What Is the Optimal Therapy for Childhood AML?

August 1, 2002
Kenan Onel, MD, PhD

Peter Steinherz, MD

Oncology, ONCOLOGY Vol 16 No 8, Volume 16, Issue 8

The past 30 years have seen tremendous advances in the treatment of pediatric leukemia. What was once an invariably fatal diagnosis is now quite curable in close to 80% of cases. Unfortunately for children with acute myelogenous leukemia (AML), most of these developments have been in the treatment of acute lymphoblastic leukemia (ALL); even today, nearly half of all children diagnosed with AML will die of the disease.

The past 30 years have seen tremendous advances in the treatment of pediatricleukemia. What was once an invariably fatal diagnosis is now quite curable inclose to 80% of cases. Unfortunately for children with acute myelogenousleukemia (AML), most of these developments have been in the treatment of acutelymphoblastic leukemia (ALL); even today, nearly half of all children diagnosedwith AML will die of the disease.

In their excellent and comprehensive article, Loeb and Arceci review thecurrent state of therapies for pediatric AML throughout the world, summarizingthe approaches taken by the various major cooperative groups. In essence, thesestrategies have involved the intensification of either the dosages or timing ofthe few chemotherapeutic agents active against AML.

Distinct Disease Entities

Several themes running through the review are worth emphasizing. First, ithas become increasingly clear that, despite their common CD34-positivehematopoietic stem cell origin, AML and ALL are biologically quite distinct, andthe successful treatment of AML hinges upon differing treatment strategies thanthose used for ALL. While ALL initially responds rapidly to a number of agents,even when they are used alone, AML is resistant to all but a few drugs. Whereasregimens involving intensive but short remission induction and consolidationphases and prolonged low-intensity maintenance cycles have been effective in thetreatment of ALL, this approach has not been successful in the treatment of AML.

Rather, it was found in early Children’s Cancer Group (CCG) studies thatsurvival was significantly improved by intensive but short courses of therapy,when directly compared against prolonged ALL-style maintenance treatments.[1] Infact, it was later shown in the Medical Research Council (MRC) AML9 study thatextended maintenance therapy might be detrimental to survival because ofprotracted periods of immunosuppression.[2]

Timed Sequential Therapy

In contrast to the rapid proliferation characteristic of ALL blasts, thepercentage of AML cells in cycle is typically quite low because of the presenceof a secreted hematopoietic inhibitory factor. This factor is no longer presentfollowing a single cycle of chemotherapy but is replaced by a factor thatstimulates proliferation.[3] These insights into the biology of AML led the CCGto adopt "timed sequential therapy" as its overall strategy in thetreatment of AML, the second major theme of this review.

With this approach, sequential cycles of chemotherapy are given after onlyshort periods of rest, without awaiting hematologic recovery. When this approachwas compared to standard timing in CCG 2891, the results were quite impressive.Event-free survival at 3 years was 42% for the intensively timed cohort, butonly 27% for the standard timing group.[4] Furthermore, among those patients whoachieved remission, the survival benefit of intensive timing remainedsignificant at 8 years (49% vs 34% for those in the standard timing group).[5]

Dose Intensification

Clearly, 49% survival is not good enough. Hence, the final theme of thisreview is dose intensification of agents known to be active against AML, and theconcomitant development of novel agents and strategies. The use of high-dosecytarabine (Ara-C) has emerged as the cornerstone of treatment in AML. Itsutility was initially demonstrated in the Pediatric Oncology Group (POG) study8498, and then extended in the subsequent POG 8821 trial.[6,7] This strategy hassubsequently been incorporated into most treatment regimens worldwide.

While the British and American cooperative groups have pursued similarapproaches to dose intensification, the strategy of the German group has been tostratify patients into standard- and high-risk groups using FAB histologycriteria, and response to induction therapy as determined by blast percentage inthe bone marrow on day 15.[8] Although they have identified putative high- andlow-risk groups by retrospective multivariate analyses, their success has beenlimited in tailoring treatment to these groups. Interestingly, they have foundthat prophylactic cranial irradiation confers a survival benefit on allpatients.[9]

Bone Marrow Transplantation

As a result of these treatment strategies, it is now possible to induceremission in over 80% of children with AML. Remission consolidation, andprolonged maintenance of remission, however, remain problematic. When intensivechemotherapy, autologous bone marrow transplantation (BMT), and allogeneic BMTare directly compared, allogeneic BMT for those patients with a matched donorappears to confer a statistically significant survival benefit. The exception tothis finding was seen in the MRC AML 10 study,[10] which showed no advantage toBMT over chemotherapy. These results, however, may have been compromised by lowcompliance with the BMT arms of the study, and an overrepresentation ofgood-risk patients in the study population.

BMT is associated with significant morbidity and mortality, however. Inaddition, an appropriate donor is not an option for many patients. Furtherchemotherapeutic dose intensification will be associated with ever greatertoxicities. Hence, innovative therapeutic modalities or improvements in existingregimens must be devised. Loeb and Arceci describe some of these potential newtherapies in the final section of their review.

Targeted Therapies

AML is not a monolithic disease. Instead, it is a series of diseasephenotypes that result from an arrest at various and differing steps along thepathways of hematopoietic cell differentiation. Correctly, the authors recognizethat novel, disease-specific treatments are most likely to result from anunderstanding of the molecular basis for the developmental blocks that result inAML.

One example of such targeted therapy is the use of all-trans-retinoicacid (ATRA, Vesanoid) in the treatment of acute promyelocytic leukemia (APL).APL is characterized by several chromosomal translocations that result in theexpression of fusion proteins involving one of the retinoic acid receptors, RAR-alpha,which is thought to interfere with a differentiation program directed byretinoic acid. By binding to this fusion product, ATRA apparently short-circuitsthis block.

Intriguingly, the most common translocation in APL fuses RAR-alpha to the PMLprotein. PML bodies themselves have recently been found by Pier Paolo Pandolfiof Memorial Sloan-Kettering and others to be tumor suppressors in their ownright. Hence, dysregulation of PML may itself play a role in the development ofAPL.

Another example of targeted therapy has proven to be much more disappointing.Imatinib mesylate (STI-571, Gleevec) is a tyrosine kinase inhibitor withactivity against the abl tyrosine kinase, overexpression of which is the resultof the Philadelphia chromosome (Ph). Although this inhibitor is very effectivein treating Ph-positive CML, it is much less useful in the treatment ofPh-positive AML or ALL. Another oncoprotein overexpressed in AML is the c-kitkinase. Although imatinib has activity against the c-kit kinase in vitro as wellas in the rare adult gastrointestinal stromal tumor (GIST), its utility againstc-kit-expressing AML blasts in vivo remains to be seen.


The optimal therapy for childhood AML remains to be determined. For theforeseeable future, dose-intensive timed sequential chemotherapy with high-dosecytarabine plus an anthracycline and other active agents is likely to remain themainstay of induction therapy. Most likely, new treatments such as anti-CD33antibodies or tumor vaccines will be of greatest efficacy once minimal residualdisease has been achieved, either as consolidation therapy on their own, or, indefined high-risk patients, in conjunction with an allogeneic BMT. However,because multiple pathways are likely to be deranged in leukemic blasts, it maywell be that specific inhibitors will not be effective against AML.

In the end, while AML may never be as curable as ALL, insights gained throughthe efforts of the cooperative groups and basic scientists in understanding thebiology of AML can only result in significant improvements in survival, while atthe same time decreasing associated morbidity.


1. Wells RJ, Woods WG, Lampkin BC, et al: Impact of high-dose cytarabine andasparaginase intensification on childhood acute myeloid leukemia: A report fromthe Children’s Cancer Group. J Clin Oncol 11(3):538-545, 1993

2. Rees JK, Gray RG, Wheatley K: Dose intensification in acute myeloidleukaemia: greater effectiveness at lower cost. Principal report of the MedicalResearch Council’s AML9 study. MRC Leukaemia in Adults Working Party. Br JHaematol 94(1):89-98, 1996.

3. Burke PJ, Diggs CH, Owens AH Jr: Factors in human serum affecting theproliferation of normal and leukemic cells. Cancer Res 33(4):800-806, 1973.

4. Woods WG, Kobrinsky N, Buckley JD, et al: Timed-sequential inductiontherapy improves postremission outcome in acute myeloid leukemia: A report fromthe Children’s Cancer Group. Blood 87(12):4979-4989, 1996.

5. Woods WG, Neudorf S, Gold S, et al: A comparison of allogeneic bone marrowtransplantation, autologous bone marrow transplantation, and aggressivechemotherapy in children with acute myeloid leukemia in remission. Blood97(1):56-62, 2001.

6. Ravindranath Y, Yeager AM, Chang MN, et al: Autologous bone marrowtransplantation versus intensive consolidation chemotherapy for acute myeloidleukemia in childhood. Pediatric Oncology Group. N Engl J Med 334(22):1428-1434,1996.

7. Ravindranath Y, Steuber CP, Krischer J, et al: High-dose cytarabine forintensification of early therapy of childhood acute myeloid leukemia: APediatric Oncology Group study. J Clin Oncol 9(4):572-580, 1991.

8. Creutzig U, Zimmermann M, Ritter J, et al: Definition of a standard-riskgroup in children with AML. Br J Haematol 104(3):630-639, 1999.

9. Creutzig U, Ritter J, Zimmermann M, et al: Does cranial irradiation reducethe risk for bone marrow relapse in acute myelogenous leukemia? Unexpectedresults of the Childhood Acute Myelogenous Leukemia study BFM-87. J Clin Oncol11(2):279-286, 1993.

10. Stevens RF, Hann IM, Wheatley K, et al: Marked improvements in outcomewith chemotherapy alone in paediatric acute myeloid leukemia: Results of theUnited Kingdom Medical Research Council’s 10th AML trial. MRC ChildhoodLeukaemia Working Party. Br J Haematol 101(1):130-140, 1998.