MED ICA L PROGR ES S

Review Article

Medical Progress mic infiltration of various tissues, including the liver

(hepatomegaly), spleen (splenomegaly), skin (leuke-
mia cutis), lymph nodes (lymphadenopathy), bone
(bone pain), gingiva, and central nervous system,
A CUTE M YELOID L EUKEMIA can produce a variety of other symptoms. An isolat-
ed mass of leukemic blasts is usually referred to as a
BOB LÖWENBERG, M.D., JAMES R. DOWNING, M.D., granulocytic sarcoma. Hyperleukocytosis (more than
AND ALAN BURNETT, M.D. 100,000 white cells per cubic millimeter) can lead to
symptoms of leukostasis, such as ocular and cere-
brovascular dysfunction or bleeding. There may also
be metabolic abnormalities (e.g., hyperuricemia and
hypocalcemia), although these are rarely found at

A
CUTE myeloid leukemia (AML) is charac- presentation.
terized by an increase in the number of my-
eloid cells in the marrow and an arrest in DIAGNOSIS
their maturation, frequently resulting in hematopoi- The primary diagnosis of AML rests on the mor-
etic insufficiency (granulocytopenia, thrombocyto- phologic identification of leukemic myeloblasts in
penia, or anemia), with or without leukocytosis. In preparations of peripheral blood and bone marrow
the United States, the annual incidence of AML is stained with Wright–Giemsa. These cells have round-
approximately 2.4 per 100,000,1 and it increases pro- to-irregular nuclei, distinct nucleoli, and very little
gressively with age, to a peak of 12.6 per 100,000 cytoplasm. The cytoplasm frequently contains fine
adults 65 years of age or older. Until the 1970s, the azurophilic granules and a variable number of Auer
diagnosis was based solely on the pathological and bodies, or rods (azurophilic granules within lyso-
cytologic examination of bone marrow and blood. somes). The presence of more than 30 percent leu-
Five-year survival rates during this period were less kemic blasts in a bone marrow aspirate is required
than 15 percent. Over the past decade, refinements for a definitive diagnosis of acute leukemia; before
in the diagnosis of subtypes of AML and advances therapy is initiated, however, several critical diagnos-
in therapeutic approaches have improved the out- tic distinctions must be made. AML must be distin-
look for patients with AML. Despite these improve- guished from acute lymphoblastic leukemia (ALL),
ments, however, the survival rate among patients myelodysplastic syndrome (MDS), or AML arising in
who are less than 65 years of age is only 40 percent. the setting of MDS, because therapeutic strategies
In this article, we will review the diagnostic criteria, and prognosis vary considerably for these diseases.
pathology, and treatment of AML, emphasizing new AML can be distinguished from ALL by demon-
findings that promise to improve the cure rates. stration of definitive commitment to the myeloid
CLINICAL PRESENTATION lineage through judicious use of morphologic, im-
munohistochemical, and immunologic methods.2-4
The clinical signs and symptoms of AML are di- The distinction of AML from MDS or MDS-related
verse and nonspecific, but they are usually directly AML is more difficult and requires careful clinical,
attributable to the leukemic infiltration of the bone morphologic, and genetic analysis. MDS is charac-
marrow, with resultant cytopenia. Typically, patients terized by ineffective hematopoiesis, and although
present with signs and symptoms of fatigue, hemor- its diagnosis rests largely on morphologic evidence
rhage, or infections and fever due to decreases in red of dysplastic maturation, characteristic cytogenetic
cells, platelets, or white cells, respectively. Pallor, fa- lesions, including the loss of all or part of chromo-
tigue, and dyspnea on exertion are common. Leuke- some 5 (del5q) or chromosome 7, the deletion of
the long arm of chromosome 20 (del20q), and the
loss of Y, are found in a substantial percentage of pa-
From the Department of Hematology, Erasmus University and Univer- tients. Conversion to AML is diagnosed when the
sity Hospital Rotterdam, Rotterdam, the Netherlands (B.L.); the Depart- percentage of myeloblasts in the marrow exceeds 30
ment of Pathology and Laboratory Medicine, St. Jude Children’s Research
Hospital, Memphis, Tenn. (J.R.D.); and the Department of Hematology,
percent. Conventional therapy for AML is much less
University of Wales, College of Medicine, Cardiff, United Kingdom (A.B.). effective against MDS-related AML, and thus it is
Address reprint requests to Dr. Löwenberg at the Department of Hema- useful to distinguish between these two conditions
tology, University Hospital Rotterdam, P.O. Box 5201, 3008 AE Rotter-
dam, the Netherlands. at the time of the initial diagnosis.
©1999, Massachusetts Medical Society. Once a diagnosis of AML is made, the morpho-

Vol ume 341 Numb e r 14 · 1051

 The New Eng land Jour nal of Medicine

TABLE 1. THE FRENCH–AMERICAN–BRITISH (FAB) CLASSIFICATION OF AML AND ASSOCIATED GENETIC ABNORMALITIES.

ASSOCIATED
TRANSLOCATIONS AND
FAB COMMON NAME REARRANGEMENTS
SUBTYPE (% OF CASES) RESULTS OF STAINING (% OF CASES) GENES INVOLVED
MYELOPER- SUDAN NONSPECIFIC
OXIDASE BLACK ESTERASE

M0 Acute myeloblastic leukemia with mini- ¡ ¡ ¡* inv(3q26) and EVI1

mal differentiation (3%) t(3;3) (1%)
M1 Acute myeloblastic leukemia without + + ¡
maturation (15–20%)
M2 Acute myeloblastic leukemia with matu- + + ¡ t(8;21) (40%), AML1-ETO,
ration (25–30%) t(6;9) (1%) DEK-CAN
M3 Acute promyelocytic leukemia + + ¡ t(15;17) (98%), PML-RARa, PLZF-
(5–10%) t(11;17) (1%), RARa, NPM RARa
t(5;17) (1%)
M4 Acute myelomonocytic leukemia (20%) + + + 11q23 (20%), MLL, DEK-CAN, EVI1
inv(3q26) and
t(3;3) (3%),
t(6;9) (1%)
M4Eo Acute myelomonocytic leukemia with + + + inv(16), t(16;16) CBFb-MYH11
abnormal eosinophils (80%)
(5–10%)
M5 Acute monocytic leukemia ¡ ¡ + 11q23 (20%), MLL, MOZ-CBP
(2–9%) t(8;16) (2%)
M6 Erythroleukemia (3–5%) + + ¡
M7 Acute megakaryocytic leukemia ¡ ¡ +† t(1;22) (5%) Unknown
(3–12%)

*Cells are positive for myeloid antigen (e.g., CD13 and CD33).
†Cells are positive for a-naphthylacetate and platelet glycoprotein IIb/IIIa or factor VIII–related antigen and negative for naphthylbutyrate.

logic and genetic subtype must be identified. AML an essential part of the routine diagnostic workup
is a heterogeneous disease caused by a variety of of patients with AML. This combination of mor-
pathogenic mechanisms. At a morphologic level, this phologic, immunologic, and genetically based diag-
heterogeneity is manifested by variability in the de- nostic approaches not only makes it possible to
gree of commitment and differentiation of the cell modify therapy according to the sensitivity of bio-
lineage. This variability has been used to define spe- logically defined subtypes, but also provides unique
cific morphologic subgroups. The most commonly markers with which to monitor a patient’s response
used method of classification is that developed by to therapy.14
the French–American–British (FAB) group (Table
1),5-9 which divides AML into nine distinct subtypes MOLECULAR PATHOGENESIS
that differ with respect to the particular myeloid lin- The importance of specific cytogenetic lesions as
eage involved and the degree of leukemic-cell differ- powerful determinants of the therapeutic response
entiation. This distinction is based on the morpholog- suggests that the mechanisms of transformation as-
ic appearance of the blasts (Fig. 1) and their reactivity sociated with these lesions is likely directly to influ-
with histochemical stains, including myeloperoxidase, ence the sensitivity of the leukemic blasts to ther-
Sudan black, and the nonspecific esterases a-naph- apeutic agents. The implication is that if we can
thylacetate and naphthylbutyrate. understand why certain genetic lesions are associat-
In addition, immunologic methods have been in- ed with a favorable outcome, we may be able to ap-
corporated into the diagnostic criteria for some FAB ply this knowledge to improve the therapeutic ap-
subgroups7-10 (Table 1). Cytogenetic analysis of leu- proach and, ultimately, the outcome among patients
kemic blasts has resulted in the identification of with AML. The recent identification of the genetic
nonrandom clonal chromosomal aberrations in a targets of common AML-associated cytogenetic ab-
large percentage of patients with AML.11-13 Some of normalities and the elucidation of their mechanisms
these lesions correlate with specific FAB subtypes. of action have begun to provide critical insights into
Moreover, several cytogenetic lesions can be used to this issue. This approach has been best exemplified
identify subgroups of patients with distinct clinical by the highly successful therapeutic use of the dif-
features and therapeutic responses. Thus, cytogenet- ferentiation-inducing agent all-trans-retinoic acid.15-22
ic or direct molecular genetic methods have become All-trans-retinoic acid targets the chimeric protein en-

1052 · S ep tem b er 3 0 , 19 9 9
 MED IC A L PROGR ES S

A B

C D E

Figure 1. Cytologic Findings in Bone Marrow Specimens and Peripheral-Blood Smears from a Patient with Subtype M2 AML and
the t(8;21)(q22;q22) Translocation.
In Panel A, a bone marrow specimen contains medium-sized blasts, cytoplasm with no granulation, and nucleoli that are sometimes
clearly visible (May–Grünwald–Giemsa, ¬1600). Panel B shows a bone marrow specimen with myeloperoxidase-stained blasts
(¬1600). Panel C shows a leukemic blast with an Auer body (arrow) (May–Grünwald–Giemsa, ¬1600). Panel D shows a blast stained
with May–Grünwald–Giemsa in a peripheral-blood smear (¬1000). Panel E shows the results of fluorescence in situ hybridization
of the cell shown in Panel D with probes specific for the breakpoint regions of chromosome 8(q22) (isolated green spot) and chro-
mosome 21(q22) (isolated magenta spot) (¬1000). The arrow indicates the chromosomal fusion (clustered green and magenta
spots).

coded by the t(15;17) translocation associated with netic lesions has provided critical insights into the
acute promyelocytic leukemia.23 pathogenesis of AML and has already helped to iden-
The most common targets of AML-associated tify subgroups for therapeutic purposes.
chromosomal translocations are genes that encode
DNA-binding transcription factors or the regulatory ALTERATIONS OF AML1-CBFb
components of transcriptional complexes.24 Trans- Cloning of the AML-associated t(8;21) transloca-
formation in each of these cases appears to result tion led to the identification of AML1, which en-
from the generation of fusion proteins that interfere codes the DNA-binding subunit of AML1-CBFb, a
in a dominant manner with the function of the wild- transcription factor that regulates a number of hem-
type protein. Study of three specific molecular ge- atopoiesis-specific genes and is essential for normal

Vol ume 341 Numb e r 14 · 1053

 The New Eng land Jour nal of Medicine

Coactivators
CBFb
AML1
Target genes
---TGTGGT---

B t(8;21)

Corepressor
Complex
CBFb
AML1 ETO
Target genes
---TGTGGT---

Inv(16)

AML1 CBFb MYH11

AML1 CBFb MYH11

Target genes
---TGTGGT---

Figure 2. The AML1-CBFb Transcription Factor.

In Panel A, in normal cells, heterodimeric AML1-CBFb transcription-factor complex binds to the
DNA sequence TGTGGT in the transcriptional regulatory region of AML1-regulated target genes
and activates transcription through the recruitment of coactivators. In Panel B, in AML cells with
the t(8;21) translocation, the N-terminal part of AML1 fuses with the C-terminal portion of ETO.
The resultant chimeric protein continues to interact with CBFb and to bind to the core enhancer
sequence; however, ETO recruits a nuclear corepressor complex and results in the dominant re-
pression of AML1-regulated target genes. Similarly, the CBFb -MYH11 chimeric protein encoded by
the inv(16) mutation continues to interact with AML1; however, instead of allowing AML1 to inter-
act with DNA, this chimeric protein recruits AML1 into functionally inactive complexes in the cy-
toplasm.

development of the hematopoietic system.25-31 Some- number of rare translocations in AML, making it
what surprising was the observation that the AML- the most frequent target of chromosomal rearrange-
associated chromosomal rearrangement inv(16) or ments in human leukemia.
its variant t(16;16) targets CBFb, the other subunit The t(8;21) translocation is found in approxi-
of this transcription-factor complex.32 More recent- mately 40 percent of patients with FAB subtype M2
ly, AML1-CBFb has been found to be the target of AML, but it is not restricted to this subtype. The
the t(12;21) translocation in pediatric ALL and a fusion gene created by this translocation joins the

1054 · S ep tem b er 3 0 , 19 9 9
 MED IC A L PROGR ES S

N-terminal part of AML1, including the DNA-bind- tations of the maturation subdomain of the receptor
ing and CBFb-interaction domains, with the C-ter- for granulocyte colony-stimulating factor (G-CSF).46-48
minal portion of the 821 gene (ETO) on chromo- It is important to emphasize that cellular transfor-
some 8 (Fig. 2).25,26 Although the resultant protein mation is a multistep process and the abnormalities
retains the ability to bind AML1-regulated target se- discussed above are insufficient by themselves to lead
quences, it does not activate transcription, but instead to leukemia. Thus, cooperating molecular genetic ab-
dominantly represses AML1-mediated activation.33-36 normalities are required. In AML, the nature of these
Transcriptional repression appears to be mediated lesions remains poorly defined. A second point to
through the direct interaction of ETO with the nu- remember is that slightly more than half of all cases
clear corepressor complex.37 of AML involve chromosomal rearrangements. In the
The inv(16)(p13;q22) and the t(16;16)(p13;q22) remaining cases, the underlying molecular genetic
mutations are mainly (but not only) seen in patients abnormalities remain to be identified.
with FAB subtype M4Eo AML.38-41 In these chro-
mosomal rearrangements, the CBFb subunit of the PROGNOSTIC FACTORS
core binding-factor complex on chromosome 16q22 A number of clinical and biologic features that re-
is fused to the smooth-muscle myosin heavy-chain flect the heterogeneity of AML are used to predict
gene MYH11 on chromosome 16p13.32 In the result- the likelihood that a patient will have a response to
ing CBFb-MYH11 chimeric product, the N-termi- treatment.49 Adverse prognostic factors include an
nal portion of CBFb, including its AML1-interaction age over 60 years, a poor performance score before
domain, is fused in-frame to a variable amount of treatment, AML resulting from prior chemotherapy
the C-terminal domain of MYH11. CBFb-MYH11 or an antecedent hematologic disorder such as MDS,
directly represses AML1-mediated transcriptional ac- and a white-cell count of more than 20,000 per cu-
tivation, in part, by sequestering AML1 into function- bic millimeter or an elevated serum lactate dehydro-
ally inactive complexes within the cytoplasm (Fig. 2).42 genase level at presentation (Table 2). Furthermore,
an assessment for multidrug resistance and immuno-
AML WITH ALTERATIONS OF THE phenotyping may provide prognostic information.
MIXED-LINEAGE LEUKEMIA GENE Detailed cytogenetic analysis of the leukemic blasts
Structural alterations involving band q23 of chro- has also been demonstrated to provide critical prog-
mosome 11 are common in patients with AML and nostic information. Although there are correlations
account for approximately 6 to 8 percent of primary between certain FAB subtypes and cytogenetic ab-
cases and up to 85 percent of secondary cases of leu- normalities, such as between FAB subtype M3 AML
kemia that develop after exposure to topoisomerase and the t(15;17) translocation, the abnormalities
II inhibitors. This chromosomal abnormality is seen themselves appear to be the more important prog-
in all FAB subtypes, but predominantly in patients nostic factor.
with M4 or M5 AML. Although more than 30 dif- Combining these clinical and laboratory data has
ferent chromosomal loci can participate in these allowed the subdivision of AML into three broad
11q23 translocations, most sites involve 6q27, 9p22,
10p12, 17q21, or 19p13.1.43,44 Usually, as the result
of these translocations, a chimera is formed that con-
sists of the 5' portion of the mixed-lineage leukemia,
or MLL, gene, fused to the 3' portion of a gene en- TABLE 2. ADVERSE PROGNOSTIC FACTORS IN PATIENTS
coded on the reciprocal chromosome. The structure WITH AML.
of MLL suggests that its normal function is likely to
be mediated at the level of DNA or DNA-associated FACTORS USED TO PREDICT RESPONSE TO
chromatin proteins. In addition to its activity in the INDUCTION CHEMOTHERAPY FACTORS USED TO PREDICT RELAPSE
regulation of gene transcription, MLL also directly Unfavorable karyotype Unfavorable karyotype
interacts with a putative antiphosphatase called Sbfl,45 Age >60 yr Age >60 yr
which acts as a positive regulator in kinase signaling Secondary AML Delayed response to induction
pathways. Poor performance score* chemotherapy
Taken together, the information on the mechanisms Features of multidrug resistance Features of multidrug resistance
of transformation induced by the t(15;17), t(8;21), White-cell count of >20,000/mm3 White-cell count of >20,000/mm3
Female sex
inv(16), and MLL rearrangements suggests that induc- Unfavorable immunophenotype
CD34 positivity† Elevated lactate dehydrogenase
tion of AML often results from alterations in tran- level
scriptional cascades that are normally involved in reg- Autonomous growth of leukemic
ulating decisions regarding the fate of cells. Other cells
mechanisms that have been identified, although sub- *The extent of a patient’s disabilities are assessed according to a well-
stantially less frequently, involve alterations of growth defined set of criteria.50
factor–signaling pathways including structural mu- †Unlike the other factors listed, this factor is considered to be minor.

Vol ume 341 Numb e r 14 · 1055

 The New Eng land Jour nal of Medicine

prognostic groups: favorable, standard (or interme- non-Hodgkin’s lymphoma, ovarian cancer, breast can-
diate), and unfavorable. Although there may be sub- cer, or multiple myeloma.56 The risk of this compli-
tle differences in the criteria used to define these cation peaks 5 to 10 years after the start of chemo-
groups, the prognostic discrimination made possible therapy. These patients frequently present with MDS,
by the presence of various cytogenetic abnormalities which may then progress to overt AML.57-59 Such a
has become more important as the efficacy of treat- course is often associated with deletions of chromo-
ment for AML has improved.51-54 somes 5 and 7. The prognosis for these patients is
The favorable prognostic subgroup, which in- considerably worse than that for patients with pri-
cludes approximately 20 percent of cases among pa- mary AML.
tients who are 60 years of age or younger, is defined A second distinct subtype of therapy-induced AML
by the presence of leukemic blasts with the t(15;17), has been identified as a complication of treatment
t(8;21), or inv(16) mutation or molecular evidence with certain regimens of topoisomerase II inhibi-
of these abnormalities. These mutations are more tors, such as the epipodophyllotoxins.60 In contrast
frequent in younger patients who have high (more to alkylating-agent–induced secondary AML, this
than 85 percent) rates of complete remission and a type develops after a relatively short latency period
relatively low risk of relapse (30 to 40 percent). (two to three years), is not preceded by MDS, and
At the other end of the spectrum is the unfavor- is frequently associated with 11q23 chromosomal
able prognostic subgroup, which includes approxi- abnormalities.
mately 15 percent of the cases among patients who
are 15 to 60 years of age. These unfavorable cases are TREATMENT
defined by the presence of leukemic blasts with cy- The primary objective in treating patients with
togenetic abnormalities involving more than two chro- AML is to induce remission and thereafter prevent
mosomes, monosomies of chromosome 5 or 7, dele- relapse. Remission is conventionally defined mor-
tion of the long arm of 5 (del5q), or abnormalities phologically by the presence of fewer than 5 percent
of the long arm of chromosome 3. These abnormal- blasts in bone marrow together with the recovery of
ities are more frequent in older patients and in pa- peripheral-blood counts. More sensitive immuno-
tients with secondary AML, but even among young- logic and molecular genetic methods are now avail-
er patients, the survival rate is less than 20 percent able, which should be able to characterize remission
at five years. They represent a considerable therapeu- status more accurately; however, they have not yet
tic challenge for which no current treatment approach been extensively validated clinically. Treatment is con-
— including transplantation — is satisfactory. ventionally divided into two phases: induction and
Between these two groups are patients who are postinduction.
characterized as having a standard (or intermediate)
risk of relapse. The leukemic blasts of these patients Induction of Remission
have either a normal karyotype or cytogenetic ab- For more than 30 years, daunorubicin and cytar-
normalities that are not included in the definition of abine have been the backbone of treatments to induce
the other subgroups. In some series this includes pa- remission. Conventionally, daunorubicin is adminis-
tients with cytogenetic abnormalities of 11q23, where- tered three times at a dose of 40 to 60 mg per square
as in others these patients are included in the unfa- meter of body-surface area during each course of
vorable prognostic subgroup. Patients who are older chemotherapy. In recent years, prospective, random-
than 60 years generally have a poor prognosis, with ized trials of alternative agents have suggested that
a probability of survival at five years of less than 10 idarubicin61-63 or mitoxantrone64 is more effective
percent.55 than daunorubicin in younger patients, although both
resulted in more prolonged cytopenia. Therefore,
SECONDARY AML the question was raised as to whether the doses used
The majority of patients have no risk factors or ex- in these comparisons were equivalent in terms of
posures that could account for the development of levels of toxicity.65 Studies directly comparing mito-
the disease and thus are considered to have primary xantrone and idarubicin are ongoing.
AML. Secondary AML may develop in patients with In most induction regimens, cytarabine is given
a hematologic disorder (e.g., severe congenital neu- intravenously in bolus doses of 100 to 200 mg per
tropenia) or an inherited disease (e.g., Bloom’s syn- square meter per day or by continuous infusion over
drome and Fanconi’s anemia), in patients who have a period of 7 to 10 days. Several groups have sug-
had MDS for at least three months, or in those who gested that escalation of the dose during this period
have been exposed to leukemogenic agents, often as would be more effective than conventional dosing
a component of therapy for an unrelated neoplasm. strategies. Two randomized trials have confirmed
For example, AML can be expected to develop in this notion, although the benefit consisted of ex-
3 to 10 percent of patients who receive alkylating tension of disease-free survival among younger pa-
agents as part of their therapy for Hodgkin’s disease, tients who could tolerate the high doses used, and

1056 · S eptem b er 3 0 , 19 9 9
 MED IC A L PROGR ES S

it was more evident in the group with a favorable risk of relapse is 30 to 40 percent.54 High-risk pa-
prognosis.53,54,66 tients do less well after transplantation than those at
With the use of daunorubicin and cytarabine or low or moderate risk, and the limited comparative
their analogues, complete remission can be routinely data available do not always show a benefit after al-
induced in 70 to 80 percent of patients who are 60 logeneic transplantation.
years of age or younger and in approximately 50 per-
Autologous Bone Marrow Transplantation
cent of older patients. There is some evidence that
the addition of etoposide to combinations of dauno- Myeloablative treatment supported by autologous
rubicin and cytarabine can further increase remis- stem-cell transplantation has been widely used in re-
sion rates.67 The use of high-dose cytarabine (3 g per cent years, particularly in Europe. Several single-
square meter twice a day) did not increase the rate center series and registry data from nonrandomized
of remission,68,69 but in one randomized study it fa- studies indicate survival rates of 45 to 55 percent.76-79
vorably influenced relapse and survival.68 Because of concern about generalized selection bi-
as, several large collaborative trials have been un-
Postinduction Therapy dertaken. Although there were variations in study
Once remission is induced, further intensive treat- design, the main objective was prospectively to com-
ment of patients with AML is essential to prevent re- pare autologous bone marrow transplantation alone
lapse. Three options are available for younger patients: or in addition to intensive chemotherapy with in-
allogeneic bone marrow transplantation from an HLA- tensive chemotherapy alone; patients for whom do-
matched related or unrelated donor, autologous bone nors were available underwent allogeneic bone mar-
marrow transplantation, or chemotherapy. row transplantation. The French,80 European,81 and
British82 studies all reported a reduced risk of re-
Allogeneic Bone Marrow Transplantation
lapse among adults who underwent autologous bone
Allogeneic bone marrow transplantation from an marrow transplantation. In spite of a higher mortal-
HLA-matched sibling has been established practice ity rate (3 to 15 percentage points higher than the
for 15 to 20 years and can cure 50 to 60 percent of rate among patients who underwent allogeneic bone
recipients.70-72 It is the most active antileukemic treat- marrow transplantation), disease-free survival was
ment currently available. The risk of relapse among also improved in two of the studies.81,82 Overall sur-
patients in first complete remission who receive an vival did not differ significantly, because salvage
HLA-matched transplant from a sibling is generally therapy with transplantation after relapse was possi-
less than 20 percent. The reduced relapse rate is the ble in the case of some patients in the chemothera-
result not only of the use of marrow-ablative high- py group.81
dose cytotoxic therapy before bone marrow trans- In a study conducted by the Pediatric Oncology
plantation, but also of the allogeneic effect mediated Group,83 the risk of relapse was reduced among chil-
by the graft against residual leukemia in the host dren who underwent allogeneic bone marrow trans-
(graft-versus-leukemia effect). However, this favor- plantation, but this was counterbalanced by a high
able effect is partially offset by the toxicity of treat- risk of death (15 percent), so that the disease-free
ment and mortality related to the complications of survival and overall survival were not improved. The
immunosuppression (e.g., infections with cytomeg- U.S. collaborative trial73 was of a similar design and
alovirus and Epstein–Barr virus) and graft-versus-host found no significant difference in disease-free surviv-
disease. Because of the possibility of graft-versus- al between allogeneic or autologous transplantation
host disease, allogeneic bone marrow transplantation and intensive chemotherapy with high-dose cytara-
is usually restricted to patients under 55 years of age. bine. Because the survival rate after relapse was bet-
No randomized comparison of allogeneic bone ter in the chemotherapy group than in the trans-
marrow transplantation with chemotherapy has been plantation group, the overall survival was also better
done, but patients with suitable donors have been in the chemotherapy group. A typical feature of all
compared with those without donors. Several stud- these trials was that only a minority of patients who
ies have reported the beneficial effect of allogeneic were in remission and could have undergone trans-
transplantation in this type of analysis. In recent years, plantation actually did so.
the use of more intensive regimens of chemotherapy Relapse remains a problem that is partly account-
has improved the results in younger patients enough ed for by the presence of residual disease in the ab-
so that in some studies66,73-75 there was no overall sur- sence of a graft-versus-leukemia effect and partly by
vival benefit for the group with donors, despite the contamination of the autograft with leukemic cells.
fact that there was a lower risk of relapse. Much energy has been devoted to finding ways to
In addition, there is increasing recognition of the purge the marrow graft of contaminating cells ex vi-
prognostic profile of patients in relation to the risk vo.79 Although gene-marking studies have demon-
of relapse.49 Transplantation is probably unnecessary strated that the autograft itself can contribute to the
in low-risk patients in first remission, for whom the risk of relapse,84 there are no comparative clinical

Vol ume 341 Numb e r 14 · 1057

 The New Eng land Jour nal of Medicine

data to confirm that techniques employed to purge notypically HLA-matched unrelated donors. Wheth-
the autograft ex vivo are effective.79 er these patients should first receive induction ther-
apy or immediately undergo transplantation has not
Chemotherapy
been settled. Some patients may not enter a second
A common feature of two of the studies in which remission and are therefore deprived of the option
no benefit of allogeneic bone marrow transplanta- of bone marrow transplantation.
tion was shown was that the chemotherapy group One study has prospectively assessed the option of
received at least one course of high-dose cytarabine. offering bone marrow transplantation as immediate
A number of nonrandomized studies have suggested treatment after relapse88 and found that survival was
that a marked dose escalation would improve effica- similar to that for transplantation after chemotherapy.
cy, despite the pharmacologic evidence that the in- However, the logistics of arranging a transplantation
tracellular concentration of the active drug metabo- on short notice could be problematic. Currently, the
lites reached the saturation point at doses of more survival rate after either autologous transplantation
than 0.5 or 1.0 g per square meter per injection.85 or allogeneic transplantation with an HLA-matched
The landmark study of this approach was the Cancer donor for patients with AML in first relapse or sec-
and Leukemia Group B (CALGB) study,86 which ond remission is about 30 percent.89,90 Experience
demonstrated that in patients with AML in remis- with the use of transplants from phenotypically HLA-
sion, four courses of cytarabine at a dose of 3 g per matched unrelated donors or partially matched relat-
square meter twice daily for three out of five days ed donors is still limited.91
was superior to equivalent courses of 400 or 100 mg
per square meter given as a continuous infusion over OLDER PATIENTS WITH AML
a period of five days. The overall survival rate four More than three fourths of patients with AML
years after randomization was 46 percent in the high- are older than 60 years. Only in recent years have
dose group. Another study of high-dose cytarabine68 there been studies focused on these patients. In this
reported similar results, although in that study the age group, there is an uneven distribution of unfa-
treatment was used to induce remission. vorable prognostic factors (e.g., cytogenetic abnor-
Thus, it would appear that there is strong evi- malities, features of drug resistance, or a history of
dence of a dose–response effect of cytarabine in pa- MDS).55,92,93 In addition, older patients cannot tol-
tients with AML — even in high-risk patients.87 This erate intensive chemotherapy well and often have in-
approach, however, can be tolerated only by younger tercurrent medical conditions that are exacerbated by
patients. Many questions remain to be resolved, such cancer chemotherapy or its sequelae. Withholding
as the optimal dose, number of doses per course, induction chemotherapy generally results in low sur-
and number of courses. There may well be different vival rates and a poor quality of life.94 Currently, pa-
levels of benefit in the different risk groups. The re- tients older than 60 years of age who have a good
sults of the CALGB study suggest that those with performance status and meet the medical criteria of
favorable cytogenetic characteristics will benefit most. adequate organ function are usually offered induction
The comparative value of high-dose cytarabine reg- chemotherapy and have an overall probability of com-
imens and autologous transplantation in the various plete remission of 50 percent.55,92,93 Among those with
prognostic subgroups remains the subject of study. a complete response, approximately 20 percent sur-
vive free of leukemia for at least two years. Patients
RELAPSE who have cytogenetic abnormalities and a high white-
When treatment fails in patients with AML, the cell count at presentation, who are more than 80
available options are dictated by age, duration of the years of age, and who are in poor general physical
first remission, and cytogenetic findings, among oth- condition55 or have drug-resistance phenotypes92
er factors.49 Patients with favorable cytogenetic char- have a low likelihood of complete remission (rates of
acteristics — that is, a t(15;17), t(8;21), or inv(16) complete response are less than 30 percent).
mutation — who were in remission for more than Prognostic factors in the elderly that are associat-
one year before relapse have an approximately 20 ed with overall survival rates of 20 percent or more
percent chance of survival after subsequent therapy. at three years include good physical condition, an age
Since all other groups have a poor response, the of 80 years or less, primary rather than secondary
highest priority should be to prevent the first re- AML, the absence of cytogenetic abnormalities, and
lapse. For children and younger adults who have a the absence of leukocytosis at diagnosis.55 There is
first relapse or those who do not have a complete re- some evidence that the use of low-dose maintenance
sponse to first-line induction therapy, the recom- chemotherapy (for instance, with low-dose cytara-
mended option is marrow-ablative (high-dose) cyto- bine) for several months after the induction of re-
toxic treatment followed by hematopoietic stem-cell mission reduces the probability of relapse.55 High-
transplantation, including autografts or allografts from dose chemotherapy is highly unlikely to improve the
genotypically HLA-matched related donors or phe- clinical outcome in older patients.86

1058 · S eptem b er 3 0 , 19 9 9
 MED IC A L PROGR ES S

Clearly, new approaches to therapy are needed to cifically, the targeting of the malignant cells with
improve the cure rates in this large cohort of pa- molecular and immunologic therapeutic strategies.
tients. In the meantime, the wisest course is to offer The development of new drugs and treatment strat-
induction therapy only to patients in adequate phys- egies and the circumvention of resistance mechanisms
ical condition and to proceed further only if they are major targets of current clinical investigation.
have a response to the first cycle of treatment with- Cross resistance of drugs to structurally unrelated
out serious adverse effects. cytotoxic agents — pleiotropic, or multidrug, resist-
ance — is common in patients with refractory leu-
USE OF HEMATOPOIETIC GROWTH kemia.110,111 Classic multidrug resistance (governed
FACTORS by the MDR1 gene) is associated with the expres-
The incidence of death from bacterial and fungal sion of the membrane marker P-glycoprotein. This
infections during and after induction therapy among molecule transports antileukemic drugs (e.g., an-
patients with hypoplasia generally ranges from 15 thracyclines and etoposide) out of the plasma mem-
percent to 25 percent among adults with AML and brane, so that high levels of expression of MDR1
increases with age. Thus, the use of hematopoietic have been associated with reduced intracellular con-
growth factors to accelerate hematopoietic recovery centrations of chemotherapeutic agents in tumor cells.
and prevent infection has attracted wide attention.95,96 Other genes involved in the mechanisms of resist-
G-CSF and granulocyte–macrophage colony-stimu- ance to chemotherapy and serving as predictors of
lating factor (GM-CSF) can stimulate the produc- treatment response are MRP, which codes for mul-
tion and activation of granulocytes and monocytes tidrug-resistance–associated protein, a transporter of
(in the case of GM-CSF) and promote their mobili- the glutathione complex,112 and LRP, which encodes
zation from the marrow to the blood circulation.97 the lung resistance protein.113-115 Although the mo-
Thrombopoietin and several other cytokines have lecular pathways leading to the development of drug
also become available for clinical studies in patients resistance in patients with AML remain largely un-
with AML. A substantial number of randomized known, drugs that reverse or abrogate resistance are
studies have evaluated the use of G-CSF or GM-CSF being developed.116,117 Phase III studies of competi-
as an adjunct to induction or consolidation cycles of tive inhibitors of P-glycoprotein (e.g., cyclosporine
chemotherapy.98-106 The duration of neutropenia was and its analogues) have recently been initiated.
consistently shorter with the use of either cytokine.
This benefit translated into fewer days of antibiot- REFERENCES
ic99,103 or antifungal106 therapy or fewer days of hos- 1. Kosary CL, Ries LAG, Miller BA, Hankey BF, Edwards BK, eds. SEER
pitalization106 in a minority of studies. None of the cancer statistics review, 1973-1992: tables and graphs. Bethesda, Md.: Na-
tional Cancer Institute, 1995. (NIH publication no. 96-2789.)
studies found that this approach reduced the num- 2. Catovsky D, Matutes E, Buccheri V, et al. A classification of acute leu-
ber of documented infections. One study reported kaemia for the 1990s. Ann Hematol 1991;62:16-21.
an increase in the rate of initial response,98 and an- 3. Ryan DH. Phenotypic heterogeneity in acute leukemia. Clin Chim Acta
1992;206:9-23.
other study reported that survival was increased.100 4. Matutes E, Pombo de Oliveira M, Foroni L, Morilla R, Catovsky D.
Although AML blast cells generally express func- The role of ultrastructural cytochemistry and monoclonal antibodies in
clarifying the nature of undifferentiated cells in acute leukaemia. Br J Hae-
tional G-CSF and GM-CSF receptors on their sur- matol 1988;69:205-11.
face,97 thus far, the fear that treatment with G-CSF 5. Bennett JM, Catovsky D, Daniel MT, et al. Proposals for the classifica-
or GM-CSF could provoke the growth of leukemic tion of the acute leukaemias. Br J Haematol 1976;33:451-8.
6. Idem. Proposed revised criteria for the classification of acute myeloid
cells in patients has not been realized. leukemia: a report of the French-American-British Cooperative Group.
In the light of these findings, neither G-CSF nor Ann Intern Med 1985;103:620-5.
GM-CSF has a standard role in the clinical care of 7. Idem. Criteria for the diagnosis of acute leukemia of megakaryocyte lin-
eage (M7): a report of the French-American-British Cooperative Group.
patients with AML. However, the use of these cyto- Ann Intern Med 1985;103:460-2.
kines might be justified in patients with serious in- 8. Bloomfield CD, Brunning RD. The revised French-American-British
classification of acute myeloid leukemia: is new better? Ann Intern Med
fections that do not respond to antimicrobial treat- 1985;103:614-6.
ment. A future role for myeloid colony-stimulating 9. Bennett JM, Catovsky D, Daniel MT, et al. Proposal for the recognition
factors in hematopoietic progenitor-cell mobilization of minimally differentiated acute myeloid leukaemia (AML-MO). Br J
Haematol 1991;78:325-9.
is suggested by studies in which autografts of pe- 10. Koike T. Megakaryoblastic leukemia: the characterization and identifi-
ripheral-blood progenitor cells (instead of marrow cation of megakaryoblasts. Blood 1984;64:683-92.
transplants) had accelerated rates of hematopoietic 11. Rowley JD. Chromosome changes in acute leukaemia. Br J Haematol
1980;44:339-46.
regeneration after mobilization with colony-stimu- 12. Yunis JJ. Recurrent chromosomal defects are found in most patients
lating factors and cytapheresis.107-109 with acute nonlymphocytic leukemia. Cancer Genet Cytogenet 1984;11:
125-37.
CONCLUSIONS 13. Bitter MA, Le Beau MM, Rowley JD, Larson RA, Golomb HM,
Vardiman JW. Associations between morphology, karyotype, and clinical
In general, the trend in the treatment of patients features in myeloid leukemias. Hum Pathol 1987;18:211-25.
14. Lo Coco F, Diverio D, Pandolfi PP, et al. Molecular evaluation of re-
with AML is toward the modification of therapy to sidual disease as a predictor of relapse in acute promyelocytic leukaemia.
treat specific subtypes of the disease and, more spe- Lancet 1992;340:1437-8.

Vol ume 341 Numb e r 14 · 1059

 The New Eng land Jour nal of Medicine

15. Grimwade D, Solomon E. Characterisation of the PML/RAR alpha 41. Costello R, Sainty D, Lecine P, et al. Detection of CBFb/MYH11 fu-
rearrangement associated with t(15;17) acute promyelocytic leukaemia. sion transcripts in acute myeloid leukemia: heterogeneity of cytological and
Curr Top Microbiol Immunol 1997;220:81-112. molecular characteristics. Leukemia 1997;11:644-50.
16. Chambon P. A decade of molecular biology of retinoic acid receptors. 42. Kanno Y, Kanno T, Sakakura C, Bae SC, Ito Y. Cytoplasmic seques-
FASEB J 1996;10:940-54. tration of the polyomavirus enhancer binding protein 2 (PEBP2)/core
17. Horlein AJ, Naar AM, Heinzel T, et al. Ligand-independent repression binding factor alpha (CBFalpha) subunit by the leukemia-related
by the thyroid hormone receptor mediated by a nuclear receptor co-repres- PEBP2/CBFbeta-SMMHC fusion protein inhibits PEBP2/CBF-mediated
sor. Nature 1995;377:397-404. transactivation. Mol Cell Biol 1998;18:4252-61.
18. Kurokawa R , Soderstrom M, Horlein A, et al. Polarity-specific activi- 43. Rubnitz JE, Behm FG, Downing JR. 11q23 Rearrangements in acute
ties of retinoic acid receptors determined by a co-repressor. Nature 1995; leukemia. Leukemia 1996;10:74-82.
377:451-4. 44. Waring PM, Cleary ML. Disruption of a homolog of trithorax by
19. Nagy L, Kao HY, Chakravarti D, et al. Nuclear receptor repression 11q23 translocations: leukemogenic and transcriptional implications. Curr
mediated by a complex containing SMRT, mSin3A, and histone deacety- Top Microbiol Immunol 1997;220:1-23.
lase. Cell 1997;89:373-80. 45. Cui X, De Vivo I, Slany R, Miyamoto A, Firestein R, Cleary ML. As-
20. Grignani F, De Matteis S, Nervi C, et al. Fusion proteins of the reti- sociation of SET domain and myotubularin-related proteins modulates
noic acid receptor-alpha recruit histone deacetylase in promyelocytic leu- growth control. Nat Genet 1998;18:331-7.
kaemia. Nature 1998;391:815-8. 46. Dong F, Brynes RK, Tidow N, Welte K, Löwenberg B, Touw IP. Mu-
21. Lin RJ, Nagy L, Inoue S, Shao W, Miller WH Jr, Evans RM. Role of tations in the gene for the granulocyte colony-stimulating–factor receptor
the histone deacetylase complex in acute promyelocytic leukaemia. Nature in patients with acute myeloid leukemia preceded by severe congenital neu-
1998;391:811-4. tropenia. N Engl J Med 1995;333:487-93.
22. He LZ, Guidez F, Tribioli C, et al. Distinct interactions of PML- 47. Dong F, Dale DC, Bonilla MA, et al. Mutations in the granulocyte
RARa and PLZF-RARa with co-repressors determine differential respons- colony-stimulating factor receptor gene in patients with severe congenital
es to RA in APL. Nat Genet 1998;18:126-35. neutropenia. Leukemia 1997;11:120-5.
23. Warrell RP Jr, de Thé H, Wang Z-Y, Degos L. Acute promyelocytic 48. Hermans MH, Ward AC, Antonissen C, Karis A, Lowenberg B, Touw
leukemia. N Engl J Med 1993;329:177-89. IP. Perturbed granulopoiesis in mice with a targeted mutation in the gran-
24. Look AT. Oncogenic transcription factors in the human acute leuke- ulocyte colony-stimulating factor receptor gene associated with severe
mias. Science 1997;278:1059-64. chronic neutropenia. Blood 1998;92:32-9.
25. Miyoshi H, Shimizu K, Kozu T, Maseki N, Kaneko Y, Ohki M. 49. van Putten WLJ, Löwenberg B. Prognostic factors in adult AML.
t(8;21) Breakpoints on chromosome 21 in acute myeloid leukemia are Blood 1997;90:Suppl 1:65a. abstract.
clustered within a limited region of a single gene, AML1. Proc Natl Acad 50. WHO Handbook for reporting results of cancer treatment. Geneva:
Sci U S A 1991;88:10431-4. World Health Organization, 1979.
26. Erickson P, Gao J, Chang KS, et al. Identification of breakpoints in 51. Yunis JJ, Brunning RD, Howe RB, Lobell M. High-resolution chro-
t(8;21) acute myelogenous leukemia and isolation of a fusion transcript, mosomes as an independent prognostic indicator in adult acute nonlym-
AML1/ETO, with similarity to Drosophila segmentation gene, runt. phocytic leukemia. N Engl J Med 1984;311:812-8.
Blood 1992;80:1825-31. 52. Keating MJ, Smith TL, Kantarjian H, et al. Cytogenetic pattern in
27. Daga A, Tighe JE, Calabi F. Leukemia/Drosophila homology. Nature acute myelogenous leukemia: a major reproducible determinant of out-
1992;356:484. come. Leukemia 1988;2:403-12.
28. Okuda T, van Deursen J, Hiebert SW, Grosveld G, Downing JR. 53. Mrozek K, Heinonen K, de la Chapelle A, Bloomfield CD. Clinical
AML1, the target of multiple chromosomal translocations in human leu- significance of cytogenetics in acute myeloid leukemia. Semin Oncol 1997;
kemia, is essential for normal fetal liver hematopoiesis. Cell 1996;84:321- 24:17-31.
30. 54. Grimwade D, Walker H, Oliver F, et al. The importance of diagnostic
29. Wang Q, Stacy T, Binder M, Marin-Padilla M, Sharpe AH, Speck NA. cytogenetics on outcome in AML: analysis of 1,612 patients entered into
Disruption of the Cbfa2 gene causes necrosis and hemorrhaging in the the MRC AML 10 trial. Blood 1998;92:2322-33.
central nervous system and blocks definitive hematopoiesis. Proc Natl Acad 55. Löwenberg B, Suciu S, Archimbaud E, et al. Mitoxantrone versus dau-
Sci U S A 1996;93:3444-9. norubicin in induction-consolidation chemotherapy — the value of low-
30. Wang Q, Stacy T, Miller JD, et al. The CBFb subunit is essential for dose cytarabine for maintenance of remission, and an assessment of prog-
CBFa2 (AML1) function in vivo. Cell 1996;87:697-708. nostic factors in acute myeloid leukemia in the elderly: final report of the
31. Sasaki K, Yagi H, Bronson RT, et al. Absence of fetal liver hematopoi- Leukemia Cooperative Group of the European Organization for the Re-
esis in mice deficient in transcriptional coactivator core binding factor beta. search and Treatment of Cancer and the Dutch-Belgian Hemato-Oncology
Proc Natl Acad Sci U S A 1996;93:12359-63. Cooperative Hovon Group Randomized Phase III Study AML-9. J Clin
32. Liu P, Tarle SA, Hajra A, et al. Fusion between transcription factor Oncol 1998;16:872-81.
CBFb/PEBP2b and a myosin heavy chain in acute myeloid leukemia. Sci- 56. van Leeuwen FE. Risk of acute myelogenous leukaemia and myelodys-
ence 1993;261:1041-4. plasia following cancer treatment. Baillieres Clin Haematol 1996;9:57-85.
33. Meyers S, Lenny N, Hiebert SW. The t(8;21) fusion protein interferes 57. Kyle RA, Pierre RV, Bayrd ED. Multiple myeloma and acute my-
with AML-1B-dependent transcriptional activation. Mol Cell Biol 1995; elomonocytic leukemia: report of four cases possibly related to melphalan.
15:1974-82. N Engl J Med 1970;283:1121-5.
34. Frank R , Zhang J, Uchida H, Meyers S, Hiebert SW, Nimer SD. The 58. Levine EG, Bloomfield CD. Leukemias and myelodysplastic syn-
AML1/ETO fusion protein blocks transactivation of the GM-CSF pro- dromes secondary to drug, radiation, and environmental exposure. Semin
moter by AML1B. Oncogene 1995;11:2667-74. Oncol 1992;19:47-84.
35. Westendorf JJ, Yamamoto CM, Lenny N, Downing JR, Selsted ME, 59. Pedersen-Bjergaard J. Radiotherapy- and chemotherapy-induced mye-
Hiebert SW. The t(8;21) fusion product, AML-1-ETO, associates with lodysplasia and acute myeloid leukemia: a review. Leuk Res 1992;16:61-5.
C/EBP-a, inhibits C/EBP-a-dependent transcription, and blocks granu- 60. Pui C-H, Ribeiro RC, Hancock ML, et al. Acute myeloid leukemia in
locytic differentiation. Mol Cell Biol 1998;18:322-33. children treated with epipodophyllotoxins for acute lymphoblastic leuke-
36. Lutterbach B, Sun D, Schuetz J, Hiebert SW. The MYND motif is re- mia. N Engl J Med 1991;325:1682-7.
quired for repression of basal transcription from the multidrug resistance 61. Berman E, Heller G, Santorsa J, et al. Results of a randomized trial
1 promoter by the t(8;21) fusion protein. Mol Cell Biol 1998;18:3604-11. comparing idarubicin and cytosine arabinoside with daunorubicin and cy-
37. Wang J, Hoshino T, Redner RL, Kajigaya S, Liu JM. Novel human nu- tosine arabinoside in adult patients with newly diagnosed acute myeloge-
clear receptor co-repressor: cloning and identification as a binding partner nous leukemia. Blood 1991;77:1666-74.
for the ETO proto-oncoprotein. Blood 1997;90:Suppl 1:245a. abstract. 62. Vogler WR, Velez-Garcia E, Weiner RS, et al. A phase III trial com-
38. Arthur DC, Bloomfield CD. Partial deletion of the long arm of chro- paring idarubicin and daunorubicin in combination with cytarabine in
mosome 16 and bone marrow eosinophilia in acute nonlymphocytic leu- acute myelogenous leukemia: a Southeastern Cancer Study Group study.
kemia: a new association. Blood 1983;61:994-8. J Clin Oncol 1992;10:1103-11.
39. Le Beau MM, Larson RA, Bitter MA, Vardiman JW, Golomb HM, 63. Wiernik PH, Banks PLC, Case DC Jr, et al. Cytarabine plus idarubicin
Rowley JD. Association of an inversion of chromosome 16 with abnormal or daunorubicin as induction and consolidation therapy for previously un-
marrow eosinophils in acute myelomonocytic leukemia: a unique cytoge- treated adult patients with acute myeloid leukemia. Blood 1992;79:313-9.
netic-clinicopathological association. N Engl J Med 1983;309:630-6. 64. Arlin Z, Case DC Jr, Moore J, et al. Randomized multicenter trial of
40. Shurtleff SA, Meyers S, Hiebert SW, et al. Heterogeneity in cytosine arabinoside with mitoxantrone or daunorubicin in previously un-
CBFb/MYH11 fusion messages encoded by the inv(16)(p13q22) and the treated adult patients with acute nonlymphocytic leukemia (ANLL). Leu-
t(16; 16)(p13;q22) in acute myelogenous leukemia. Blood 1995;85:3695- kemia 1990;4:177-83.
703. 65. The AML Collaborative Group. A systemic collaborative overview of

1060 · S eptem b er 3 0 , 19 9 9
 MED IC A L PROGR ES S

randomized trials comparing idarubicin with daunorubicin (or other an- plantation for patients with acute myeloid leukemia in untreated first re-
thracyclines) as induction therapy for acute myeloid leukaemia. Br J Hae- lapse or in second complete remission. J Clin Oncol 1993;11:1353-60.
matol 1998;103:100-9. 89. Clift RA, Buckner CD, Thomas ED, et al. The treatment of acute non-
66. Burnett AK, Goldstone AH, Stevens R, et al. Allo and auto BMT re- lymphoblastic leukemia by allogeneic marrow transplantation. Bone Mar-
duce relapse risk in AML in CR1 but do not significantly improve overall row Transplant 1987;2:243-58.
survival: results of the MRC AML 10 Trial. Br J Haematol 1996;93:Suppl 90. Allogeneic bone marrow transplantation for leukaemia in Europe: re-
2:313. abstract. port from the Working Party on Leukaemia, European Group for Bone
67. Bishop JF, Lowenthal PM, Joshua D, et al. Etoposide in acute non- Marrow Transplantation. Lancet 1988;1:1379-82.
lymphocytic leukemia. Blood 1990;75:27-32. 91. Kernan NA, Bartsch G, Ash RC, et al. Analysis of 462 transplantations
68. Bishop JF, Matthews JP, Young GA, et al. A randomized study of high- from unrelated donors facilitated by the National Marrow Donor Program.
dose cytarabine in induction in acute myeloid leukemia. Blood 1996;87: N Engl J Med 1993;328:593-602.
1710-7. 92. Leith CP, Kopecky KJ, Godwin J, et al. Acute myeloid leukemia in the
69. Weick JK, Kopecky KJ, Appelbaum FR, et al. A randomized investi- elderly: assessment of multidrug resistance (MDR1) and cytogenetics dis-
gation of high-dose versus standard-dose cytosine arabinoside with dauno- tinguishes biologic subgroups with remarkably distinct responses to stand-
rubicin in patients with previously untreated acute myeloid leukemia: a ard chemotherapy: a Southwest Oncology Group study. Blood 1997;89:
Southwest Oncology Group study. Blood 1996;88:2841-51. 3323-9.
70. Thomas ED, Buckner CD, Clift RA, et al. Marrow transplantation for 93. Löwenberg B. Treatment of the elderly patient with acute leukaemia.
acute nonlymphoblastic leukemia in first remission. N Engl J Med 1979; Baillieres Clin Haematol 1996;9:147-59.
301:597-9. 94. Löwenberg B, Zittoun R, Kerkhofs H, et al. On the value of intensive
71. Appelbaum FR, Dahlberg S, Thomas ED, et al. Bone marrow trans- remission-induction chemotherapy in elderly patients of 65+ years with
plantation or chemotherapy after remission induction for adults with acute acute myeloid leukemia: a randomized phase III study of the European Or-
nonlymphoblastic leukemia: a prospective comparison. Ann Intern Med ganization for Research and Treatment of Cancer Leukemia Group. J Clin
1984;101:581-8. Oncol 1989;7:1268-74.
72. Champlin RE, Ho WG, Gale RP, et al. Treatment of acute myeloge- 95. Terpstra WE, Löwenberg B. Application of myeloid growth factors in
nous leukemia: a prospective controlled trial of bone marrow transplanta- the treatment of acute myeloid leukemia. Leukemia 1997;11:315-27.
tion versus consolidation chemotherapy. Ann Intern Med 1985;102:285- 96. Schiffer CA. Hematopoietic growth factors as adjuncts to the treat-
91. ment of acute myeloid leukemia. Blood 1996;88:3675-85.
73. Cassileth PA, Harrington DP, Appelbaum FR, et al. Chemotherapy 97. Löwenberg B, Touw IP. Hematopoietic growth factors and their recep-
compared with autologous or allogeneic bone marrow transplantation in tors in acute leukemia. Blood 1993;81:281-92.
the management of acute myeloid leukemia in first remission. N Engl J 98. Dombret H, Chastang C, Fenaux P, et al. A controlled study of re-
Med 1998;339:1649-56. combinant human granulocyte colony-stimulating factor in elderly patients
74. Keating S, de Witte T, Suciu S, et al. The influence of HLA-matched after treatment for acute myelogenous leukemia. N Engl J Med 1995;332:
sibling donor availability on treatment outcome for patients with AML: an 1678-83.
analysis of the AML 8A study of the EORTC Leukaemia Cooperative 99. Godwin JE, Kopecky KJ, Head DR, et al. A double-blind placebo-
Group and GIMEMA. Br J Haematol 1998;102:1344-53. controlled trial of granulocyte colony-stimulating factor in elderly patients
75. Stevens RF, Hann IM, Wheatley K, Gray RG. Marked improvements with previously untreated acute myeloid leukemia: a Southwest Oncology
in outcome with chemotherapy alone in paediatric acute myeloid leu- Group study (9031). Blood 1998;91:3607-15.
kaemia: results of the United Kingdom Medical Research Council’s 10th 100. Rowe JM, Andersen JW, Mazza JJ, et al. A randomized placebo-con-
AML trial. Br J Haematol 1998;101:130-40. trolled phase III study of granulocyte-macrophage colony-stimulating fac-
76. Löwenberg B, Abels J, van Bekkum DW, et al. Transplantation of non- tor in adult patients (>55 to 70 years of age) with acute myelogenous leu-
purified autologous bone marrow in patients with AML in first remission. kemia: a study of the Eastern Cooperative Oncology Group (E1490).
Cancer 1984;54:2840-3. Blood 1995;86:457-62.
77. Burnett AK, Tansey P, Watkins R, et al. Transplantation of unpurged 101. Stone RM, Berg DT, George SL, et al. Granulocyte–macrophage
autologous bone-marrow in acute myeloid leukaemia in first remission. colony-stimulating factor after initial chemotherapy for elderly patients
Lancet 1984;2:1068-70. with primary acute myelogenous leukemia. N Engl J Med 1995;332:1671-
78. Löwenberg B, Verdonck LJ, Dekker AW, et al. Autologous bone mar- 7.
row transplantation in acute myeloid leukemia in first remission: results of 102. Löwenberg B, Boogaerts MA, Daenen SMGJ, et al. Value of different
a Dutch prospective study. J Clin Oncol 1990;8:287-94. modalities of granulocyte-macrophage colony-stimulating factor applied
79. Gorin NC, Labopin M, Meloni G, et al. Autologous bone marrow during or after induction therapy of acute myeloid leukemia. J Clin Oncol
transplantation for acute myeloblastic leukemia in Europe: further evi- 1997;15:3496-506.
dence of the role of marrow purging by mafosfamide. Leukemia 1991;5: 103. Witz F, Sadoun A, Perrin MC, et al. A placebo-controlled study of
896-904. recombinant human granulocyte-macrophage colony-stimulating factor
80. Harousseau J-L, Cahn J-Y, Pignon B, et al. Comparison of autologous administered during and after induction treatment for de novo acute mye-
bone marrow transplantation and intensive chemotherapy as postremission logenous leukemia in elderly patients. Blood 1998;91:2722-30.
therapy in adult acute myeloid leukemia. Blood 1997;90:2978-86. 104. Zittoun R, Suciu S, Mandelli F, et al. Granulocyte-macrophage col-
81. Zittoun RA, Mandelli F, Willemze R, et al. Autologous or allogeneic ony-stimulating factor associated with induction treatment of acute mye-
bone marrow transplantation compared with intensive chemotherapy in logenous leukemia: a randomized trial by the European Organization for
acute myelogenous leukemia. N Engl J Med 1995;332:217-23. Research and Treatment of Cancer Leukemia Cooperative Group. J Clin
82. Burnett AK, Goldstone AH, Stevens RMF, et al. Randomised compar- Oncol 1996;14:2150-9.
ison of addition of autologous bone-marrow transplantation to intensive 105. Löwenberg B, Suciu S, Archimbaud E, et al. Use of recombinant
chemotherapy for acute myeloid leukaemia in first remission: results of granulocyte-macrophage colony-stimulating factor during and after remis-
MRC AML 10 trial. Lancet 1998;351:700-8. sion induction chemotherapy in patients aged 61 years and older with
83. Ravindranath Y, Yeager AM, Chang MN, et al. Autologous bone mar- acute myeloid leukemia (AML): final report of AML-11, a phase III ran-
row transplantation versus intensive consolidation chemotherapy for acute domized study of the Leukemia Cooperative Group of the European Or-
myeloid leukemia in childhood. N Engl J Med 1996;334:1428-34. ganisation for the Research and Treatment of Cancer (EORTC-LCG) and
84. Brenner MK, Rill DR, Moen RC, et al. Gene-marking to trace origin the Dutch Belgian Hemato-Oncology Cooperative Group (HOVON).
of relapse after autologous bone-marrow transplantation. Lancet 1993;341: Blood 1997;90:2952-61.
85-6. 106. Heil G, Hoelzer D, Sanz MA, et al. A randomized, double-blind,
85. Plunkett W, Liliemark JO, Adams TM, et al. Saturation of 1-b-D-ara- placebo-controlled, phase III study of filgrastim in remission induction and
binofuranosylcytosine 5'-triphosphate accumulation in leukemia cells dur- consolidation therapy for adults with de novo acute myeloid leukemia.
ing high-dose 1-b-D-arabinofuranosylcytosine therapy. Cancer Res 1987; Blood 1997;90:4710-8.
47:3005-11. 107. Sanz MA, de la Rubia J, Sanz GF, et al. Busulfan and cyclophospha-
86. Mayer RJ, Davies RB, Schiffer CA, et al. Intensive postremission che- mide followed by autologous blood stem-cell transplantation for patients
motherapy in adults with acute myeloid leukemia. N Engl J Med 1994; with acute myeloblastic leukemia in first complete remission: a report from
331:896-903. a single institution. J Clin Oncol 1993;11:1661-7.
87. Kern W, Aul C, Maschmeyer G, et al. Superiority of high-dose over 108. Demirer T, Petersen FB, Bensinger WI, et al. Autologous transplan-
intermediate-dose cytosine arabinoside in the treatment of patients with tation with peripheral blood stem cells collected after granulocyte colony-
high-risk acute myeloid leukemia: results of an age-adjusted prospective stimulating factor in patients with acute myelogenous leukemia. Bone Mar-
randomized comparison. Leukemia 1998;12:1049-55. row Transplant 1996;18:29-34.
88. Petersen FB, Lynch MHE, Clift RA, et al. Autologous marrow trans- 109. Vellenga E, van Putten WLJ, Boogaerts MA, et al. Peripheral blood

Vol ume 341 Numb e r 14 · 1061

 The New Eng land Jour nal of Medicine

stem cell transplantation as an alternative to autologous marrow transplan- 114. List AF, Spier CS, Grogan TM, et al. Overexpression of the major
tation in the treatment of acute myeloid leukemia? Bone Marrow Trans- vault transporter protein lung-resistance protein predicts treatment out-
plant 1999;23:1279-82. come in acute myeloid leukemia. Blood 1996;87:2464-9.
110. Löwenberg B, Sonneveld P. Resistance to chemotherapy in acute leu- 115. Filipits M, Pohl G, Stranzl T, et al. Expression of the lung resistance
kemia. Curr Opin Oncol 1998;10:31-5. protein predicts poor outcome in de novo acute myeloid leukemia. Blood
111. Sonneveld P. Multidrug resistance in acute myeloid leukaemia. 1998;91:1508-13.
Baillieres Clin Haematol 1996;9:185-203. 116. Solary E, Witz B, Caillot D, et al. Combination of quinine as a po-
112. Cole SP, Bhardwaj G, Gerlach JH, et al. Overexpression of a trans- tential reversing agent with mitoxantrone and cytarabine for the treatment
porter gene in a multidrug-resistant human lung cancer cell line. Science of acute leukemias: a randomized multicenter study. Blood 1996;88:1198-
1992;258:1650-4. 205.
113. Scheper RJ, Broxterman HJ, Scheffer GL, et al. Overexpression of a 117. List AF, Spier C, Greer J, et al. Phase I/II trial of cyclosporin as a
M(r) 110,000 vesicular protein in non-P-glycoprotein-mediated multidrug chemotherapy-resistance modifier in acute leukemia. J Clin Oncol 1993;11:
resistance. Cancer Res 1993;53:1475-9. 1652-60.

1062 · S eptem b er 3 0 , 19 9 9