O u t c o m e i n C... A f t e r I n i t i...

Outcome in Children With Relapsed Acute Myeloid Leukemia
A f t e r I n i t i a l T r e a t m e n t W i t h t h e F r e n c h L e u c é m i e A i q u ë
M y é l o ı̈ d e E n f a n t ( L A M E ) 8 9 / 9 1 P r o t o c o l o f t h e F r e n c h S o c i e t y
of Pediatric Hematology and Immunology
By Nathalie Aladjidi, Anne Auvrignon, Thierry Leblanc, Yves Perel, Antoine Bénard, Pierre Bordigoni, Virginie Gandemer,
Isabelle Thuret, Jean Hugues Dalle, Christophe Piguet, Brigitte Pautard, André Baruchel, and Guy Leverger
Purpose: After present first-line therapies for childhood
acute myeloid leukemia (AML), nearly 40% of patients still
relapse. The goals of this retrospective study were to determine whether these children could be treated successfully
with a salvage regimen and to establish the optimal therapeutic strategy.
Patients and Methods: In the multicentric, prospective,
Leucémie Aiquë Myéloı̈de Enfant 89/91 protocol, 106 of the
308 children enrolled between 1988 and 1998 relapsed.
Initial treatment after the first complete remission (CR1) had
been allogenic HLA-identical bone marrow transplantation
(BMT; n ⴝ 21) or chemotherapy (n ⴝ 85). Treatment procedures were scheduled according to the choice of each participating institution.
Results: When reinduction therapy was attempted, second complete remission (CR2) was obtained in 71% of
patients (68 of 96 patients). BMT was performed in 53 (78%)
of these 68 patients (autograft, mainly harvested in CR1,
n ⴝ 25; matched sibling-donor BMT, n ⴝ 12; or alternativedonor BMT, n ⴝ 16). The 5-year overall survival (OS) rate for
all 106 patients was 33%, and the disease-free survival
(DFS) rate for children in CR2 was 45%. Multivariate analysis of re-treated children showed that the 5-year OS was
higher if the CR1 had been longer than 12 months compared
with less than 12 months (54% v 24%, respectively; P ⴝ
.001) and lower if maintenance therapy had been given
after CR1 compared with chemotherapy without maintenance therapy or HLA-identical BMT (12% v 40% v 52%,
respectively; P ⴝ .002). For patients attaining CR2, the
5-year DFS rate was not significantly different for matched
sibling-donor BMT (60%), autograft (47%), or alternativedonor BMT (44%).
Conclusion: After aggressive first-line therapy, one third
of unselected, relapsing AML children could be cured. Further prospective trials are warranted to define the optimal
reinduction regimen and megadose chemotherapy and to
evaluate the late effects of these therapies.
J Clin Oncol 21:4377-4385. © 2003 by American
Society of Clinical Oncology.
N THE main collaborative groups for treatment of acute
myeloid leukemia (AML) in children, with the use of
intensive induction and consolidation therapy, results have
steadily improved over the last decade. Complete remission (CR)
and long-term event-free survival rates have reached 75% to
90% and 40% to 51%, respectively.1-8 In the French multicentric
prospective Leucémie Aiquë Myéloı̈de Enfant (LAME) 89/91
protocol, the first CR (CR1) rate was 90%, and the 5-year overall
survival (OS), event-free survival, and disease-free survival
(DFS) rates were 61%, 48%, and 55%, respectively.8 Children
with a matched sibling donor were given allogenic bone marrow
transplantation (BMT) in CR1 and had a better outcome compared with children treated with chemotherapy alone, including
high-dose cytarabine (DFS, 72% v 48%, respectively, P ⫽ .02;
relapse rate, 26% v 47%, respectively, P ⫽ .04).3
Relapse remains the main cause of treatment failure and
occurs in approximately 40% of patients who had attained a CR.
Intensive reinduction therapy and BMT for second CR (CR2) are
still associated with undoubted treatment-related mortality and
morbidity despite improved modern supportive-care procedures.
The 2-year OS rate rarely exceeds 20%.9 Important questions for
the management of relapsing AML in children are still unanswered. Is CR2 attainment possible for most of them? Which
postremission consolidation therapy should be proposed? To
help answer these questions, a retrospective study has been
conducted that has recorded the treatment and outcome of all
unselected children who experienced a relapse after initial
treatment in LAME 89/91.
I
PATIENTS AND METHODS
Patients
Between December 1988 and December 1998, 308 children were enrolled
onto the French multicentric prospective LAME 89/91 protocol, the preliminary results of which have been previously reported.3 Inclusion criteria were
as follows: age under 20 years, de novo and previously untreated AML, and
French-American-British (FAB) subtype ranging from M1 to M6. Briefly,
children who had attained CR after induction chemotherapy were treated
From the University Hospital Centers of Bordeaux, Paris-Trousseau,
Paris-Saint Louis, Nancy, Rennes, Marseille, Lille, Limoges, and Amiens,
and the Public Health Institute of Epidemiology and Development, Bordeaux, France.
Submitted November 13, 2002; accepted September 8, 2003.
Authors’ disclosures of potential conflicts of interest are found at the end
of this article.
Address reprint requests to Nathalie Aladjidi, MD, Unité d’Onco-Hématologie, Département de Pédiatrie, Hôpital des Enfants, Groupe Hospitalier
Pellegrin, Place Amélie-Raba Léon, 33076 Bordeaux, France; e-mail:
[email protected].
© 2003 by American Society of Clinical Oncology.
0732-183X/03/2123-4377/$20.00
Journal of Clinical Oncology, Vol 21, No 23 (December 1), 2003: pp 4377-4385
DOI: 10.1200/JCO.2003.11.066
Downloaded from jco.ascopubs.org on August 22, 2014. For personal use only. No other uses without permission.
Copyright © 2003 American Society of Clinical Oncology. All rights reserved.
4377
4378
ALADJIDI ET AL
either with allogeneic BMT, if they had an HLA-identical sibling donor, or
with chemotherapy including high-dose cytarabine (Fig 1). Between 1991
and 1996, in those who had been treated with chemotherapy, children were
randomly assigned to either receive further maintenance chemotherapy or not
to receive maintenance therapy. CNS prophylaxis was administered to
patients with an initial white blood cell count higher than 50 ⫻ 109/L or with
an M4 or M5 FAB subtype and consisted of five courses of intrathecal (IT)
chemotherapy (cytarabine, methotrexate, and corticosteroids). Patients with
CNS disease on diagnosis were given three additional IT doses and 24 Gy of
cranial irradiation. An autologous marrow was harvested during CR1 for
almost all children after completion of treatment. Children under 1 year old
were given two thirds of the induction doses, consolidation with high-dose
cytarabine without amsacrine, and, from 1993 on, an autologous BMT if they
had no HLA-matched donor.
The analysis of relapsed children was completed by June 1, 2001. The
clinical and biologic data were retrospectively collated from medical charts.
There was no single recommended reinduction chemotherapy schedule for
the management of relapsed disease, and procedures varied slightly according to the different institutions. Briefly, the main groups of treatments were
standard-dose cytarabine plus other agents (group 1), high-dose cytarabine
plus other agents (group 2), fludarabine plus other agents (group 3),
etoposide plus carboplatin plus other agents (group 4), and miscellaneous
treatments (group 5). However, all the children in CR2 were scheduled to
receive high-dose chemotherapy followed by BMT, either matched siblingdonor BMT, if available, or alternative-donor BMT, or autograft.
Methods
Patients were evaluated for CR2 attainment after completion of the
reinduction therapy. CR was defined as the absence of clinical symptoms of
leukemia, peripheral absolute neutrophil count more than 1 ⫻ 109/L, platelet
count more than 100 ⫻ 109/L, and normocellular marrow with less than 5%
blasts. OS after relapse was the time from relapse to death. DFS was the time
from CR2 to any relevant event (death or relapse). For the long-term
survivors, the following medical data were recorded: height, body weight,
school performance, occurrence and grading of chronic graft-versus-host
disease (GVHD), and hematologic, cardiac, endocrine, neurologic, pulmonary, or ophthalmologic disorders.
The potential prognostic factors for CR2 attainment and OS in the patients
treated with a curative prospect were sex, age at initial diagnosis, leukocyte
counts (threshold, 50 ⫻ 109/L), FAB subtypes (M1 to M6) and chromosome
analysis at initial diagnosis (favorable, t[8;21], t[15;17], inv[16], v others),
postremission treatment in CR1 (HLA-identical allograft or conventional
chemotherapy with or without maintenance), CR1 duration (threshold, 12
months), medullary or extramedullary sites of relapse, blast count at relapse
(threshold, 1.8 ⫻ 109/L), and reinduction therapy. Reinduction therapy was
analyzed in the following three main groups because of the high number of
children who were given several lines of treatments: group A, patients given
only cytarabine⫹fludarabine-based courses (groups 1 to 3 in Table 1); group
B, patients who were not given cytarabine⫹fludarabine-based courses
(groups 4 and 5 in Table 1), and group C, patients who were given either
group A–type or group B–type reinduction.
Statistical Analysis
Time to an event was calculated in months. Patients who remained
event-free until June 1, 2001, were right censored at that date. Univariate
analysis of the association between potential prognostic factors and outcomes
(CR2 attainment, OS, and DFS) was performed using the Kaplan-Meier
method and log-rank test. All variables associated with a P value less than
.25 were included in multivariate analyses. These analyses were performed
using a Cox proportional hazards model. The proportionality of hazard was
graphically checked by examining log(⫺log[S(t)]) versus t. Analyses were
performed using SAS software, version 8.2 (SAS Institute, Inc, Cary, NC).
Fig 1. The Leucémie Aiquë Myéloı̈de Enfant 89/91 protocol for childhood acute myeloid leukemia. Abbreviations: IV, intravenous; VP16, etoposide; BMT, bone
marrow transplantation.
Downloaded from jco.ascopubs.org on August 22, 2014. For personal use only. No other uses without permission.
Copyright © 2003 American Society of Clinical Oncology. All rights reserved.
4379
OUTCOME IN RELAPSED CHILDHOOD AML
Table 1.
Response to Reinduction Therapy in Relapsed AML Patients
Reinduction
No. of Courses
(N ⫽ 134)
No. of Patients Who
Achieved CR2
(n ⫽ 96)
CR2 Rate/Course (%)
Group 1: cytarabine ID ⫹ others
Cytarabine ID CIV ⫹ amsacrine
LAME 89/91 induction
LAME 89/91 first consolidation
Cytarabine ID ⫹ idarubicin ⫹ asparaginase
Cytarabine ID ⫹ VP16 ⫹ idarubicin
DCTER
Group 2: cytarabine HD ⫹ others
Cytarabine HD ⫹ amsacrine
Cytarabine HD ⫹ asparaginase ⫹ amsacrine
Cytarabine HD ⫹ idarubicin-amsacrine or idarubicin-VP16-mitoxantrone or VP16-mitoxantrone
LAME 89/91 second consolidation
Group 3: fludarabine ⫹ others
IDA-FLAG
Cytarabine ⫹ fludarabine
Group 4: VP16 ⫹ carboplatin ⫹ others
VP16 ⫹ carboplatin ⫹ bisantrene
VP16 ⫹ carboplatin
Group 5: miscellaneous, as first reinduction course
Autograft
Donor lymphocyte infusion
Others
40
17
5
2
4
10
2
25
15
3
3
4
7
6
1
23
12
11
38
4
4
30
27
12
3
1
4
6
1
19
11
3
2
2
5
5
0
6
4
2
8
0
0
8
67*
70
60
50
100
60
50
72*
73
100
66
50
71*
83
0
26*
33
18
21
0
0
26
Abbreviations: AML, acute myeloid leukemia; CR2, second complete remission; ID, intermediate dose; CIV, continuous intravenously; LAME, Leucémie Aiquë Myéloı̈de
Enfant; VP16, etoposide; DCTER, dexamethasone, ID cytarabine, etoposide, rubidomycin; HD, high dose; IDA-FLAG, idarubicine, granulocyte colony-stimulating factor,
fludarabine, HD, cytarabine.
*P is not significant.
RESULTS
Patient Characteristics at Relapse
By June 1, 2001, 106 children from 19 institutions had
relapsed (Table 2). The median duration of CR1 was 10 months
(range, 1 to 57 months). Over two thirds of all relapses (n ⫽ 66)
occurred within 1 year from CR1, whereas relapse rates in the
second, third, and fourth years from CR1 were 27%, 7%, and
4%, respectively. To date, the latest relapse was recorded 57
months after CR1. Relapses occurred off-therapy in 67% of
patients and on-therapy in 33% of patients.
The main circumstances of diagnosis of relapse were routine
hemograms (n ⫽ 59), poor tolerance of maintenance therapy
(n ⫽ 27), and medullary failure and cutaneous or neurologic
symptoms (n ⫽ 30). Bone marrow was the only site of relapse in
84 children. Extramedullary relapses occurred in 22 children,
seven after allogenic BMT (10%) and 15 after chemotherapy in
CR1 (7%, P ⫽ not significant). Five of the extramedullary
relapses were isolated (CNS, n ⫽ 1; eye, n ⫽ 1; skin, n ⫽ 2; and
testis, n ⫽ 1), and 17 were combined with bone marrow
involvement, mainly simultaneously (CNS, n ⫽ 10; skin, n ⫽ 4;
testis, n ⫽ 1; CNS and eye, n ⫽ 1; and CNS, eye, and skin, n ⫽
1). The median leukocyte count at relapse was 4.1 ⫻ 109/L
(range, 0.5 to 350 ⫻ 109/L), 68% of children had circulating
blasts, and 80% had associated cytopenia. FAB subtypes differed
from initial diagnosis in the following 12 patients: seven had
different FAB subtypes, one had biphenotypic leukemia, two had
pre-B acute lymphoblastic leukemia (ALL), and two had sec-
ondary therapy–related AML with 11q23-specific breakpoint
that we chose to analyze with relapses. Cytogenetic analyses
were available in 75% of patients at relapse, and clonal evolution
was evident in most of them.
Reinduction Outcome
Ten children (9%) received palliative treatment only because
of a poor initial prognosis as a result of one or more of the
following conditions: age less than 1 year at initial diagnosis
(n ⫽ 5), extreme precocity of relapse (n ⫽ 5), poor initial
characteristics (n ⫽ 7), and unfavorable clinical context, including one extensive paraplegia caused by isolated paraspinal
relapse (n ⫽ 3). Supportive care was associated with the
following agents: low-dose cytarabine, oral etoposide, thioguanine, or idarubicin. All of the patients died within 11 months
(median, 2.5 months) after diagnosis of relapse.
Reinduction therapy was attempted in 96 children (91%) with
the aim of attaining a CR2 (Table 1). The patients had relapsed
either after chemotherapy (n ⫽ 77) or after allograft in CR1 (n ⫽
19). Sixty-seven patients were given one course, 18 patients
were given two courses, and 11 patients were given more than
three courses. BMT and donor lymphocyte infusion (DLI)
were performed as the first attempt at treatment in five and
four children in relapse, respectively. All-trans-retinoic acid
(ATRA) was administered to six children with acute promyelocytic leukemia (APL). Patients with isolated extramedullary relapses were as intensively re-treated at the time of
relapse as the other relapsed patients.
Downloaded from jco.ascopubs.org on August 22, 2014. For personal use only. No other uses without permission.
Copyright © 2003 American Society of Clinical Oncology. All rights reserved.
4380
ALADJIDI ET AL
Table 2.
Patients Who Relapsed After LAME 89/91 Protocol: Characteristics According to CR1 Treatment (N ⴝ 106)
Allograft (n ⫽ 21)
Characteristic
Sex
Female
Male
Age
Median, years
Range, years
⬍ 1 year
1-10 years
⬎ 10 years
Leukocytes, ⫻ 109/L*
Median
Range
FAB*
M1
M2
M3
M4
M4eo
M5
M6
Karyotype*
t(8,21)
inv(16)
11q23
t(15,17)
CR1 duration, months
Median
Range
ⱕ 6 months
6-12 months
13-24 months
⬎ 24 months
Chemotherapy (n ⫽ 80)
Autograft (n ⫽ 5)
No. of Patients
%
No. of Patients
%
No. of Patients
%
9
12
42
58
39
41
49
51
0
5
0
100
6.2
1-15.7
0
15
6
6.7
0.2-18.7
0
71
29
12
39
29
34
2.1-220
0.7
0.4-10.8
15
49
36
4
0
1
40.8
2.1-480
80
0
20
96
11-160
4
9
1
3
1
3
0
19
43
5
14
5
14
0
14
15
4
11
8
25
3
17
19
5
14
10
31
4
1
0
1
0
0
3
0
20
0
20
0
0
60
6
0
1
1
29
0
5
5
7
3
6
9
9
4
7
11
0
1
0
2
0
20
0
40
12
4-41
1
10
7
3
10
2-57
5
48
33
14
16
36
21
7
6
4-25
20
45
26
9
1
3
0
1
20
60
0
20
Abbreviations: LAME, Leucémie Aiquë Myéloı̈de Enfant; CR1, first complete remission; FAB, French-American-British.
*At initial diagnosis.
Fourteen toxic deaths were observed before CR2 attainment.
They occurred after a mean period of 75 days (range, 9 to 163
days) from the beginning of the first course and mainly in
patients exposed to secondary intense therapy. Long-lasting
marrow aplasia (median, 42 days; range, 10 to 187 days) was
the major cause of morbidity, resulting in severe infection
(WHO grade 3 and 4) in 40 children, with 15 cases of
septicemia, 10 cases of disseminated fungal infections, and 10
cases of pneumonia.
CR2 was attained in 68 patients, resulting in a CR2 rate of
71%. Fifty-one percent of children were in CR2 after a single
course of therapy, 67% were in CR2 after two courses of
therapy, and 71% were in CR2 after three or more courses of
therapy. Higher CR2 rates were obtained with standard-dose
cytarabine plus other agents (group 1), high-dose cytarabine plus
other agents (group 2), and fludarabine plus other agents (group
3). No patient who received a transplantation or DLI in relapse
attained CR2. Three children with AML M3 entered CR2 after
ATRA. Only one child given single-agent therapy attained CR2;
he was treated with low-dose cytarabine for an isolated cutaneous relapse.
Treatment Allocation in CR2
BMT was performed in 78% of children (n ⫽ 53) who
attained CR2 (50% of the total patient population). Matched
sibling-donor BMTs were performed by choice (n ⫽ 12, including two HLA-identical cord blood BMTs), but when no matched
sibling donor was available, autograft (n ⫽ 25; harvested in CR1
in 21 patients and in CR2 in four patients) or alternative-donor
BMT were performed (n ⫽ 16; 11 unrelated donors, three
HLA-mismatched familial donors, and two unrelated cord blood
cells). In nine children, this was a second transplantation (Table
3). Before transplantation, children were given a mean number
of two consolidation courses of chemotherapy. The median time
from CR2 to transplantation was 58 days (range, 10 to 203 days).
Various conditioning regimens were administered, and 32 of the
53 patients who received transplantation were given total-body
irradiation (TBI).
After matched sibling-donor BMT, eight patients developed
acute GVHD, two patients developed moderate nonextensive
chronic GVHD, and two patients developed veno-occlusive
disease. After alternative-donor BMT, nine patients developed
acute GVHD, seven patients developed chronic GVHD, includ-
Downloaded from jco.ascopubs.org on August 22, 2014. For personal use only. No other uses without permission.
Copyright © 2003 American Society of Clinical Oncology. All rights reserved.
4381
OUTCOME IN RELAPSED CHILDHOOD AML
Table 3.
Patients Who Relapsed After LAME 89/91 Protocol: Outcome
According to Post-CR1 Treatment (N ⴝ 106)
No. of Patients
Outcome
Palliative treatment
CR2 attainment if attempted
Attempted
Achieved
Aplasia duration before CR2, days
Median
Range
CR2 duration younger than CR1
Post-CR2 therapy
Matched sibling donor
Alternative donor
Autograft
Patients alive in CR
CR duration, months
Median
Range
Allograft
(n ⫽ 21)
Chemotherapy
(n ⫽ 80)
Autograft
(n ⫽ 5)
2
7
1
19
16
73
49
4
3
45
32-175
10
41
19-187
34
30
10-49
3
6
—
—
8
5
14
25
22
1
2
—
3
60
39-79
57
9-114
83
38-100
Abbreviations: LAME, Leucémie Aiquë Myéloı̈de Enfant; CR1, first complete
remission; CR2, second complete remission; CR, complete remission.
months). By June 1, 2001, 33 of 106 children were alive; 30 were
in CR2, and three were in CR3, with a median follow-up of 65
months after CR2 (range, 11 to 116 months). The 5-year DFS
rate for children in CR2 was 45% (Fig 2). Most of the patients
experienced a longer CR2 than CR1. Four children were alive
and well without BMT (one after DLI, two with ALL, and one
after high-dose cytarabine and interleukin-2). Among patients
who received transplantations, the 5-year DFS rate was not
significantly different among those who received matched sibling-donor allograft BMT (60%), autograft (47%), or phenoidentical allograft BMT (44%; Table 4).
After BMT, 29 patients were long-term survivors, with a
median follow-up of 65 months (range, 12 to 115 months). Eight
children were alive after two BMTs. Of the 53 patients who
underwent BMT of some form in CR2, 16 of 32 patients are alive
after a TBI-containing regimen (48%), and 11 of 19 patients are
alive after a non–TBI-containing regimen (57%; missing data,
n ⫽ 2). At the time of evaluation, long-term sequelae according
to the type of BMT are listed in Table 5. Three cured children
ing four severe and extensive cases, and two patients developed
Epstein-Barr virus–induced lymphoproliferative syndrome. Marrow failed to engraft in two children who required a second
BMT. Toxic morbidity after autograft consisted mainly of grade
3 to 4 infections in half of the children, as well as prolonged
marrow aplasia lasting 18 and 24 months in two patients. For the
entire cohort of grafted children, these procedures resulted in
eight toxic deaths (15%).
After CR2 attainment, seven patients were not grafted; four
previously allografted patients were given DLI, two patients with
ALL followed the high-risk ALL protocol, and one patient was
given high-dose cytarabine chemotherapy and interleukin-2.
Thirty-three children experienced a second relapse after CR2
(11 patients before BMT and 22 after BMT) after a median
period of 172 days (range, 61 to 477 days). Palliative treatment
was initiated in 24 children, with a median survival of 4 months
(range, 1 to 21 months). A third-line treatment was attempted
with nine children. Three of these patients had relapsed before
the programmed HLA-identical BMT; one patient entered third
CR (CR3), was grafted, and is alive and well at 36 months after
relapse; and two patients were grafted in relapse and died of
AML. Six patients relapsed after BMT in CR2; two are alive and
well in CR3 36 and 95 months after CR3. One of these patients
was a 2.6-year-old girl who relapsed from AML M2 12 months
after unrelated-donor BMT in CR2; she was treated successfully
with salvage high-dose cytarabine reinduction and continuous
maintenance therapy with low-dose and high-dose cytarabine for
2 years. The other patient was a 3-year-old boy who experienced
a CNS relapse of AML M5 with t(9,11) after autograft in CR2;
he was re-treated with salvage IT chemotherapy, three high-dose
cytarabine courses, and cerebral radiotherapy.
Global Results
The 5-year OS rate for all 106 patients was 33%, with a
median follow-up from relapse of 12 months (range, 1 to 115
Fig 2. Outcome of relapsed children after initial treatment in the Leucémie
Aiquë Myéloı̈de Enfant 89/91 protocol. (A) Five-year overall survival rate, 33%
(N ⴝ 106). (B) Five-year disease-free survival rate after second complete remission
attainment, 45% (n ⴝ 68).
Downloaded from jco.ascopubs.org on August 22, 2014. For personal use only. No other uses without permission.
Copyright © 2003 American Society of Clinical Oncology. All rights reserved.
4382
ALADJIDI ET AL
Table 4.
Treatment
Matched sibling BMT, n ⫽ 12
%
No.
Autografts, n ⫽ 25
%
No.
Alternative BMT, n ⫽ 16
%
No.
Outcome According to Treatment in CR2 (n ⴝ 53 grafted children)
Relapses
AML Deaths
Toxic Deaths
Alive in CR2
5-Year DFS
33
4
25
3
16
2
58
7
60
—
52
13
40
10
4
1
48
12
47
—
31
5
19
3
31
5
43
7
44
—
NOTE. P is not significant.
Abbreviations: CR2, second complete remission; AML, acute myeloid leukemia; DFS, disease-free survival; BMT, bone marrow
transplantation.
have presented with major consequences of their treatment; a
12-year-old boy has severe neurologic impairment of multifactorial origin, a 12.6-year-old boy has persistent aplastic anemia 6
years after autograft harvested in CR2, and a 9.7-year-old girl
has chronic hemolytic autoimmune anemia and panhypopituitarism 7 years after marrow unrelated-donor BMT.
Prognostic Factors Analysis
Multivariate analysis of 96 re-treated children showed that a CR1
duration of more than 12 months had a significant impact on CR2
attainment (hazard ratio [HR] ⫽ 2, P ⫽ .008), as did the type of
reinduction therapy (HR for groups A, B, and C ⫽ 1, 0.6, and 0.4,
respectively; P ⫽ .05; Table 6). The 2-year OS rate was significantly lower in patients with a CR1 duration of less than 12 months
(HR ⫽ 2.5, P ⫽ .002) and in patients who received maintenance
therapy after CR1 compared with chemotherapy without maintenance therapy and allograft BMT (HR ⫽ 3 v 1.5 v 1, respectively;
P ⫽ .0004; Fig 3). Three of the four re-treated children with isolated
Table 5.
Late Effects in 29 Alive Children in CR2 or CR3 According to the
Conditioning Regimen
Late Effects
Median age, years
Weight ⬍ normal
Height ⬍ normal
School retardation ⬎ 1 year
Growth hormone deficiency
Hypothyroidism
Hypogonadism
Bone marrow failure
Chronic GVHD
Myocardiopathy
Pulmonary disorders
Neurologic disorders
ENT disorders
Cataract
Posttransfusional hepatitis C
TBI-Containing
Regimen, CNS RT
(n ⫽ 19)*
Non–TBI-Containing
Regimen, No CNS RT
(n ⫽ 10)
11.9
4
6
7
5
4
6
2
0
0
2
2
2
2
1
16.6
1
1
1
0
0
1
0
1
0
1
1
0
0
0
NOTE. Median follow-up, 65 months.
Abbreviations: CR2, second complete remission; CR3, third complete remission;
TBI, total-body irradiation; RT, radiotherapy; GVHD, graft-versus-host disease; ENT,
ear, nose, and throat.
*Includes eight children after two BMTs and two children in CR3.
extramedullary relapses were alive in CR, but the extramedullary
site of relapse had no significant influence on the outcome. Sex, age
at initial diagnosis, leukocytosis, FAB subtype, and medullary
karyotype had no significant impact on CR2 attainment and survival. When the peripheral-blood blasts count at relapse was
included in the multivariate analysis, a high count of over 1.8 ⫻
109/L had a slight impact on survival (HR ⫽ 2.1, P ⫽ .03).
To avoid bias caused by the particular subgroups of 106
relapsing patients, the same analysis was conducted excluding
nine of the patients (five patients with APL who had not been
given first-line ATRA, two with 11q23 secondary AML, and two
with ALL). For all 97 AML patients who relapsed, the CR2 rate,
when remission was attempted, was 68%, the 5-year OS rate was
29%, and the 5-year DFS rate for children in CR2 was 39%; CR1
duration and postremission treatment in CR1 influenced the
outcome significantly.
DISCUSSION
In the LAME 89/91 protocol, the 5-year OS rate in relapsing
children was 33%, and the DFS rate was 45%, indicating that
recovery from relapsed AML is not an unrealistic goal. There are
few reports describing outcome in unselected relapsed AML in
children after similar intensive and effective front-line therapy.
Our policy was to offer a curative strategy to most patients. Our
results are to be compared with those recently reported by the
German and British groups,10,11 which have provided similar
CR2 rates in re-treated patients (71% v 51% and 69%, respectively) and percentages of grafts in CR2 (79% v 83% and 72%,
respectively). We speculate that the low rate of purely palliative
treatments in our experience (9% v 20% and 27%, respectively)
may have contributed to a better OS rate (33% at 5 years v 21%
and 24% at 2 years, respectively).10,11
Lethal toxicity was high in these intensively treated children,
reaching 9% after the first reinduction course and 15% after BMT.
The toxic mortality rate for re-treatment of relapsing AML patients
with intensive therapies is between 7% and 50%, with the highest
figures in transplantation patients.9 Long-lasting marrow aplasia and
the high percentage of fungal infections in our study led us to
recommend laminar flow-air isolation for most reinduction courses
and strong preventive measures against Candida infection. Potentially serious pneumonia was observed in 10% to 30% of children
Downloaded from jco.ascopubs.org on August 22, 2014. For personal use only. No other uses without permission.
Copyright © 2003 American Society of Clinical Oncology. All rights reserved.
4383
OUTCOME IN RELAPSED CHILDHOOD AML
Table 6.
Univariate and Multivariate Analysis of Prognostic Factors in 96 Relapsing AML Children
Univariate
Analysis
Factor
Age
⬍ 1 year
⬎ 1 year
Leukocytosis*
⬍ 50 ⫻ 109/L
⬎ 50 ⫻ 109/L
FAB subtypes*
M1 ⫹ M2
M3
M4 ⫹ M5
Karyotype*
Favorable
Others
Post-CR1 therapy
Allograft
Chemotherapy
Chemotherapy ⫹ maintenance therapy
CR1 duration
⬍ 12 months
⬎ 12 months
Reinduction therapy†
Group A
Group B
Group C
No. of Patients
% CR2 Attainment
HR
P
Multivariate
Analysis
HR
.23
11
85
58
75
1.7
1
60
36
78
65
1
1.4
39
6
51
78
100
66
1
0.4
1.3
19
47
30
83
77
60
1
1
1.6
54
42
63
85
2
1
54
25
16
89
56
47
1
2
3
18
43
2.1
1
43
36
1
1.3
53
83
25
1
0.4
2.2
1.1
58
33
1
2.4
.08
58
50
13
1
1
2.5
.02
26
59
2.3
1
.002
51
28
25
1
1.6
2
.08
1
0.5
1.7
.11
1
1.7
.004
.002
1.5
1
3
.001
.05
1
1.7
2.5
.34
.005
.003
2
1
P
1.5
1
.003
.06
1.1
1
2
HR
.38
.94
1
1
P
Multivariate
Analysis
.04
.68
1
.19
1
1.4
HR
.36
1
1.3
.07
83
70
2-Year OS (%)
.76
1.1
1
.2
29
66
P
Univariate
Analysis
.002
2.5
1
.27
.82
1
1
1.2
NOTE. Multivariate analysis was performed on 95 patients.
Abbreviations: AML, acute myeloid leukemia; CR2, second complete remission; HR, hazard ratio; OS, overall survival; FAB, French-American-British; CR1, first complete
remission.
*At initial diagnosis. Favorable cytogenetics indicates t(8,21), inv(16), and t(15,17)
†Groups A, B, and C are described in Patients and Methods, under Methods.
with relapsed AML, which may be precipitated by high doses of
cytarabine and dysregulation of fluid balance.12
There are no data in the medical literature describing the
long-term sequelae of AML in CR2. Marrow transplantation was
widely performed in patients who attained CR2 in our study
(78%), even after previous allograft in CR1. Moderate to severe
long-term toxicity was observed in half of the patients alive after
BMT. Among these patients highly treated with either a
second BMT or alternative-donor BMT or a TBI-containing
regimen, numerous aggressive procedures were implicated.
Long-term consequences of treatment on the heart, eyes, and
endocrine or pulmonary systems, mainly in cases of TBI use,
require attentive monitoring.13,14
In accordance with previously published data, the duration of
CR1 was the main independent prognostic factor in relapsed
AML. Previous Berlin-Frankfurt-Munster studies found that the
2-year survival rate was 10% for early relapses before 18 months
and 40% for relapses 18 months after CR1.11 Age less than 1
year at initial diagnosis was a strong predictor of poor outcome
after relapse because of multifactorial etiologies, including
treatment with palliative therapy only. The outcome of extramedullary and medullary relapses did not differ significantly.
It seems that the same intensive strategy of CR2 reinduction and
marrow transplantation should be recommended.15 There is
widespread agreement that a better prognosis is associated with
t(8,21), t(15,17), and inv(16) at initial diagnosis. These karyotypes form the basis of front-line risk-directed therapy in most
multicentric trials.6,16,17 These differences in outcome persist in
case of relapse.
The reasons explaining why our patients who relapsed after
HLA-identical allograft in CR1 had a slightly better 5-year OS
rate remain unclear but may include heterogeneity in primary
risk factors and statistical biases. Maintenance therapy after CR1
had a strongly unfavorable impact on survival. In a study of the
first 268 enrolled children previously reported by Perel et al,8
survival and DFS were shown to be worse in children who had
been randomly assigned to maintenance therapy. This unfavorable outcome, without increasing mortality or relapse
rates, was related to the increased rate of failure to attain a
CR2. The Children’s Cancer Group had previously reported
the inferiority of maintenance therapy in AML.18 The mechanisms of the deleterious effects of maintenance therapy are
unclear, but prolonged exposure to low-dose cytarabine may
increase mdr gene product expression and the incidence of
refractory diseases.19
CR2 was obtained in 71% of patients when it was attempted,
and this should be compared with the results of other European
groups.10,11 The best results in our experience were obtained
with the use of high or intermediate doses of cytarabine plus
anthracycline or the fludarabine, cytarabine, and granulocyte
Downloaded from jco.ascopubs.org on August 22, 2014. For personal use only. No other uses without permission.
Copyright © 2003 American Society of Clinical Oncology. All rights reserved.
4384
ALADJIDI ET AL
Fig 3. (A) Impact of post-first complete remission (CR1) therapy on outcome in
children with relapsed acute myeloid leukemia (5-year overall survival rates:
allograft, 52%; chemotherapy with no maintenance, 40%; chemotherapy with
maintenance, 12%; n ⴝ 96; P ⴝ .002). (B) Impact of CR1 duration on outcome of
children with acute myeloid leukemia (5-year overall survival rates: CR1 more than
12 months, 54%; CR1 less than 12 months, 24%; n ⴝ 96; P ⴝ .001).
colony-stimulating factor regimen, but the diversity of regimens
administered and the retrospective nonrandomized nature of our
study did not enable us to evidence the superiority of any one
regimen in particular or to assess differences in toxicities.
Prospective studies of reinduction therapy for relapsed childhood
AML are rare. Such children are often assimilated with refractory children in small retrospective series. Clearly effective
reinduction regimens most commonly combine cytarabine with
anthracycline or fludarabine. CR2 rates of 40% to 75% are
usually attained.9 The synergic efficacy of the cytarabine, fludarabine, idarubicin, and granulocyte colony-stimulating factor
regimen was recently reported by several centers.20-23 High-dose
liposomal daunorubicin has given promising preliminary results;
in a recent phase I to II study of adult relapsing AML, it
produced a 29% CR2 rate when used alone.24 Because of the
extremely small number of AML relapsing children (30 to 50
children a year in main European countries), there is clearly a
need for large, multicentric, pediatric trials to address these
important issues of activity and toxicity.
Seventy-eight percent of CR2 children received transplantation. On the whole, an allograft from a matched sibling donor
remains the best treatment after CR2 attainment, providing a
5-year DFS of 60%. It is worth noting that this includes a high
proportion of patients previously treated with HLA-identical
BMT in CR1. In a large series of second allografts for leukemia
including both adults and children, the risks of relapse and
transplantation-related mortality were as high as 44% ⫾ 12%
and 45% ⫾ 9%, respectively.25 In accordance with our study,
much lower rates were reported in children and may be considered acceptable for these high-risk patients. However, it should
be noted that more than half of the patients will experience some
late sequelae.
When a matched sibling donor was not available, autograft or
alternative-donor BMT resulted in similar DFS rates. The high
rate of transplantation-related mortality in alternative-donor
BMT (31%) was counterbalanced by a low relapse rate (31%).
Thus, the DFS rate was still at 44%, possibly because of an
enhanced antileukemic effect. Significant improvements are to
be expected with strict selection of alternative sources of stem
cells and optimal control of GVHD. Furthermore, recent studies
conducted on AML cells suggest that blocking the Fas/Fas
ligand pathway could be used to prevent GVHD without impairing the graft-versus-leukemia effect.26
On the other hand, autologous BMT seems to be a reasonable
choice in AML, providing the lowest rate of transplantationrelated mortality (4%) and a 5-year DFS rate of 47%. Other
series of selected children reported a DFS rate of approximately
40% after autograft in CR2 for childhood AML.27 These findings
further argue for a role of dose-intensity in AML relapse.
However, residual leukemia cells in the graft contribute to
posttransplantation relapse. Treatment of the graft to decrease
the number of infused leukemia cells is controversial but was
reported in AML to improve survival, with similar transplantation-related mortality and lower risks of treatment failure.28 An
alternative procedure, as reported in our study, is to perform the
marrow harvest after CR1 (in vivo purging). The monitoring of
residual disease after CR1 may contribute to the evaluation of
this approach.
Cure of relapsed AML of children is a realistic goal even after
aggressive first-line therapy. When CR2 is attained after initial
intensive reinduction chemotherapy, prompt intensification with
allograft or autograft provides similar DFS. Thus, for most patients,
we recommend a therapeutic strategy including intensive reinduction and megadose intensification. In unfavorable cases, the CR2
Downloaded from jco.ascopubs.org on August 22, 2014. For personal use only. No other uses without permission.
Copyright © 2003 American Society of Clinical Oncology. All rights reserved.
4385
OUTCOME IN RELAPSED CHILDHOOD AML
rate remains low with the use of conventional intensive chemotherapy. At the present time, therapeutic options may include further
intensified schedules but also modulation of multidrug resistance
mechanisms,29 immunomodulatory approaches,30 or anti-CD33
monoclonal antibodies.31 For survivors, the increased intensity and
duration of treatment certainly result in long-term morbid consequences. The risk of impaired quality of life may have an impact on
medical decisions and warrants further studies.
APPENDIX
The appendix is included in the full-text version of this article, available
on-line at www.jco.org. It is not included in the PDF (via Adobe® Acrobat
Reader®) version.
AUTHORS’ DISCLOSURES OF POTENTIAL
CONFLICTS OF INTEREST
The authors indicated no potential conflicts of interest.
REFERENCES
1. Wells RJ, Woods WG, Buckley JD, et al: Treatment of newly
diagnosed children and adolescents with acute myeloid leukemia: A Children’s Cancer Group Study. J Clin Oncol 12:2367-2377, 1994
2. Behar C, Suciu S, Benoit Y, et al: Mitoxantrone-containing regimen for
treatment of childhood acute leukemia (AML) and analysis of prognosis
factors: Results of the EORTC Children Leukemia Cooperative Study 58872.
Med Pediatr Oncol 26:173-179, 1996
3. Michel G, Baruchel A, Tabone MD, et al: Allogeneic bone marrow
transplantation vs aggressive post-remission chemotherapy for children with
acute myeloblastic leukemia in first complete remission: A prospective study
from the French Society of Pediatric Hematology and Immunology (SHIP).
Bone Marrow Transplant 17:191-196, 1996
4. Ravindranath Y, Yaeger AM, Chang MN, et al: Autologous bone
marrow transplantation versus intensive consolidation chemotherapy for
acute myeloid leukemia in childhood: Pediatric Oncology Group. N Engl
J Med 334:1428-1434, 1996
5. Stevens RF, Hann IM, Wheatley K, et al: Marked improvement in
outcome with chemotherapy alone in pediatric acute myeloid leukemia:
Results of the United Kingdom Medical Research Council’s 10th AML trial
(MRC AML 10)-Adult and Childhood Leukemia Working Parties of the
Medical Research Council. Br J Haematol 101:130-140, 1998
6. Creutzig U, Ritter J, Zimmermann M, et al: Improved treatment results
in high-risk pediatric acute myeloid leukemia after intensification with
high-dose cytarabine and mitoxantrone: Results of Study Acute Myeloid
Leukemia-Berlin-Frankfurt-Münster 93. J Clin Oncol 19:2705-2713, 2001
7. Woods WG, Neudorf S, Gold S, et al: A comparison of allogeneic bone
marrow transplantation, autologous bone marrow transplantation, and aggressive chemotherapy in children with acute myeloid leukemia in remission.
Blood 97:56-62, 2001
8. Perel Y, Auvrignon A, Leblanc T, et al: 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—Leucámie Aiquë Myéloı̈de Enfant. J Clin Oncol 20:2774-2782, 2002
9. Webb DK: Management of relapsed acute myeloid leukemia. Br J
Haematol 106:851-859, 1999
10. Webb KH, Wheatley K, Harrison G, et al: Outcome for children with
relapsed acute myeloid leukemia following initial therapy in the Medical
Research Council (MRC) AML 10 Trial. Leukemia 13:25-31, 1999
11. Stahnke K, Boos J, Bender-Götze C, et al: Duration of first remission
predicts remission rates and long term survival in children with relapsed
acute myelogenous leukemia. Leukemia 12:1534-1538, 1998
12. Shearer P, Katz J, Bozeman P, et al: Pulmonary insufficiency complicating therapy with high dose cytosine arabinoside in five pediatric patients with
relapsed acute myelogenous leukemia. Cancer 74:1953-1958, 1994
13. Mertens AC, Yasui Y, Neglia JP, et al: Late mortality experience in
five-year survivors of childhood adolescent after cancer: The Childhood
Cancer Survivor Study. J Clin Oncol 19:3161-3162, 2001
14. Michel G, Socie G, Gebhard F, et al: Late effects of allogeneic bone
marrow transplantation for children with acute leukemia in first complete
remission: Impact of conditioning regimen without total-body irradiation—
Report from the Societe Française de Greffe de Moelle. J Clin Oncol
15:2238-2246, 1997
15. Ginsberg JP, Lange BJ: Isolated extramedullary relapse in AML: A
retrospective analysis. J Pediatr Hematol Oncol 20:389, 1998 (abstr)
16. Raimondi SC, Chang MN, Ravindranath Y, et al: Chromosomal
abnormalities in 478 children with acute myeloid leukemia: Clinical characteristics and treatment outcome in a cooperative Pediatric Oncology Group
study-POG 8821. Blood 94:3707-3716, 1999
17. Webb DK, Harrison G, Stevens RF, et al: Relationships between age
at diagnosis, clinical features, and outcome of therapy in children treated in
the Medical Research Council AML 10 and 12 trials for acute myeloid
leukemia. Blood 98:1714-1720, 2001
18. Wells RJ, Woods WJ, Lampkin BC, 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 11:538-545, 1993
19. Hu XF, Slater A, Kantharidis P, et al: Altered multidrug resistance
phenotype caused by anthracycline analogues and cytosine arabinoside in
myeloid leukemia. Blood 93:4086-4095, 1999
20. Fleischhack G, Hasan C, Graf N, et al: IDA-FLAG (idarubicin, fludarabine, cytarabine, G-CSF), an effective remission-induction therapy for poorprognosis AML of childhood prior to allogeneic or autologous bone marrow
transplantation: Experience of a phase II trial. Br J Haematol 102:647-655, 1998
21. McCarthy AJ, Pitcher LA, Hann IM, et al: FLAG (fludarabine,
high-dose cytarabine and G-CSF) for refractory and high-risk relapsed acute
leukemia in children. Med Pediatr Oncol 32:411-415, 1999
22. Yalman N, Sarper N, Devecioglu O, et al: Fludarabine, cytarabine,
G-CSF and idarubicin (FLAG-IDA) for the treatment of relapsed or poor risk
childhood acute leukemia. Turk J Pediatr 42:198-204, 2000
23. Luczynski W, Muszynska-Roslan K, Kravczuk-Rybak M, et al:
Results of IDA-FLAG program in the treatment of recurrent acute myeloblastic leukemia: Preliminary report. Med Sci Monit 7:125-129, 2001
24. Cortes J, Estey E, O’Brien S, et al: High-dose liposomal daunorubicin
and high-dose cytarabine combination in patients with refractory or relapsed
acute myelogenous leukemia. Cancer 92:7-14, 2001
25. Michallet M, Tanguy ML, Socié G, et al: Second allogeneic hematopoietic stem cell transplantation in relapsed acute and chronic leukemias for
patients who underwent a first allogeneic bone marrow transplantation: A
survey of the Société Française de Greffe de Moelle. Br J Haematol
108:400-407, 2000
26. Hsieh MH, Korngold AR: Differential use of FasL- and perforinmediated cytolytic mechanisms by T-cell subsets involved in graft-versusmyeloid leukemia responses. Blood 96:1047-1055, 2000
27. Vignetti M, Rondelli R, Locatelli F, et al: Autologous bone marrow
transplantation in children with acute myeloid leukemia: Report from the
Italian National Pediatric Registry (AIEOP-BMT). Bone Marrow Transplant
18:59-62, 1996 (suppl 2)
28. Miller CB, Rowlings PA, Zhang MJ, et al: The effect of graft purging
with 4-hydroperoxycyclophosphamide in autologous bone marrow transplantation for acute myelogenous leukemia. Exp Hematol 29:1336-1346, 2001
29. Dahl GV, Lacayo NJ, Brophy N, et al: Mitoxantrone, etoposide and
cyclosporine therapy in pediatric patients with recurrent or refractory acute
myeloid leukemia. J Clin Oncol 18:1867-1875, 2000
30. Sievers EL, Lange BJ, Sondel PM, et al: Children’s Cancer Group
trials of interleukin-2 therapy to prevent relapse of acute myelogenous
leukemia. Cancer J Sci Am 6:S39-S44, 2000 (suppl 1)
31. Larson RA: Current use and future development of gemtuzumab
ozogamicin. Semin Hematol 38:24-31, 2001
Downloaded from jco.ascopubs.org on August 22, 2014. For personal use only. No other uses without permission.
Copyright © 2003 American Society of Clinical Oncology. All rights reserved.