolerate intensive treatment and hold a high risk for
long-term sequelae after intensive chemotherapy, cranial irra-
diation and even more after hematopoietic stem cell transplan-
tation (HSCT).4 Whereas it is well known that infants with
ALL have worse outcome than older children,5 it was suggested
that this might not be the case in AML.6,7 During
the past decade, prognosis improved not only in standard-
risk (SR) but also in HR pediatric AML.8 This report describes
the improved outcome of 108 infants with AML treated on
protocols AML-BFM-98 and -2004 in comparison with older
children.
Patients and methods
Patients
Our analysis includes patients from Germany, Austria, Switzerland
and the Czech Republic, with de novo AML diagnosed between
July 1998 and March 2010, treated on protocols AML-BFM-98,
-98-Interim (combined to AML-BFM-98) and -2004. Written
informed consent from parents or guardians was obtained at study
entry.
Out of 1155 patients o18 years enrolled (Down syndrome
excluded), 125 were infants o1 year; 136 were 1–o2 and 354
were 2–o10 years old. The upper age limit was selected in
order to exclude adolescents.
Bone marrow morphology and cytogenetics were centrally
reviewed (D Reinhardt, Hannover, and J Harbott, Giessen,
Germany). Cytogenetic analysis, as described,9 was successfully
done in 92% (n ¼ 563) of the included patients aged o10 years.
Treatment plan
Patients were treated on protocols AML-BFM-98, -98-Interim
and -2004.10,11 Treatment regimens were largely similar
(overview of study AML-BFM-2004, Figure 1). Induction with
AIE (cytarabine, idarubicin, etoposide) was given in all studies,
whereas ADxE (Dx ¼ liposomal daunorubicin) was only
randomly assigned in study-2004. Drug dosage in infants
was generally calculated according to the bodyweight instead
of body surface; only for high-dose (HD) cytarabine was the
m2-dose given age-adjusted (from 20 in o3-month- to 60% in

Induction 1 Induction 2
SR AI
ADxE
R1 S
AIE
HR HAM
IT
Day 1 15 21-28
Figure 1 Treatment schedule of study AML-BFM-2004. ADxE, cytarabine/liposomal daunorubicin/etoposide; AIE, cytarabine/idarubicin/
etoposide; AI, cytarabine (0.5 g/m2)/idarubicin; 2-CDA, 2-chlorodeoxyadenosine; HAE, high-dose cytarabine/etoposide; HAM, high-dose
cytarabine (3 g/m2)/mitoxantrone; hAM, high-dose cytarabine (1 g/m2)/mitoxantrone; IT, intrathecal therapy; R1, first random assignment;
R2, second random assignment; R3, third random assignment.
411-month-old children because of a reduced cytarabine
clearance in this age group).12 Allogeneic (allo)-HSCT from
a matched family donor (in poor responders also from an
unrelated donor) in 1st remission (CR1) was restricted to HR
patients. Cranial irradiation with 12 or 15 Gy had to be
postponed until the age of 15 months in order to reduce
neurotoxicity.11 Intrathecal cytarabine was given 12 times
(11  in SR patients) in age-related dosage adjustments. The
98-Interim regimen, which was identical to the standard arm of
study-2004, was applied after closure of study-98 (7/2003–2/
2004) until start of study-2004.
Until 9/2001, relapses were treated on protocol AML-BFM-
REZ-97, with a double induction regimen with intermediate-
dose cytarabine and liposomal daunorubicin, followed by
HSCT after consolidation.13 From October 2001, patients were
enrolled in the international protocol relapsed AML-2001/01
with a randomized FLAG (fludarabine, cytarabine, granulocyte-
colony-stimulating factor) or FLAG/DNX (plus liposomal dauno-
rubicin) induction.14,15
There were no major changes in supportive-care strategies
between the studies. The protocol guidelines recommend
general prophylaxis for pneumocystis carinii pneumonia, and
also to consider prophylaxis against viridans group streptococci
and antifungal prophylaxis.16,17
Definition and statistics
SR group included all patients with AML FAB M1/2 and Auer
rods, M3, M4eo, t(15;17)/PML-RARA, t(8;21)/AML1-ETO or
inv(16)/CBFB/MYH1 with the exception of slow early respon-
ders with 45% bone-marrow blasts on day 15 (FAB M3
excluded). All others were considered HR patients.
CR was defined according to the CALGB criteria.18 Survival
and event-free survival (EFS) were calculated from date of
diagnosis to death of any cause or last follow-up (survival) or
until first event (EFS, failure to achieve remission, resistant
leukemia, relapse, secondary malignancy or death of any
cause). Probabilities of survival were estimated using the
Kaplan–Meier method, with s.e. calculated according to Green-
wood,19 and were compared with the log-rank test. Cumulative
incidence (CI) functions of relapse and secondary malignancy
were constructed by the Kalbfleisch and Prentice method.19
Functions were compared with the Gray’s test. Cox regression
analysis was used for multivariate analysis. Toxicity was
Outcome in infants with AML
U Creutzig et al
655
AML-BFM 2004
Consolidation Intensification
12 Gy
HAE R3
haM
18 Gy
AI/ Maintenance 1 year
cytarabine IT
2-CDA
R2 haM HAE
AI HSCT from HLA-identical
siblings
42-56 88 ~112 ~140
assessed according to the NCI common toxicity criteria version
2.0 (http://ctep.cancer.gov/reporting).
Univariate analysis was conducted by the Wilcoxon test for
quantitative variables and Fisher’s exact test for qualitative
variables. When frequencies were sufficiently high, w2 statistics
were used. Computations were performed using SAS (Statistical
Analysis System Version 9.1, SAS Institute Inc., Cary, NC, USA).
Results
Patient characteristics
A total of 125 infants with AML (median age 0.6 years,
interquartile range 0.30.7 years) were enrolled in studies
AML-BFM-98 (n ¼ 59) and AML-BFM-2004 (n ¼ 66). Twelve
neonates (11%) who presented during the first 4 weeks of life
were included. Patient characteristics were similar in both
studies (data not shown). Most infants (120/125 ¼ 96%) were
classified as HR patients. Four of the five SR patients had
favorable cytogenetics with inv(16) and one presented with
t(15;17). Patient characteristics in infants aged 1–3, 4–6, 7–9 or
10–12 months were similar, with the exception that the five SR
patients were all in the highest (10–12 months) age group.
Table 1a shows the clinical and biological features in infants
compared with the older cohorts. Infants and patients aged
1–o2 years had similar characteristics. In both cohorts the
percentage of patients with hyperleukocytosis, FAB M4/M5 and
FAB M7 subtypes, as well as initial CNS and extramedullary
organ involvement (mainly of the skin), was significantly higher
than in older children. According to cytogenetic data (Table 1b),
more than half of all patients o1 year of age showed a MLL
gene rearrangement. In most of these patients t(9;11) was
identified; however, unlike infant ALL, no t(4;11) was observed.
MLL rearrangements were most frequent in o2-year olds and
significantly less frequent in older children (Po0.001).
The median follow-up period was 5.1 (0.6–12.1) years for
infants and 4.9 (0.3–12.2) years for the other age groups (1–o2
years and 2–o10 years).
Treatment results
Five-year survival in the total group (n ¼ 125) of infants was
69±4%. Survival improved from study-98 to -2004 from
61±6% (n ¼ 59) to 75±6% (n ¼ 66), Plogrank 0.14 (Figure 2a).
EFS rates were 44±6% and 51±6% (Plogrank 0.66), respectively
Leukemia
 Outcome in infants with AML
U Creutzig et al
656
Table 1a Initial patient data in infants from studies AML-BFM-98 and -2004 compared with older patients

Age groups Infants 0–o1 year 1–o2 years 2–o10 years P-value

No. of patients 125 136 354

Leukocytes (  103/ml), median (Q1–Q3) 29 080 (8800–120 000) 17 445 (6950–58 950) 17 100 (5700–54 200) 0.004

N (%) N (%) N (%)

Leukocytes X100 000/ml 35 (28) 25 (18) 51 (14) 0.003

Gender: male 61 (49) 68 (50) 190 (54) 0.57
CNS leukemiaa 27/114 (24) 17/127 (13) 25/342 (7) 0.00003
Extramedullary organ involvement 44 (36) 41 (31) 72 (21) 0.0015

FAB types o0.00001

M0 9 (7) 2 (2) 11 (3)
M1 7 (6) 4 (3) 61 (17)
M2 4 (3) 12 (9) 106 (30)
M3 1 (1) 3 (2) 20 (6)
M4 27 (22) 26 (19) 67 (19)
M5 53 (42) 52 (38) 59 (17)
M6 2 (2) 4 (3) 7 (2)
M7 21 (17) 32 (24) 20 (6)
Other 1 (1) 1 (1) 3 (1)

BM blasts day 1545%a 12/104 (12) 12/119 (10) 55/314 (18) 0.089

Risk groups
SR 5 (4) 20 (15) 145 (41) o0.00001
HR 120 (95) 116 (85) 209 (59)
Abbreviations: BM, bone marrow; HR, high risk; SR, standard risk; WBC, white blood cell count.
a
Patients with data.

Table 1b Cytogenetic subgroups in infants of study AML-BFM-98 and -2004 compared with older age groups

Cytogenetic aberration Age, n (%)

o1 year 1–p2 years 2–p10 years

Total no. of patients 125 136 354

Patients without cytogenetic/molecular genetic data 14 (11) 8 (6) 30 (9)

Total no. of patients with cytogenetic data 111 (100) 128 (100) 324 (100)
Normal karyotype 10 (9) 26 (20) 67 (21)
t(8;21) or AML1-ETO F F 62 (19)
t(15;17) or PML-RARa 1 (1) 3 (2) 19 (6)
Inversion 16, t(16;16) or CBFb/MYH11 4 (4) 11 (9) 29 (9)

Patients with MLL rearrangement 62 (56) 51 (40) 57 (18)

t(9;11) and/or MLL/AF9 20 (18) 30 (23) 25 (8)
t(11;19) (q23;p13) 4 (4) 1 (1) 3 (1)
t(10;11) and/or MLL/AF10 10 (9) 7 (6) 5 (2)
t(1;11) (variable;q23) 6 (5) 3 (2) 2 (1)
MLL rearrangement with other translocation partner 17 (15) 7 (6) 19 (6)
MLL rearrangement with unknown translocation partner 5 (5) 3 (2) 3 (1)

Others 34 (31) 37 (29) 90 (28)

t(1;22)(p13;q13) 5 (5) 1 (1) 1 (0.3)
Trisomy 8 1 (1) 7 (6) 13 (4)
der 12p 5 (5) 1 (1) 4 (0.3)
Complex karyotypesa 14 (13) 11 (9) 22 (7)
Other cytogenetic aberrations 9 (9) 17 (12) 50 (15)
Abbreviations: AML, acute myeloid leukemia; MLL, mixed lineage leukemia.
a
Three or more aberrations, at least one structural aberration, without favorable genetics, without MLL rearrangement.

(Figure 2b). A total of 43 patients relapsed (5-year CI 34±5%),
with no significant differences in both studies (CI 37±6% vs
33±6%, P(Gray) 0.58). CNS relapse occurred in 14 patients
(in 6 of them after initial CNS involvement); 6 relapses were
combined with blasts in the bone marrow and 8 isolated in the
CNS. Eight of the 14 CNS relapses were very early relapses
(that is, within 6 months after diagnosis, before CNS irradiation
was scheduled (nevertheless, 1 patient got irradiation)), the other
6 occurred later (in 2 of 36 irradiated and 3 of 44 non-irradiated
patients with an EFS 46 months).

Leukemia
 Outcome in infants with AML
U Creutzig et al
657
1.0 1.0
0.9 0.9

Event Free Survival (probability)

0.8 0.75, SE=0.06 0.8
Survival (probability)

0.7 0.7

0.6 0.6
0.61, SE=0.06
0.5 0.5
0.4
0.4
0.3
0.3
0.2
0.2
0.1
0.1 p: 1-2 0.53 1-3 0.0052 2-3 0.036
Log-Rank p = 0.14 0.0
0.0
0 1 2 3 4 5 6 7 8 9 10 11 12 13
0 1 2 3 4 6 5 7 9 8 10 11 12 13 years
years
Age < 1 year 0.48, SE=0.05 (N=125, 67 events)
AML-BFM 98 (N= 59, 24 events) Age 1-<2 years 0.48, SE=0.04 (N=136, 67 events)
AML-BFM 04 (N= 66, 16 events) Age >=2 years 0.57, SE=0.03 (N=354, 139 events)

1.0
1.0
0.9
Event Free Survival (probability)

0.9

Event Free Survival (probability)

0.8 0.8
0.7 0.7
0.6 0.6
0.51, SE=0.06
0.5 0.5
0.4 0.44, SE=0.06 0.4
0.3 0.3
0.2 0.2
0.1 0.1
Log-Rank p = 0.66 p: 1-2 0.88 1-3 0.60 2-3 0.71
0.0 0.0
0 1 2 3 4 5 6 7 8 9 10 11 12 0 1 2 3 4 5 6 7 8 9 10 11 12
years years

AML-BFM 98 (N= 59, 35 events) Age < 1 year 0.47, SE=0.05 (N=120, 64 events)
AML-BFM 04 (N= 66, 32 events) Age 1-<2 years 0.45, SE=0.05 (N=116, 61 events)
Age >=2 years 0.45, SE=0.04 (N=209, 105 events)
Figure 2 Estimated probability of (a) survival (pOS), (b) event-free
survival (pEFS) in infants of studies AML-BFM-98 and -2004. Figure 3 Estimated probability of event-free survival (pEFS) (a) in all
infants compared with the total group (b) in high-risk infants compared
with the high risk group of 1–o2 year old and 2–o10-year-old
patients of studies AML-BFM-98 and-2004.
Results according to age groups
There was no difference in survival, EFS and CI for relapse in
infants aged 0–3 (n ¼ 26), 4–6 (n ¼ 34), 7–9 (n ¼ 36) or 10–12
(n ¼ 29) months (Supplementary Figure S1). In addition, results 43±6% vs 52±7%, Plogrank 0.48, survival 71±7% vs 66±8%,
of 12 newborns (7 of them diagnosed during the first 4 days of Plogrank 0.59, respectively (Supplementary Figure S2). In addi-
life) did not differ from those of older infants: continuous CR tion, results in different MLL fusions (which comprised less
(n ¼ 7), early death (n ¼ 1), nonresponse (n ¼ 1), death in CCR numbers of patients) were not significantly different in the infant
(n ¼ 2) and relapse (n ¼ 1). group (data not shown).
When comparing EFS in infants with that in older children,
EFS was significantly higher in children aged 2–o10 years than Hematopoietic stem cell transplantation. There was no
in those aged 1–o2 years and o1 year, respectively, Plogrank difference in the number and percentage of patients undergoing
0.036 (Figure 3a). However, the majority of patients o2 years of HSCT in both studies. In total, 50 patients received an allo-
age were HR patients. Results for early death and nonresponse, HSCT: in CR1 n ¼ 14 (matched related: 7, unrelated: 7), after
5-year survival, pEFS and CI of relapse were similar in HR partial-/nonresponse n ¼ 8 and after relapse n ¼ 30 (2 after HSCT
infants and HR patients of the older age groups (1–o2 or 2–o10 in CR1). Median age at transplantation for CR1 patients was
years, Table 2, Figure 3b). The CI of death in remission (3%) 0.46 (0.4–1.7), for nonresponders 0.4 (0.3–0.8), and for relapse
and secondary malignancies were also similar. Only the rate patients 1.1 (0.5–3.3) years, respectively. In all, 13 of 14 CR1
of relapses with CNS involvement was higher in infants and patients survived. Two children relapsed and one developed a
1–o2 year-old children than in 2–10-year-old children (Table 2). secondary malignancy. Two of 8 patients transplanted after
Survival after relapse in HR infants was 50±8% and was thus initial non-/partial response and 20/30 patients transplanted
higher than in older HR patients (although not significantly in after relapse are still alive.
42-year-old children, Figure 4).

Secondary malignancy
Results in cytogenetic subgroups. Results in HR infants Two patients developed secondary malignancies (10 year CI
with or without MLL gene rearrangements were similar: EFS 5±3%). One case presented as T-ALL after 5.7 years and is still

Leukemia
 Outcome in infants with AML
U Creutzig et al
658
Table 2 Results in high-risk infants compared with older high-risk patients from studies AML-BFM-98 and -2004

Infants CI at 5 years 1–o2 years CI at 5 years 2–o10 years CI at 5 years P-value

N % (s.e.) % (s.e.) % (s.e.) (Gray/log-rank)

Patients 120 116 209

Early death 9 7.5 3 2.6 6 2.9 0.08
Non-responders 9 7.5 8 6.9 22 10.5 0.42
CR achieved 102 85.0 105 90.5 181 86.6 0.77
Death in CCR (CI) 4 3.3 (2) 4 4.2 (2) 4 2.5 (1) 0.65
Relapse (CI) 40 34 (5) 45 41 (4) 72 39 (4) 0.57
Relapse with CNS involvement (CI) 14 12 (3) 15 13 (3) 5 8 4 (1) 0.007
Second malignancy 2 1.7 1 0.9 1 0.5 0.54
PSurvival at 5 years 66 (4) 64 (5) 66 (4) 0.90a
0.74b
0.59c
PEFS at 5 years 47 (5) 45 (5) 45 (4) 0.88a
0.60b
0.71c
Abbreviations: CI, cumulative incidence; CR, complete remission; PEFS, probability of event-free survival.
a
o1-year vs 1–2-year-olds.
b
o1-year vs 42–10-year-olds.
c
1–2 year vs 42–10-year-olds.
The significant value is highlighted in bold.

1.0 Multivariate analysis

0.9 In a Cox regression analysis of EFS and survival, the following
0.8 risk factors were included: age (o1 and 1–o2 years),
hyperleukocytosis (X100 000/ml), blast count after induction
Survival (probability)

0.7
0.6
0.5
0.4
0.3
0.2
0.1
p: 1-2 0.037 1-3 0.32 2-3 0.11
0.0
0 1 2 3 4 5 6 7
years
Age < 1 year 0.50, SE=0.08 (N= 40, 19 events)
Age 1-<2 years 0.32, SE=0.07 (N= 45, 29 events)
Age >=2 years 0.35, SE=0.06 (N= 72, 41 events)
Figure 4 Estimated probability of overall survival (pOS) after relapse
in the high-risk groups of infants, 1–o2 year old and 2–o10 year old
patients.
treated on the ALL-BFM 2000 HR arm. The other patient
developed a squamous-cell carcinoma on the basis of extensive
chronic graft-vs-host-disease 5 years after HSCT and died.
Toxicity
The rate of grade III/IV acute toxicities after induction was
highest in infants and decreased in the age groups of 1–o2
and 2–o10 years: general condition 50% vs 43% vs 31%
(Pw2 0.007); severe infection 42% vs 33% vs 29% (Pw2 0.08);
pulmonary problems, including ventilation support 34% vs 24%
vs 14% (Pw2 0.01). Severe mucositis occurred less often than
in older children. There were three infants with cardiac
arrhythmias, however, only one of them showed cardiac
dysfunction (Table 3). The cumulative incidence of late effects
out of 67/81 (83%) infants surviving at least 1.5 years was
not statistically different for patients with or without HSCT
(59±11% vs 43±9%, P(Gray)0.39).
(45%), favorable cytogenetics, MLL (t(9;11), t(11;19), other
MLL), very HR cytogenetics/molecular genetics (definition see
Table 4) and HSCT as a time-dependent covariable. Only blast
count after induction 45% and favorable cyto-/molecular
genetics were of prognostic significance; HSCT was significant
for EFS but not for survival (Table 4). In a Cox model for EFS in
infants including the same variables with the age groups 0–o6
vs 6–o12 months as covariables, the latter was not significant.
8 9 10 11
Discussion
Results of studies AML-BFM-98 and -2004 show that the
outcome in infants could be improved compared with the
earlier studies,20 leading to a 5-year survival rate of 75% in
study AML-BFM-2004. These results further demonstrate that
intensive AML treatment is feasible and toxicities are manage-
able in this young age group. Importantly, even a second
intensive salvage therapy after non-response and relapse can
successfully be performed, resulting in a 5-year survival rate of
50% after relapse.
The achieved outcome data are excellent because infants and
1–2-year-old children with AML exhibited mainly unfavorable
clinical and cyto-/molecular genetic risk factors (Tables 1a
and b).2,20 Against this background, 1–2-year-olds were sepa-
rately compared with older children. They present significantly
more often with hepatosplenomegaly and extramedullary organ
involvement and show a high incidence of the FAB subtypes
M4/M5 and M7. This correlates with the high frequency of
11q23 or MLL rearrangements21 and leads to the predominance
of HR patients in this cohort. Patients with favorable cyto-
genetics were rarely reported.2 We did not see any infant
with t(8;21) and only 1 patient with t(15;17), which is
very unusual.22,23 Inversion 16 was found in 4 infants, aged
49 months.
Results in studies AML-BFM-98 and -2004 were similar within
the different age cohorts in infancy, and were also similar when

Leukemia
 Outcome in infants with AML
U Creutzig et al
659
Table 3 Toxicity after Induction in high-risk infants compared with older high-risk patients

Tox. gr. 3/4 Infants, n / total % 1–2 years, n/total % o2–10 years, n / total % P-value

General condition 49/99 50 44/102 43 52/168 31 0.007

S-GOT/s-GPT 4/97 4 12/100 12 14/168 8 0.12
Arrhythmia 3/94 3 2/98 2 4/154 3 0.92
Cardiac function (CHF) 1/83 1 4/86 5 3/137 3 0.42
Echocardiography 0/79 0 0/81 0 2/138 1 0.50
Cardiac total 3/97 3 4/100 4 4/158 3 0.80
Central neurotoxicity 3/99 3 6/103 6 3/166 2 0.18
Infection 42/101 42 36/100 36 48/168 29 0.08
Pulmonary problems 19/56 34 13/55 24 13/96 14 0.01
Mucositis 14/97 14 23/102 23 39/168 23 0.20
Significant values are highlighted in bold.

Table 4 Multivariate analysis including all patients

Feature Cox EFS Cox OS

RR 95% CI P (w2) RR 95% CI P (w2)

Age group infants 1.13 0.77–1.67 0.53 0.98 0.58–1.67 0.95

Age group 1–o2 year(s) 1.26 0.87–1.82 0.22 1.26 0.78–2.04 0.35
WBC X100 000/ml 1.18 0.81–1.71 0.39 0.94 0.56–1.59 0.83
t(9;11)+MLL/AF9 0.85 0.54–1.35 0.50 0.93 0.52–1.65 0.79
t(11;19)+MLL/ENL 0.60 0.15–2.45 0.48 0.52 0.07–3.77 0.52
Other MLL 1.17 0.78–1.75 0.44 0.74 0.41–1.33 0.31
Cytogenetic/molecular genetic favorable 0.31 0.19–0.50 0.000 0.16 0.07–0.36 0.000
Cytogenetic/molecular genetic VHR 1.50 0.84–2.70 0.17 1.76 0.86–3.63 0.12
BM blasts day 15 2. 465 1.73–3.52 0.000 2.13 1.33–3.40 0.002
HSCT* 0.46 0.26–0.80 0.006 0.65 0.34–1.22 0.18
Abbreviations: EFS, event-free survival; HSCT*, hematopoietic stem cell transplantation as a time-dependent covariable; RR, risk ratio; ; MLL,
mixed lineage leukemia; OS, overall survival; VHR, very high cytogenetics (definition: patients without favorable cytogenetics and at least one of the
following aberration: t(9;22)(q34;q11), t(7;12)(q36;p13), t(4;11)(q21;q23), t(6;11)(q27;q23), t(6;9)(p23,q24), t(8;16)(p11;q13), -7, -12p); WBC, white
blood cell count.
Significant values are highlighted in bold.

comparing them with older HR patients. There was also no
difference in EFS between the studies. However, overall survival
increased from study AML-BFM-98 to -2004 (Figure 2a), which
was partly due to better and more consequent treatment in
initial non-/partial -responders, including HSCT after blast-cell
reduction. In both studies, results after relapse were even better
in infants than in older children (Figure 4). This was not
expected and indicated that infants can tolerate intensive
relapse treatment including HSCT. There was also an improved
outcome for all AML children with MLL gene rearrangement in
study-2004 compared with study-98. Recently, unlike in other
patients, allo-HSCT in CR1 seemed to be beneficial for patients
with 11q23 aberrations (Klusmann JH, in press). This fits to the
good results after HSCT in CR1 in infants reported here and
further indicates that acute toxicities can be better handled than
in the past.15
One problem in infants is the high rate of patients with CNS
involvement, which may be due to the high frequency of
monoblastic leukemias, as these cells are able to penetrate
directly through the cytoplasm of the endothelial cells into the
brain.24 There is no satisfying treatment option, as cranial
irradiation is associated with neurological late effects. In addition,
CNS irradiation was performed late (in median 0.38 years after
diagnosis) in our treatment schedules and CNS relapses occur
much earlier than other relapses (median time to relapse 0.61 vs
1.1 years). Triple intrathecal therapy may be an option, because
low rates of CNS relapses have been reported.25
Acute toxicities in infants, apparent by the early death rate,
were unacceptably high in the past (15% in patients o2 years in
studies AML-BFM-83/87)26 and remained slightly higher than in
older patients; however, considering the high percentage of
infants with hyperleukocytosis, there was no difference in the
older age group.26 Compared to older children acute toxicities
in infants during induction chemotherapy were higher,
concerning severe infections and pulmonary toxicities (earlier
and more frequent respiratory support was necessary, probably
due to a higher pulmonary susceptibility and less tolerance to
compensate ventilation problems in infants).27 This also
indicates that close monitoring of the pulmonary condition is
required. However, the toxic death rate during intensive
chemotherapy and after HSCT was in the same range. Even in
neonates intensive chemotherapy was successfully applied,
which may be also because of the age-related dosage adjust-
ments of cytarabine,12 thus avoiding too much toxicity.
Late effects concerning especially neurological/psychological
impairments have been reported in both, the non-transplant as
well as in the transplant infant group. A high rate of severe late
effects after HSCT, for example, growth-hormone deficiency,
hypothyroidism, abnormal pubertal development, osteopenia/
osteoporosis, short stature, dental abnormalities and cataracts,
were described in young survivors of pediatric HSCT.4,28,29
In addition, radiation-induced neuropsychological sequelae
should be considered in very young patients. In our studies,
cranial irradiation was postponed until the age of 15 months for
these anticipated late effects and performed in a minority of
patients. Long-term data on CNS toxicity are not yet available
for the current infant group, however, the previous analysis from
study AML-BFM-87 has shown a tendency toward more learning

Leukemia
 Outcome in infants with AML
U Creutzig et al
660
problems (deficits in concentration) in irradiated patients aged
o5 years.30
In view of these considerations, the aim should be to optimize
treatment for infants with AML, applying the most effective
treatment with the lowest side effects.25,31,32 This implies to use
HSCT in CR1 only in selected patients in whom a clear benefit
can be expected (further analyses with higher number of patients
are needed to define this group). In addition, in future AML-BFM
studies, CNS irradiation will be substituted by triple-intrathecal
therapy.11
In comparison with the favorable outcome of infants with
AML, infants with ALL have a dismal prognosis (EFS 40±9% vs
84±1% in children aged 1–5 years), which is explained by the
high MLL incidence of 80% (EFS 40±10% in MLL-ALL).33,34
Besides all types of MLL rearrangements, an age of o6 months
was an independent risk factor in ALL. In contrast to ALL, MLL-
positivity was not an unfavorable risk factor in infant AML,
which might be due to different fusion partners of this gene in
AML and ALL (for example, t(4;11) was not found in our AML
patient populations). In addition, very young age was neither an
independent nor an associated risk factor for EFS in AML. In this
context, it has to be mentioned that the cumulative doses
of cytarabine and antracyclines were much lower in the ALL
regimens (for example, in the INTERFANT-99 regimen) than in
the AML treatment. Nevertheless, similar toxicities occurred
during ALL-regimens.34
Concerning other pediatric AML studies, experiences were
similar. Five-year EFS rates for infants with AML were generally
o50% for patients treated during the 80s and 90s until 1995
and improved during the late 90s.29 A superior outcome (3-year
EFS: 72%) in 35 infants with AML treated with three high-dose-
cytarabine courses was reported from Japan.6 Data from the
children cancer study group were favorable in infants regarding
survival after CR (71% with allo-HSCT, 61% chemotherapy
alone).35 The British Medical Research Council showed a 5-year
EFS of 59% in 57 AML infants,7 and the Nordic Society for
Paediatric Haematology and Oncology showed a 5-year EFS of
44 % in 25 infants.36 It is difficult to compare these results,
because of various criteria, for example, shorter follow-ups
or presenting only results after CR or possible exclusion of
neonates or early deaths.
In summary, the outcome of pediatric AML HR patients,
after both frontline and relapse treatment, has improved
considerably over the last 10–15 years. This has translated
into an improved outcome in infants with AML. Intensive
AML treatment is feasible in this young cohort, and toxicities are
manageable. The favorable outcome is partly due to improved
supportive care and better salvage therapy after nonresponse
and relapse.
Conflict of interest
The authors declare no conflict of interest.
Acknowledgements
We thank the colleagues, data managers and technicians of
the participating hospitals for their valuable cooperation,
Jans-Enno Müller for competent data management and Ursula
Bernsmann for her help in preparing the manuscript. The studies
were supported by the Deutsche Krebshilfe e.V. and partly by
grant from the Czech Ministry of Education (MSM0021620813).
Leukemia
References
1 Sorensen PH, Chen CS, Smith FO, Arthur DC, Domer PH,
Bernstein ID et al. Molecular rearrangements of the MLL gene
are present in most cases of infant acute myeloid leukemia and are
strongly correlated with monocytic or myelomonocytic pheno-
types. J Clin Invest 1994; 93: 429–437.
2 Pui CH, Raimondi SC, Srivastava DK, Tong X, Behm FG, Razzouk B
et al. Prognostic factors in infants with acute myeloid leukemia.
Leukemia 2000; 14: 684–687.
3 Greaves MF, Wiemels J. Origins of chromosome translocations in
childhood leukaemia. Nat Rev Cancer 2003; 3: 639–649.
4 Sanders JE, Im HJ, Hoffmeister PA, Gooley TA, Woolfrey AE,
Carpenter PA et al. Allogeneic hematopoietic cell transplantation
for infants with acute lymphoblastic leukemia. Blood 2005; 105:
3749–3756.
5 Pieters R. Infant acute lymphoblastic leukemia: lessons learned
and future directions. Curr Hematol Malig Rep 2009; 4: 167–174.
6 Kawasaki H, Isoyama K, Eguchi M, Hibi S, Kinukawa N, Kosaka Y
et al. Superior outcome of infant acute myeloid leukemia with
intensive chemotherapy: results of the japan infant leukemia study
group. Blood 2001; 98: 3589–3594.
7 Webb DK, Harrison G, Stevens R, Gibson BG, Hann I, Wheatley K.
Relationship between age at diagnosis, clinical features, and
outcome of therapy in children treated in the medical research
concil AML 10 and 12 trials for acute myeloid leukemia. Blood
2001; 98: 1714–1720.
8 Creutzig U, Zimmermann M, Ritter J, Reinhardt D, Hermann J,
Henze G et al. Treatment strategies and long-term results in
pediatric patients treated in four consecutive AML-BFM trials.
Leukemia 2005; 19: 2030–2042.
9 von Neuhoff C, Reinhardt D, Sander A, Zimmermann M, Bradtke J,
Betts DR et al. Prognostic impact of specific chromosomal
aberrations in a large group of pediatric patients with acute
myeloid leukemia treated uniformly according to trial AML-BFM
98. J Clin Oncol 2010; 28: 2682–2689.
10 Creutzig U, Zimmermann M, Lehrnbecher T, Graf N, Hermann J,
Niemeyer CM et al. Less toxicity by optimizing chemotherapy,
but not by addition of granulocyte colony-stimulating factor in
children and adolescents with acute myeloid leukemia: results of
AML-BFM 98. J Clin Oncol 2006; 24: 4499–4506.
11 Creutzig U, Zimmermann M, Bourquin JP, Dworzak MN,
Fleischhack G, von Neuhoff C et al. CNS irradiation in pediatric
acute myleoid leukemia: equal results by 12 or 18 Gy in studies
AML-BFM98 and 2004. Pediatr. Blood Cancer 2011; 57: 986–992.
12 Periclou AP, Avramis VI. NONMEM population pharmacokinetic
studies of cytosine arabinoside after high-dose and after loading bolus
followed by continuous infusion of the drug in pediatric patients with
leukemias. Cancer Chemother Pharmacol 1996; 39: 42–50.
13 Reinhardt D, Hempel G, Fleischhack G, Schulz A, Boos J, Creutzig
U. Liposomal daunorubicine combined with cytarabine in the
treatment of relapsed/refractory acute myeloid leukemia in
children. Klin Padiatr 2002; 214: 188–194.
14 Kaspers GJL, Zimmermann M, Reinhardt D, Gibson B, Tamminga R,
Aleinikova O et al. Addition of liposomal daunorubicin
(DaunoXomes) to FLAG significantly improves treatment res-
ponse in pediatric relapsed AML: final results from the interna-
tional randomised phase III study relapsed AML 2001/01
[abstract]. Blood 2009; 114: 13–14.
15 Sander A, Zimmermann M, Dworzak M, Fleischhack G,
von Neuhoff C, Reinhardt D et al. Consequent and intensified
relapse therapy improved survival in pediatric AML: results of
relapse treatment in 379 patients of three consecutive AML-BFM
trials. Leukemia 2010; 24: 1422–1428.
16 Groll AH, Ritter J, Muller FM. Prevention of fungal infections in
children and adolescents with cancer. Klin Padiatr 2001; 213
(Suppl 1): A50–A68.
17 Simon A, Fleischhack G, Marklein G, Ritter J. Antimicrobial
prophylaxis of bacterial infections in pediatric oncology patients.
Klin Padiatr 2001; 213(Suppl 1): A22–A37.
18 Cheson BD, Cassileth PA, Head DR, Schiffer CA, Bennett JM,
Bloomfield CD et al. Report of the national cancer institute-
sponsored workshop on definitions and response in acute myeloid
leukemia. J Clin Oncol 1990; 8: 813–819.

19 Kalbfleisch J, Prentice R. The statistical analysis of failure time
data. Wiley: New York, NY, 2002.
20 Vormoor J, Ritter J, Creutzig U, Boos J, Heyen P, Ludwig W-D et al.
Acute myelogenous leukaemia in children under 2 yearsFexperi-
ences of the west german AML studies BFM-78, -83 and -87.
Br J Cancer 1992; 66(Suppl XVIII): 63–67.
21 Balgobind BV, Raimondi SC, Harbott J, Zimmermann M, Alonzo TA,
Auvrignon A et al. Novel prognostic subgroups in childhood 11q23/
MLL-rearranged acute myeloid leukemia: results of an international
retrospective study. Blood 2009; 114: 2489–2496.
22 Testi AM, Biondi A, Lo CF, Moleti ML, Giona F, Vignetti M et al.
GIMEMA-AIEOPAIDA protocol for the treatment of newly
diagnosed acute promyelocytic leukemia (APL) in children. Blood
2005; 106: 447–453.
23 Creutzig U, Zimmermann M, Dworzak M, Urban C, Henze G,
Kremens B et al. Favourable outcome of patients with childhood
acute promyelocytic leukaemia after treatment with reduced
cumulative anthracycline doses. Br J Haematol 2010; 149:
399–409.
24 Abbott NJ, Patabendige AA, Dolman DE, Yusof SR, Begley DJ.
Structure and function of the blood-brain barrier. Neurobiol Dis
2010; 37: 13–25.
25 Pui CH, Howard SC. Current management and challenges of
malignant disease in the CNS in paediatric leukaemia. Lancet
Oncol 2008; 9: 257–268.
26 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.
27 Holm-Knudson RJ, Rasmussen LS. Paediatric airway manage-
ment:basic aspects. Acta Anaesthesiol Scand 2009; 53: 1–9.
28 Baker KS, Bresters D, Sande JE. The burden of cure: long-term
side effects following hematopoietic stem cell transplantation
(HSCT) in children. Pediatr Clin North Am 2010; 57: 323–342.
Supplementary Information accompanies the paper on the Leukemia website (http://www.nature.com/leu)
Outcome in infants with AML
U Creutzig et al
661
29 Ishii E, Kawasaki H, Isoyama K, Eguchi-Ishimae M, Eguchi M.
Recent advances in the treatment of infant acute myeloid
leukemia. Leuk Lymphoma 2003; 44: 741–748.
30 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.
31 Duffner PK. Long-term effects of radiation therapy on cognitive
and endocrine function in children with leukemia and brain
tumors. Neurologist 2004; 10: 293–310.
32 Pui CH, Pei D, Sandlund JT, Ribeiro RC, Rubnitz JE, Raimondi SC
et al. Long-term results of St Jude Total Therapy Studies 11, 12,
13A, 13B, and 14 for childhood acute lymphoblastic leukemia.
Leukemia 2010; 24: 371–382.
33 Moricke A, Reiter A, Zimmermann M, Gadner H, Stanulla M,
Dordelmann M et al. Risk-adjusted therapy of acute lymphoblastic
leukemia can decrease treatment burden and improve survival:
treatment results of 2169 unselected pediatric and adolescent
patients enrolled in the trial ALL-BFM 95. Blood 2008; 111:
4477–4489.
34 Pieters R, Schrappe M, De Lorenzo P, Hann I, De Rossi G,
Felice M et al. A treatment protocol for infants younger than 1 year
with acute lymphoblastic leukaemia (Interfant-99): an observa-
tional study and a multicentre randomised trial. Lancet 2007; 370:
240–250.
35 Woods WG, Neudorf S, Gold S, Sanders J, Buckley JD, Barnard DR
et al. A comparison of allogeneic bone marrow transplantation,
autologous bone marrow transplantation, and aggressive chemo-
therapy in children with acute myeloid leukemia in remission.
Blood 2001; 97: 56–62.
36 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.
Leukemia