• myeloid leukaemia;
  • childhood leukaemia;
  • clinical trial;
  • AML;
  • therapy


  1. Top of page
  2. Summary
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Author contributions
  8. Disclosure
  9. References
  10. Appendix

The Medical Research Council Acute Myeloid Leukaemia 12 (MRC AML12) trial (children) addressed the optimal anthracenedione/anthracycline in induction and the optimal number of courses of consolidation chemotherapy. 504 children (<16 years) with AML were randomized between mitoxantrone/cytarabine/etoposide or daunorubicin/cytarabine/etoposide as induction chemotherapy and 270 entered a second randomization between a total of four or five courses of treatment. Ten-year event-free (EFS) and overall survival (OS) was 54% and 63% respectively; the relapse rate was 35%. There was no difference in complete remission rate between the induction regimens, but there was a benefit for mitoxantrone with regard to relapse rate [32% vs. 39%; Hazard ratio (HR) 0·73; 95% confidence interval (CI) 0·54, 1·00] and disease-free survival (DFS; 63% vs. 55%; HR 0·72; 95% CI 0·54, 0·96). However, this did not translate into a better EFS or OS (HR 0·84; 95% CI 0·63, 1·12). Results of the second randomization did not show a survival benefit for a fifth course of treatment (HR 1·01; 95% CI 0·63, 1·62), suggesting a ceiling of benefit for conventional chemotherapy and demonstrating the need for new agents. EFS was superior compared to the preceding trial AML10, partly due to fewer deaths in remission, highlighting the importance of supportive care.

The combination of increasingly intensive anthracycline- and cytosine-based chemotherapy and advances in supportive care have lead to a dramatic improvement in the survival of children with acute myeloid leukaemia (AML) (Stevens et al, 1998; Creutzig et al, 2005a; Lie et al, 2005). The AML12 trial aimed to further the improvement of outcome in AML and followed the Medical Research Council (MRC) tradition of treating patients with AML between 0 and 59 years within one study. We report the final results of the AML12 trial in children, which confirm and expand preliminary reports (Gibson et al, 2005). Although the whole trial has been published (Burnett et al, 2010) minimal data were included for children, who comprised one quarter of the patients. Key issues specific to children are presented in this report. The first objective of AML12 was to compare the efficacy and toxicity of mitoxantrone with daunorubicin, both given in combination with cytosine and etoposide (MAE versus ADE). ADE was the preferred regimen from the preceding trial, AML10 (Stevens et al, 1998). The second objective addressed the number of courses. The significant and similar reduction in relapse rate (RR) observed for both autologous (A-SCT) and allogeneic (allo-SCT) stem cell transplantation in AML10 suggested that the benefit for transplantation might be one of additional treatment rather than a specific effect of transplantation. AML12 tested this hypothesis by randomizing patients to four or five courses of treatment in total. The additional course of treatment for children was high dose (HD) cytosine with asparaginase (CLASP) because of acceptable toxicity in children and absence of additional anthracycline exposure (Lange et al, 2008).

Risk group stratification was based on karyotype and response to the first course of treatment (Wheatley et al, 1999):

  • 1
     Good risk – any patient with favourable genetic abnormalities – t(8;21), inv (16), t(15;17) or French-American-British (FAB) classification type M3 morphology – irrespective of bone marrow status after Course 1 or the presence of other genetic abnormalities.
  • 2
     Standard risk – any patient not in either good or poor risk groups.
  • 3
     Poor risk – any patient with more than 15% blasts in the bone marrow after Course 1 or with adverse genetic abnormalities –−5, −7, del(5q), abn(3q), complex (≥5 abnormalities) – and without favourable genetic abnormalities.

Allo-SCT in first remission was restricted to standard and poor risk patients with matched sibling donors. A-SCT was abandoned because of its lack of survival advantage and its substantial morbidity in children.

Patients and methods

  1. Top of page
  2. Summary
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Author contributions
  8. Disclosure
  9. References
  10. Appendix


Five hundred and sixty-four children with newly diagnosed AML were recruited from 30 centres in the United Kingdom (MRC, n = 445), Eire (MRC, n = 44) the Netherlands (Dutch Childhood Oncology Group, n = 63) and New Zealand (n = 12) between April 1995 and May 2002. The upper age limit for inclusion was the 16th birthday. Children with de novo AML, AML of Down syndrome (DS), acute promyelocytic leukaemia (APL), secondary AML, or aggressive myelodysplastic syndrome [MDS; refractory anaemia with excess blasts (RAEB), RAEB in transformation (RAEB-t)] for whom intensive-type therapy was considered appropriate were eligible; patients who had had previous treatment for any form of leukaemia, were in blast transformation of chronic myeloid leukaemia, had concurrent malignancy or were considered unfit for intensive chemotherapy were ineligible. Five hundred and twenty-nine patients entered the trial before the start of induction chemotherapy; the remainder entering prior to the consolidation phase. The trial was ethically approved centrally and by the ethical committees of the participating institutions and all patients or parents gave written informed consent.


Morphological diagnosis, FAB typing, immunophenotyping and cytogenetics were performed locally and centrally reviewed. Cytogenetics was coordinated by the UK Leukaemia Research Cytogenetics Group and the Dutch Working Group on Cancer Genetics and Cytogenetics. Patients were classified into cytogenetic risk groups: favourable, intermediate and adverse [Table I and as previously described (Grimwade et al, 1999)]. Central nervous system (CNS) disease was defined by the presence of >5 × 106/l blasts in a cerebrospinal fluid (CSF) cytospin or the presence of unexplained cranial nerve palsy.

Table I.   Characteristics of Paediatric patients entered into MRC AML12.
 All patients in trialPatients in MAE versus ADE randomizationPatients in 5 vs. 4 courses randomization
MAEADE5 courses4 courses
  1. WCC, white cell count; CNS, central nervous system; APL, acute promyelocytic leukaemia: FAB, French-American-British; AUL, acute undifferentiated leukaemia; AML, acute myeloid leukaemia; AMML, acute myelomonocytic leukaemia; AMoL, acute monocytic leukaemia; ALL, acute lymphoblastic leukaemia; RAEB-t, refractory anaemia with excess blasts in transformation.

  2. Numbers are given as n (%) Unknown values are not listed. Percentages exclude unknown values from the denominator.

Total number of patients564 (100)251253135135
Gender (male)302 (54)145 (58)124 (49)64 (47)78 (58)
Age (years)
 0–1130 (23)57 (23)59 (23)34 (25)33 (24)
 2–9252 (45)115 (46)107 (42)70 (52)59 (44)
 10–16182 (32)79 (31)87 (34)31 (23)43 (32)
WCC (×109/l)
 <10227 (41)100 (41)102 (41)46 (35)50 (38)
 10 – <99·9231 (42)102 (42)111 (45)65 (50)63 (47)
 ≥10091 (17)41 (17)34 (14)21 (16)20 (15)
CNS leukaemia (yes)32 (6)17 (7)10 (4)7 (5)12 (9)
Down syndrome (yes)33 (6)12 (5)15 (60 (0)1 (1)
 Favourable118 (24)61 (27)47 (23)36 (30)33 (28)
  t(8;21)53 (11)29 (13)19 (9)13 (11)16 (14)
  inv(16)29 (6)16 (7)13 (6)12 (10)8 (7)
  t(15;17) (APL)36 (7)16 (7)15 (7)11 (9)9 (8)
 Intermediate292 (60)130 (58)129 (62)73 (60)68 (59)
 Adverse74 (15)32 (14)32 (15)13 (11)15 (13)
  Monosomy 714 (3)6 (3)6 (3)0 (0)1 (1)
  del (5q)2 (<1)0 (0)2 (1)1 (1)0 (0)
  abn (3q)10 (2)7 (3)2 (1)1 (1)2 (2)
  Complex karyotype(≥5 abnormalities)36 (7)14 (6)18 (9)7 (6)10 (9)
  >1 adverse abnormality12 (2)5 (2)4 (2)4 (3)2 (2)
FAB type
 AUL23 (4)8 (3)14 (6)3 (2)2 (2)
 AML (undifferentiated)58 (11)27 (11)25 (10)17 (13)7 (5)
 AML (differentiated)125 (23)54 (22)59 (24)31 (23)35 (27)
 APL49 (9)20 (8)22 (9)14 (11)14 (11)
 AMML88 (16)40 (17)38 (15)23 (17)20 (16)
 AMoL116 (21)53 (22)51 (21)31 (23)33 (26)
 Erythroleukaemia12 (2)5 (25 (21 (1)1 (1)
 Megakaryocytic33 (6)15 (6)15 (6)7 (5)6 (5)
 Bilineage1 (<1)1 (<1)0 (0)0 (0)0 (0)
 ALL3 (1)0 (0)3 (1)0 (0)1 (1)
 RAEB-t35 (6)18 (7)15 (6)5 (4)10 (8)
 Other1 (<1)0 (0)1 (<1)0 (0)0 (0)


Details of therapy are shown in Fig 1. Induction chemotherapy consisted of two courses of either ADE or MAE. Children progressed to Course 2 on recovery of the neutrophil count to 1·0 × 109/l and the platelet count to 100 × 109/l. Patients not in complete remission (CR) after two courses of treatment were considered to have failed treatment and were not eligible for further randomization within AML12, but could receive alternative therapy and continued to be followed up within the trial.


Figure 1.  Treatment schema for MRC AML12. Rand, randomization; CR, complete remission; SCT, stem cell transplantation.

Download figure to PowerPoint

Following Course 3 (MACE – amsacrine, cytosine, etoposide) patients were randomized to receive either one or two more courses of therapy (i.e. four or five courses in total). For patients allocated to five courses, the fourth course was CLASP. For both groups the final course of therapy was either MidaC (mitoxantrone, cytarabine) or allo-SCT, depending on the patient’s risk group and on the availability of a matched sibling donor. Allo-SCT was restricted to standard and poor risk patients with a matched sibling donor. Children <1 year of age had all chemotherapy doses reduced by 25%. Details of number of patients randomized into each comparison are given in the CONSORT Diagram (Fig 2).


Figure 2.  CONSORT diagram. CR, complete remission.

Download figure to PowerPoint

Central nervous system-directed therapy was by a total of three courses of ‘triple’ intrathecal therapy (methotrexate, cytarabine and hydrocortisone) at age-adjusted doses, one after each of the first three courses of chemotherapy. Children with CNS disease at diagnosis received two courses of ‘triple’ intrathecal therapy each week until the CSF was clear plus two further courses. A minimum of six courses was given in the three-week period following diagnosis. This was followed by monthly courses of triple intrathecal chemotherapy until after the final course of systemic chemotherapy had been completed. It was recommended that children aged 2 years or over with CNS disease received cranial irradiation (CRT, 2400 cGy) after the final course of chemotherapy, except for those receiving total body irradiation as part of SCT conditioning. Children <2 years of age were not eligible for CRT.


Endpoints used for this trial included complete remission (CR), partial remission (PR), overall survival (OS), disease-free survival (DFS) and event-free survival (EFS) as previously defined (Gibson et al, 2005). CR was defined as a bone marrow with <5% leukaemic cells and evidence of regeneration of normal haemopoietic cells. Neutrophil and platelet parameters were not included in the definition. For those achieving remission, relapse risk (RR) and death in CR (DCR) were calculated as the cumulative probability of relapse/death in remission respectively, starting from date of CR. For the consolidation randomization, OS, DFS, RR and DCR were measured from the date of the consolidation randomization. Toxicity was assessed according to the World Health Organization (WHO) toxicity criteria (WHO, 1979).

Statistical methods

Primary analysis of all randomizations was performed on an ‘intent to treat’ basis. Time-to-event data were compared using the logrank test, and medians/survival percentages by Kaplan–Meier estimates. For DCR and RR, survival probabilities were calculated by the method of cumulative incidence in the presence of competing risks (Kalbfleisch & Prentice, 1980). Survival probabilities are at 10 years, unless otherwise stated. Estimates of treatment effects are given with 95% confidence intervals (CIs), unless otherwise stated. Continuous data was analysed using either the Wilcoxon rank-sum test, or the two-sample T-test, as appropriate. Dichotomous data was compared using either the Chi-square test, or the Fisher exact test. As the paediatric part of the AML12 was not powered to give results on its own, presentation of results of randomizations will be focussed on point estimates with 95% CIs, rather than hypothesis testing.


  1. Top of page
  2. Summary
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Author contributions
  8. Disclosure
  9. References
  10. Appendix

Patient characteristics

Patient demographics are given in Table I.


For the MAE versus ADE randomization, 98% of patients in each arm started Course 1; for Course 2, compliance was 96% for MAE and 93% for ADE. In the consolidation randomization, compliance was 94% on the 4-course arm and 87% on the 5-course arm (Fig 2). Only 44 children underwent a matched sibling transplant in first CR; 18 unrelated and two haplo-identical transplants were performed as protocol violations.

Overall outcome

Of the patients registered before the start of treatment, 92% achieved remission; 73% of remissions occurred after Course 1 and 20% after Course 2. Failure to achieve remission was equally due to early deaths (4%) and resistant disease (4%). Surviving patients have been followed up for a median of 9·5 years, with >99% followed up for at least 2 years. The 10-year survival probabilities for patients registered before treatment were: DFS 59%, EFS 54%, OS 63%. 10-year relapse risk was 35%. The majority of relapses occurred within the first two years; RR at 1 year was 21% and 31% at 2 years. The site was specified in 167 of the 180 relapses: 89% isolated bone marrow, 7% isolated or combined CNS involvement, with the remainder at other sites, with or without bone marrow disease.

Patients with APL and patients with DS AML were eligible for AML12, but when excluded, 10-year survival probabilities were: DFS 56%, EFS 51%, OS 61% and RR 38%.

Induction randomization: ADE versus MAE

See Table II and Fig 3A for results of the ADE versus MAE randomization. CR rates were comparable between the arms. Twice as many children experienced induction death after MAE as with ADE, but the confidence interval was wide and consistent with no underlying difference. For RR and DFS, the point estimate of treatment effect suggested a benefit for MAE, but the confidence intervals would be compatible with anything from a negligible, through to a clear benefit for MAE. It was unclear whether there was a benefit for MAE with regard to EFS (HR 0·79, 95% CI 0·61, 1·03) and OS (HR 0·84, 95% CI 0·63, 1·12). Confidence intervals for these endpoints would be compatible with anything from a moderate benefit for patients receiving MAE, through to a small detriment for MAE.

Table II.   Results of the randomized comparisons.
Outcome measureMAE versus ADE randomization
MAE (= 251), %ADE (= 253), %OR/HR & 95% CIP-value
  1. CR, complete remission; DFS, disease-free survival; EFS, event-free survival; OS, overall survival.

  2. All time to event percentages are at 10 years. HRs and ORs are calculated such that a value <1·00 indicates a benefit for either MAE or for five courses.

Induction death631·78 (0·76, 4·2)0·2
Resistant disease440·91 (0·38, 2·2)1·0
CR rate90921·30 (0·70, 2·4)0·4
Death in CR560·68 (0·32, 1·44)0·3
Relapse Risk32390·73 (0·54, 1·00)0·05
DFS63550·72 (0·54, 0·96)0·03
EFS57510·79 (0·61, 1·03)0·08
OS65610·84 (0·63, 1·12)0·2
 5 versus 4 courses randomization
5 courses (= 135), %4 Courses (= 135), %HR & 95% CIP-value
Death in CR120·64 (0·11, 3·7)0·6
Relapse risk36361·01 (0·67, 1·50)1·0
DFS63620·98 (0·67, 1·45)0·9
OS74741·01 (0·63, 1·62)1·0

Figure 3.  Overall Survival by (A) induction randomization and (B) consolidation randomization.

Download figure to PowerPoint

As a secondary analysis, data was reanalysed excluding patients with DS AML and APL. For all the above endpoints, point estimates in this analysis were slightly less favourable toward MAE, and the associated confidence intervals all included the possibility of a detrimental effect of MAE.

Response rates following the first course of chemotherapy were calculated to investigate whether mitozantrone or daunorubicin might result in superior early disease clearance. On the MAE arm they were: 67% CR, 16% PR, 7% RD, 5% Deaths, 4% Unknown. On the ADE arm they were: 63% CR, 15% PR, 10% RD, 3% Deaths, 9% Unknown.

Consolidation randomization: five versus four courses of treatment in total

There were no significant differences between the arms (Table II, Fig 3B), but confidence intervals were wide. There was no interaction between the two randomizations (test for interaction P = 0·7 based on OS). Due to low compliance in the 5-course arm, a secondary per protocol analysis was also carried out. This showed no significant differences to the intention-to-treat analysis.

Prognostic factors

Analysis of prognostic factors was carried out on patients who entered the trial before treatment. Multivariate analysis of OS showed that age, white cell count (WCC) and cytogenetics were all significant predictors of outcome (Table III contains the corresponding univariate results). Logistic regression showed that cytogenetics and WCC also had prognostic value with regard to CR rates (Table III). The CR rate was similar throughout the age range. Children aged 0–1 experienced over twice the rate of induction death compared with older children (8% vs. 3%, 2= 0·06), but no resistant disease (0% vs. 5%, 2P = 0·007). As expected, WCC was a significant prognostic factor. Patients with higher counts at the start of treatment had a lower CR rate, a higher RR and a worse OS.

Table III.   Outcome by prognostic/risk factors.
FactorNumber of patientsCR rateInduction deathResistant diseaseRR at 10 yearsOS at 10 years
  1. –, not applicable; n/a, not available; WCC, white cell count; CR, complete remission; RR, relapse risk; OS, overall survival.

  2. Notes: This table excludes patients who registered after starting their treatment.

  3. Excludes patients with missing values for the relevant variables.

Age (years)
P-value 0·30·60·030·50·06
WCC count (×109/l)
P-value 0·020·10·09<0·001<0·001
 Other favourable8098%2%0%17%84%
 Other intermediate17291%5%3%40%60%
P-value 0·0010·9<0·001<0·001<0·001
Outcome of course 1
 Complete remission34030%76%
 Partial remission8741%55%
 Resistant disease40n/a25%
P-value    <0·001<0·001
Risk group
 Good risk11820%83%
 Standard risk21235%70%
 Poor risk9350%39%
P-value    <0·001<0·001


The cytogenetic results of children treated on this trial and AML10 have been published in detail elsewhere (Harrison et al, 2010). Outcome of children with adverse risk abnormalities, although relatively poor, was significantly better than for adults in the same adverse risk group: for OS, HR for adverse cytogenetics in children was 1·9 vs. 3·2 in adults, test for heterogeneity P = 0·02; for EFS, HR was 1·5 vs. 3·3, heterogeneity P < 0·001.

Risk group stratification

The MRC AML10-derived risk group stratification retained its prognostic significance in MRC AML12. 118 patients had good, 212 standard and 93 poor risk disease. See Table III for 10-year OS and RR by risk group, and by outcome of Course 1.

Acute promyelocytic leukaemia (APL)–FAB type M3

Acute promyelocytic leukaemia represented 9% of children entered into MRC AML12. All 49 children received all-trans retinoic acid (ATRA) at a dose of 45 mg/m2 during induction chemotherapy until remission was achieved, or for a maximum of 60 d. Only two patients failed to achieve remission due to early death: haemorrhage and infection. The 10-year RR for these patients was 25%; DCR was 9% and 10-year OS was 73%.

Down syndrome (DS)

Thirty-three (6%) of children entered into MRC AML12 had DS. These children were allocated four courses of chemotherapy only and were not eligible for SCT, but were eligible for the induction randomization. Details and outcome have been published (Rao et al, 2006).


Treatment-related mortality (TRM) was significant at 10% for all trial entrants. 4% (22) of children died during induction and 6% (32) in first CR. Induction deaths included five children who died prior to starting therapy (haemorrhage, n = 2) or within days of starting treatment from significant complications at presentation (infection, n = 2, intracranial haemorrhage/disseminated intravascular coagulopathy, n = 1). Deaths in the remaining 17 patients were due to infection (n = 11), haemorrhage (n = 4); the cause was not given in two children.

Table IV gives details of toxicity and resource usage by course. Whilst there was little difference between the treatments in terms of the WHO toxicity scores, MAE had a somewhat higher supportive care resource usage in Course 1, and a markedly higher resource usage in Course 2.

Table IV.   Toxicity and resource usage by courses 1–2.
MeasureCourse 1: MAECourse 1: ADEP-valueCourse 2: MAECourse 2: ADEP-valueCourse 3: MACECourse 4: CLASPCourse 4: MidACCourse 5: MidAC
  1. Data is based on intent-treat population.

  2. Units of resource usage, figures are medians, P values are from Wilcoxon rank sum test.

  3. Days to recovery – figures are Kaplan–Meier median estimates, P-values are from logrank test.

  4. WHO toxicity – figures are % grade 3/4 toxicity, P-values are from Wilcoxon rank sum test.

  5. *There was a high proportion of missing data for these variables.

  6. †Cardiac function toxicity is defined according to World Health Organization (WHO) criteria.

  7. ‡Although medians were the same, P value implied MAE used more units of blood (mean 6·2 vs. 5·5).

WHO toxicity
 Oral toxicity26%20%0·822%15%0·0432%16%15%11%
 Cardiac function†3%4%0·71%1%0·81%1%5%4%
 Other toxicity*26%28%0·811%16%0·523%26%21%31%
Supportive care
 Units of blood550·05‡32<0·0001n/an/an/an/a
 Units of platelets11100·364<0·0001n/an/an/an/a
 Units of antibiotics21190·101611<0·0001n/an/an/an/a
 Days to neutrophil recovery to 1·0 × 109/l24210·0032921<0·000124263133
 Days to platelet recovery to 100 × 109/l19170·042716<0·000124283239

Sixteen of the 32 patients who died in CR were treated with chemotherapy alone. The main cause of death in these patients was infection (n = 11). Two patients died of cardiac failure including a DS patient with complex congenital heart disease. A third patient required treatment with angiotensin-converting enzyme (ACE) inhibitors for cardiac failure associated with sepsis. All three cardiac events followed the third course of chemotherapy or later. Cardiac monitoring was recommended with 2D echocardiography prior to chemotherapy, after each course of treatment, at the end of treatment and at 1 and 5 years after the end of treatment. Compliance was disappointingly poor and the prevalence of subclinical cardiotoxicity could not be defined.

Sixteen patients died in CR following a SCT; sibling (n = 4), matched unrelated (n = 7), mismatched unrelated (n = 3) and haploidentical (n = 2). The trial permitted matched sibling donor SCT only. Eleven of these 16 SCT patients only achieved morphological remission after receiving 2–4 courses of chemotherapy. They then received 4–6 courses of treatment, which included additional protocol or alternative chemotherapy before proceeding to SCT, increasing their risk of morbidity and mortality.

Comparison with MRC AML10

Children in AML12 experienced better survival than those in AML10 (63% vs. 56%). Outcomes for other endpoints were: deaths in CR (6% vs. 10%); DFS (59% vs. 51%), EFS (54% vs. 47%) and RR (35% vs. 39%). When APL and DS patients were excluded 10-year survival probabilities for AML12 vs. AML10 were: OS 61% vs. 54%, DFS 56% vs. 51%, EFS 51% vs. 47% and RR 38% vs. 39% and deaths in remission were 5% vs. 10%.


  1. Top of page
  2. Summary
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Author contributions
  8. Disclosure
  9. References
  10. Appendix

Induction therapy, comprising an anthracycline/anthracenedione and cytarabine with etoposide or thioguanine as a third drug, achieves remission rates of up to 90% in children with AML (Stevens et al, 1998; Creutzig et al, 2005a; Lie et al, 2005). Several studies have compared daunorubicin with an alternative anthracycline or anthracenedione, especially idarubicin or mitoxantrone, but these alternatives have not been convincingly shown to improve outcome (Arlin et al, 1990; Löwenberg et al, 1998; The AML Collaborative Group., 1998; Mandelle et al, 2009), and there are no randomized studies comparing mitoxantrone with daunorubicin in induction in childhood AML.

This paper presents a subgroup analysis of the AML12 trial and the results should be considered alongside those of the whole trial (Burnett et al, 2010) and interpreted cautiously. The comparisons in the paediatric part of the trial have limited power and a non-significant result does not necessarily imply no effect. There was a significant reduction in relapse with MAE, but due to additional TRM this did not translate into a significant improvement in EFS or OS. Whilst the confidence intervals around treatment effects in children for both OS and EFS did not rule out the possibility that MAE had a detrimental effect, the point estimates of the effects were consistent with the possibility of a benefit for MAE. These results are broadly consistent with those in the trial as a whole, which showed a clear reduction in relapse with MAE, but additional TRM leading to no significant difference in DFS or OS. A subgroup analysis of the overall study based on age suggested a possible trend towards MAE being more favourable in younger patients (test for trend P = 0·08), and thus it is possible that young patients were better able to tolerate the chemotherapy than adults, and therefore benefited from any increased anti-leukaemic effect of MAE.

Mitoxantrone was associated with increased myelosuppression and a greater need for transfusions and antibiotics, but there was no significant difference in non-haematological toxicity or TRM. The increase in haematological toxicity was most marked after the second course (Burnett et al, 2010). The increased haematological toxicity and reduced relapse with mitoxantrone may indicate that the dose chosen in AML12 was more intensive than that of daunorubicin, and that higher doses of daunorubicin might be similarly effective; this has been demonstrated for idarubicin in adults where high dose daunorubicin was equivalent (Ohtake et al, 2011). Any increase in treatment intensity was solely related to the dose selected and not due to the time interval between courses; intensive timing of chemotherapy has been shown to increase anti-leukaemic effect in some studies (Smith et al, 2005).

Post-induction consolidation therapy has been shown to reduce relapse and prolong survival in AML, but the optimal number of courses is uncertain; in recent paediatric trials this has ranged from two to eight courses of intensive chemotherapy (Wells et al, 1994; Stevens et al, 1998; Creutzig et al, 2005a; Lie et al, 2005; Tomizawa et al, 2007). In AML10, the standard arm was four courses in total, and the role of a fifth course of chemotherapy was tested in AML12. There was no difference in any measure of outcome between children treated with either four or five courses of chemotherapy at 10 years of follow-up, although the study did not have adequate power to answer this question for children alone.

The OS and EFS for children treated in AML12 were similar to those in contemporary studies from the majority of major collaborative groups (Creutzig et al, 2005a; Lie et al, 2005). In the AML99 trial, the Japanese Childhood Collaborative Study Group achieved OS of 76% and EFS of 61% at 5 years, but 37% of children had favourable cytogenetics (t(8:21) or inversion (16) (Tsukimoto et al, 2009)). In the more recent St Jude AML02 study (recruitment period 2002–2008), OS was 71% and EFS 63% with therapy guided by minimal residual disease (MRD), but median follow up was short at only 3 years (Rubnitz et al, 2010). Preliminary unpublished reports of other recent trials [MRC AML15, BFM (Berlin-Frankfürt-Münster) AML 2004, NOPHO (Nordic Society of Paediatric Haematology and Oncology) AML 2004] likewise show favourable OS, but suggest that limited improvements owe much to better supportive care and salvage therapy.

Ten percent of children treated in AML12 died from treatment-related causes, 4% during induction and 6% in remission. The induction death rate and incidence of resistant disease were the same as that during AML10 (Riley et al, 1999), and there was no difference between AML12 and AML10 in the rate of complete remission. In AML10, the TRM fell significantly between the first (18%) and second (10%) halves of the trial due to fewer deaths in remission, and the TRM in AML12 (10%) was closely similar to the second half of AML10. Infection was the main cause of death in both trials, and the improvement was due to a reduction in infection-related deaths (11% in AML10, 5% in AML12). Heart failure, apart from that associated with severe sepsis, was a very rare cause of death in both trials. Haemorrhagic deaths occurred early, before or during the first course, related to high count, usually monocytic disease.

A high proportion of deaths in remission followed allo-SCT. Whilst fewer children received a SCT in AML12 compared to AML10 (11% vs. 20%), transplant-related death contributed the same percentage of deaths to both trials, because children registered in AML12 who received SCT were more likely to have an unrelated donor procedure. MRC trials have progressively restricted the use of SCT as consolidation therapy because of the absence of clear benefit and risks of mortality and morbidity.

Relapse in the CNS alone or combined with another site affected 7% of children in AML12, where all children received triple intrathecal chemotherapy (ITT) and those with CNS disease at diagnosis were recommended to receive CRT if aged over 2 years. Other groups have reported low rates of CNS relapse with ITT alone, and MRC/UK regimens no longer recommend CRT for CNS disease (Tsukimoto et al, 2009; Rubnitz et al, 2010), in line with the experience of the Children’s Oncology Group (COG) (AS Gamis, COG AML Committee, Children’s Mercy Hospital Kansas City, USA, personal communication).

Cytogenetics was the most important determinant of prognosis (Grimwade et al, 1999; Grimwade et al, 2010; Harrison et al, 2010). Adverse risk karyotypes in children were associated with less unfavourable outcomes than adults. This could be explained by the observation that karyotype complexity in the absence of other favourable and unfavourable abnormalities (the largest group) was shown to be intermediate risk in children treated according to MRC AML10 and 12 (Grimwade et al, 1999). WCC was also an important determinant of outcome on multivariate analysis. Risk group stratification continues to evolve as the prognostic significance of MRD and new molecular markers are defined and the MRC now employs a risk index based on response after course 1, cytogenetics, age, WCC, gender and de novo/secondary AML.

Children aged under one year and those with DS had a high risk of TRM, primarily due to gut toxicity and increased infection (Webb et al, 2001; Webb, 2005; Rao et al, 2006). This effect in children under one year of age was despite dose reductions stipulated in the protocol. Both groups of patients had less resistant disease, less relapse and better DFS. It is possible that the dose adjustments in AML12 resulted in more intensive chemotherapy for infants compared to older children, but as survival in infants was superior, supportive care has been stressed rather than further dose reductions (Webb et al, 2001). In future trials children with DS should receive reduced intensity therapy, especially less anthracycline (Creutzig et al, 2005b, Webb, 2005; Abildgaard et al, 2006) and contemporary studies by the BFM and COG have adopted this approach.

Children with APL had a low induction death rate (4%), and a high CR rate (95%), probably due to the routine use of ATRA. Survival was inferior to the best-published series (OS 89% and EFS 76% at 10 years) (Testi et al, 2005) due to a relatively high relapse rate and deaths in remission. This subgroup now receives disease-specific therapy based on the GIMEMA (Gruppo Italiano Malattie Ematologiche dell’Adulto) – AIEOP-AIDA Associazione Italiana di Ematologia e Oncologia Pediatrica – ATRA + idirubicin) experience (Testi et al, 2005), but with reduced anthracycline, molecular monitoring and selective use of arsenic trioxide, in line with contemporary studies in Europe and the United States.


  1. Top of page
  2. Summary
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Author contributions
  8. Disclosure
  9. References
  10. Appendix

The authors are grateful to the participating clinicians of the MRC Childhood Leukaemia Working Party and the Dutch Childhood Oncology Group, participating patients and their parents, and Professor Alan Burnett, Chairman of the United Kingdom Adult Leukaemia Working Party, for his support. This study was supported by United Kingdom Medical Research Council.

Author contributions

  1. Top of page
  2. Summary
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Author contributions
  8. Disclosure
  9. References
  10. Appendix

BESG, DKHW, AJH, SSNdeG and KW wrote the paper; AJH carried out the statistical analyses; BESG, DKHW and SSNdeG contributed to coordination of the trial; CJH contributed the cytogenetic data. All authors critically reviewed the manuscript and gave final approval of the paper.


  1. Top of page
  2. Summary
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Author contributions
  8. Disclosure
  9. References
  10. Appendix
  • Abildgaard, L., Ellebaek, E., Gustafsson, G., Abramsson, J., Hovi, L., Jonmundsson, G., Zeller, B. & Hasle, H. (2006) Optimal treatment intensity in children with Down syndrome and myeloid leukaemia; data from 56 children treated on NOPHO-AML protocols and a review of the literature. Annals of Hematology, 85, 275280.
  • Arlin, Z., Case, D.C., Moore, J., Wiernik, P., Feldman, E., Saletan, S., Desai, P., Sia, L. & Cartwright, K. (1990) Randomised multicenter trial of cytosine arabinoside with mitoxantrone or daunorubicin in previously untreated adult patients with acute nonlymphocytic leukemia (ANLL). Lederle Cooperative Group. Leukemia, 4, 177183.
  • Burnett, A.K., Hills, R., Milligan, D.W., Goldstone, A.H., Prentice, A.G., McMullin, M.-F., Duncombe, A., Gibson, B. & Wheatley, K. (2010) Attempts to optimize induction and consolidation treatment in acute myeloid leukemia: results of the MRC AML12 trial. Journal of Clinical Oncology, 28, 586595.
  • Creutzig, U., Zimmermann, M., Ritter, J., Reinhardt, D., Hermann, J., Henze, G., Jurgens, H., Kabisch, H., Reiter, A., Riehm, H., Gadner, H. & Schellong, G. (2005a) Treatment strategies and long-term results in paediatric patients treated in four consecutive AML-BFM trials. Leukemia, 19, 20302042.
  • Creutzig, U., Reinhardt, D., Diekamp, S., Dworzak, M., Stary, J. & Zimmermann, M. (2005b) AML patients with Down syndrome have a high cure rate with AML-BFM therapy with reduced dose intensity. Leukaemia, 19, 13551366.
  • Gibson, B.E.S., Wheatley, K., Hann, I.M., Stevens, R.F., Webb, D., Hills, R.K., de Graaf, S.S.N., Harrison, C.J.; for the UK Childhood Leukaemia Working Party and the Dutch Childhood Oncology Group. (2005) Treatment strategy and long-term results in paediatric patients treated in consecutive UK AML trials. Leukemia 19: 21302138.
  • Grimwade, D., Walker, H., Oliver, F., Wheatley, K., Harrison, C., Harrison, G., Rees, J., Hann, I., Stevens, R., Burnett, A. & Goldstone, A. (1999) The importance of diagnostic cytogenetics on outcome in AML: analysis of 1612 patients entered into the MRC AML10 trial. Blood, 92, 23222333.
  • Grimwade, D., Hills, R.K., Moorman, A.V., Walker, H., Chatters, S., Goldstone, A.H., Wheatley, K., Harrison, C.J., Burnett, A.K.; on behalf of the National Cancer Research Institute Adult Leukaemia Working Group. (2010) Refinement of cytogenetic classification in acute myeloid leukemia: determination of prognostic significance of rare recurring chromosomal abnormalities amongst 5,876 younger adult patients treated in the UK Medical Research Council trials. Blood, 116, 254265.
  • Harrison, C.J., Hills, R.K., Moorman, A.V., Grimwade, D., Hann, I., Webb, D.K.H., Stevens, R., Wheatley, K., de Graff, S.S.N., van Berg, E., Burnett, A.K. & Gibson, B.E.S. (2010) Cytogenetics of childhood acute myeloid leukemia: United Kingdom Medical Research Council trials AML10 and 12. Journal of Clinical Oncology, 28, 26742681.
  • Kalbfleisch, J.D. & Prentice, R. (1980) The Statistical Analysis of Failure Time Data. John Wiley & Sons, Hoboken, NJ, USA.
  • Lange, B.J., Smith, F.O., Feusner, J., Barnard, D.R., Dinndorf, P., Feig, S., Heerema, N.A., Arndt, C., Arceci, R.J., Seibel, N., Weiman, M., Dusenbery, K., Shannon, K., Luna-Fineman, S., Gerbing, R.B. & Alonzo, T.A. (2008) Outcomes in CCG-2961, a Children’s Oncology Group Phase 3 Trial for untreated pediatric acute myeloid leukemia: a report from the Children’s Oncology Group. Blood, 111, 10441053.
  • Lie, S.O., Abrahamsson, J., Clausen, N., Forestier, E., Hasle, H., Hovi, L., Jonmundsson, G., Mellander, L., Slimes, M.A., Yssing, M., Zeller, B. & Gustafsson, G. (2005) Long-term results in children with AML: NOPHO-AML Study Group – report of 3 consecutive trials. Leukemia, 19, 20902100.
  • Löwenberg, B., Suciu, S., Archimbaud, E., Haak, H., Stryckmans, P., de Cataldo, R., Dekker, A.W., Berneman, Z.N., Thyss, A., van der Lelle, J., Sonneveld, P., Visani, G., Fillet, G., Hayat, M., Hagemeijer, A., Solbu, G. & Zittoun, R. (1998) Mitoxantrone versus daunorubicin in induction-consolidation chemotherapy - the value of low dose cytarabine for maintenance of remission and an assessment of prognostic factors in acute myeloid leukemia in the elderly: final report. European Organization for the Research and Treatment of Cancer and the Dutch-Belgian Hemato-Oncology Cooperative HOVON Group. Journal of Clinical Oncology, 16, 872881.
  • Mandelle, F., Vignetti, M., Suciu, S., Stasi, R., Petti, M.-C., Meloni, G., Muus, P., Marmont, F., Marie, J.-P., Labar, B., Thomas, X., di Raimondi, F., Willemze, R., Liso, V., Ferrara, F., Baila, L., Fazi, P., Zittoun, R., Amadori, S. & de Witte, T. (2009) Daunorubicin versus mitoxantrone versus idarubicin as induction and consolidation chemotherapy for adults with acute myeloid leukemia: the EORTC and GIMEMA groups study AML-10. Journal of Clinical Oncology, 32, 53975403.
  • Ohtake, S., Miyawaki, S., Fujita, H., Kiyoi, H., Shinagawa, K., Usui, N., Okumura, H., Miyamura, K., Nakaseko, C., Miyazaki, Y., Fujieda, A., Nagai, T., Yamane, T., Taniwaki, M., Takahashi, M., Yagasaki, F., Kimura, Y., Asou, N., Sakamaki, H., Handa, H., Honda, S., Ohnishi, K., Naoe, T. & Ohno, R. (2011) Randomized study of induction therapy comparing standard-dose idarubicin with high dose daunorubicin in adult patients with previously untreated acute myeloid leukemia. JALSG AML201 Study. Blood, 117, 22992300.
  • Rao, A., Hills, R.K., Stiller, C., Gibson, B.E., de Graaf, S.S., Hann, I.M., O’Marcaigh, A., Wheatley, K. & Webb, D.K. (2006) Treatment for myeloid leukaemia of Down syndrome: population-based experience in the UK and results from the Medical Research Council AML10 and 12 trials. British Journal of Haematology, 132, 576583.
  • Riley, L., Hann, I.M., Wheatley, K. & Stevens, R.F. (1999) Treatment-related deaths during induction and first remission of acute myeloid leukaemia in children treated on the Tenth Medical Research Council acute myeloid leukaemia trial (MRC AML10). The MRC Childhood Leukaemia Working Party. British Journal of Haematology, 106, 436444.
  • Rubnitz, J.E., Inaba, H., Dahl, G., Ribeiro, R.C., Bowman, W.P., Taub, J., Pounds, S., Razzouk, B.I., Lacayo, N.J., Cao, X., Meshinchi, S., Degar, B., Airewele, G., Raimondi, S.C., Onciu, M., Coustan-Smith, E., Downing, J.R., Leung, W., Pui, C.H. & Campana, D. (2010) Minimal residual disease directed therapy for childhood acute myeloid leukemia: results of the AML02 multicentre trial. Lancet Oncology, 11, 543552.
  • Smith, F.O., Alonzo, T.A., Gerbing, R.B., Woods, W.G., Arceci, R.J. & Children’s Cancer Group. (2005) Long-term results of children with acute myeloid leukaemia: a report of three consecutive Phase III studies by the Children’s Cancer Group: CCG 251, CCG 213 and CCG 2891. Leukaemia, 19, 20542062.
  • Stevens, R.F., Hann, I.M., Wheatley, K. & Gray, R.G. (1998) Marked improvements in outcome with chemotherapy alone in paediatric acute myeloid leukaemia: results of the United Kingdom Medical Research Council’s 10th AML trial. British Journal of Haematology, 101, 130140.
  • Testi, A.M., Biondi, A., Lo Coco, F., Moleti, M.L., Giona, F., Vignetti, M., Menna, G., Locatelli, F., Pession, A., Barisone, E., De Rossi, G., Diverio, D., Micalizzi, C., Arico, M., Basso, G., Foa, R. & Mandelli, F. (2005) GIMEMA-AIEOPAIDA protocol for the treatment of newly diagnosed acute promyelocytic leukemia (APL) in children. Blood, 106, 447453.
  • The AML Collaborative Group. (1998) A systematic collaborative overview of randomised trials comparing idarubicin with daunorubicin (or other anthracyclines) as induction therapy for acute myeloid leukaemia. British Journal of Haematology, 103, 100109.
  • Tomizawa, D., Tabuchi, K., Kinoshita, A., Hanada, R., Kigasawa, H., Tsukimoto, I. & Tsuchida, M. (2007) Repetitive cycles of high-dose cytarabine are effective for childhood acute myeloid leukemia: long-term outcome of the children with AML treated on two consecutive trials of Tokyo Children’s Cancer Study Group. Pediatric Blood & Cancer, 49, 127132.
  • Tsukimoto, I., Tawa, A., Horibe, K., Tabuchi, K., Kigasawa, H., Tsuchida, M., Yabe, H., Nakayama, H., Kudo, K., Kobayashi, R., Hamamoto, K., Imaizumi, M., Morimoto, A., Tsuchiya, S. & Hanada, R. (2009) Risk-stratified therapy and the intensive use of cytarabine improves the outcome in childhood acute myeloid leukemia: the AML99 trial from the Japanese Childhood AML cooperative group. Journal of Clinical Oncology, 27, 40074013.
  • Webb, D.K. (2005) Optimizing therapy for myeloid disorders of Down syndrome. British Journal of Haematology, 131, 37.
  • Webb, D.K., Harrison, G., Stevens, R.F., Gibson, B.G., Hann, I.M., Wheatley, K. & MRC Childhood Leukemia Working Party. (2001) Relationships between age at diagnosis, clinical features, and outcome of therapy in children treated in the Medical Research Council AML10 and 12 trials for acute myeloid leukaemia. Blood, 98, 17141720.
  • Wells, R.J., Woods, W.G., Buckley, J.D., Odom, L.F., Benjamin, D., Bernstein, I., Betcher, D., Feig, S., Kim, T., Ruymann, F., Smithson, W., Srivastava, A., Tannous, R., Buckley, C.M., Whitt, K., Wolff, L. & Lampkin, B.C. (1994) Treatment of newly diagnosed children and adolescents with acute myeloid leukaemia: a Children’s Cancer Group study. Journal of Clinical Oncology, 12, 23672377.
  • Wheatley, K., Burnett, A.K., Goldstone, A.H., Gray, R., Hann, I.M., Harrison, C.J., Rees, J.K., Stevens, R.F. & Walker, H. (1999) A simple, robust, validated and highly predictive index for the determination of risk-directed therapy in acute myeloid leukaemia derived from the MRC AML10 trial. United Kingdom Medical Research Council’s Adult and Childhood Leukaemia Working Parties. British Journal of Haematology, 107, 6979.
  • World Health Organization. (1979) Handbook for Reporting Results of Cancer Treatment. World Health Organization, Geneva, Switzerland, 1522.


  1. Top of page
  2. Summary
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Author contributions
  8. Disclosure
  9. References
  10. Appendix

Appendix 1

The following institutions and members participated in the study:

Membership of the MRC Childhood Leukaemia Working Party (1998–2002): M Caswell, J Chessells, P Darbyshire, S Dempsey, J Durrant, O Eden, K Forman, M Gattens, B Gibson, N Goulden, C Haworth, I Hann, C Harrison, F Hill, R Hills, M Jenney, D King, S Kinsey, D Lancaster, J Lilleyman, M Madden, S Mellor, M Madi, C Mitchell, M Morgan, A Oakhill, M Radford, S Richards, V Saha, J Simpson, O Smith, R Stevens, A Thomas, A Vora, D Webb, K Wheatley, A Will, K Windybank. Director of the UK Leukaemia Research Cytogenetics Group was C Harrison.

Membership of the Dutch Childhood Oncology Group ANLL-97 Committee (1998–2002): H van den Berg, M Bierings, A van der Does-van den Berg, G Kardos, MH van Weel-Sipman (deceased) and SSN de Graaf. Chair of the Dutch Workgroup on Cancer Genetics and Cytogenetics was E van den Berg.

Participants of the New Zealand Centres : Professor DNJ Hart, Dr RL Spearing, Dr MJ Sullivan, Dr L Teague and Dr L Pitcher.