Estimated new cases and deaths from AML in the United States in 2017: 
Advances in the treatment of AML (also called acute myelogenous leukemia, acute nonlymphocytic leukemia [ANLL]) have resulted in substantially improved complete remission (CR) rates.  Treatment should be sufficiently aggressive to achieve CR because partial remission offers no substantial survival benefit. Approximately 60% to 70% of adults with AML can be expected to attain CR status following appropriate induction therapy. More than 25% of adults with AML (about 45% of those who attain CR) can be expected to survive 3 or more years and may be cured. Remission rates in adult AML are inversely related to age, with an expected remission rate of more than 65% for those younger than 60 years. Data suggest that once attained, duration of remission may be shorter in older patients. Increased morbidity and mortality during induction appear to be directly related to age. Other adverse prognostic factors include central nervous system involvement with leukemia, systemic infection at diagnosis, elevated white blood cell count (>100,000/mm3), treatment-induced AML, and history of myelodysplastic syndromes or another antecedent hematological disorder. Patients with leukemias that express the progenitor cell antigen CD34 and/or the P-glycoprotein (MDR1 gene product) have an inferior outcome.    AML associated with an internal tandem duplication of the FLT3 gene (FLT3/ITD mutation) has an inferior outcome that is attributed to a higher relapse rate.  
Cytogenetic analysis provides some of the strongest prognostic information available, predicting outcome of both remission induction and postremission therapy, as seen in a trial from the Southwest Oncology Group (SWOG) and the Eastern Cooperative Oncology Group (ECOG) (E-3489).  Cytogenetic abnormalities that indicate a good prognosis include t(8; 21), inv(16) or t(16;16), and t(15;17). Normal cytogenetics portends average-risk AML. Patients with AML that is characterized by deletions of the long arms or monosomies of chromosomes 5 or 7; by translocations or inversions of chromosome 3, t(6; 9), t(9; 22); or by abnormalities of chromosome 11q23 have particularly poor prognoses with chemotherapy. These cytogenetic subgroups, as seen in the trial from the Medical Research Council (MRC-LEUK-AML11), predict clinical outcome in older patients with AML as well as in younger patients.  The fusion genes formed in t(8; 21) and inv(16) can be detected by reverse transcriptase–polymerase chain reaction (RT–PCR) or fluorescence in situ hybridization (FISH), which will indicate the presence of these genetic alterations in some patients in whom standard cytogenetics was technically inadequate. RT–PCR does not appear to identify significant numbers of patients with good-risk fusion genes who have normal cytogenetics. 
The classification of AML has been revised by a group of pathologists and clinicians under the auspices of the World Health Organization (WHO).  While elements of the French-American-British classification have been retained (i.e., morphology, immunophenotype, cytogenetics and clinical features), the WHO classification incorporates more recent discoveries regarding the genetics and clinical features of AML in an attempt to define entities that are biologically homogeneous and that have prognostic and therapeutic relevance.    Each criterion has prognostic and treatment implications but, for practical purposes, antileukemic therapy is similar for all subtypes.
A long-term follow-up of 30 patients who had AML that was in remission for at least 10 years has demonstrated a 13% incidence of secondary malignancies. Of 31 younger-than-40-years, long-term, female survivors of AML or acute lymphoblastic leukemia, 26 resumed normal menstruation following completion of therapy. Among 36 live offspring of survivors, 2 congenital problems occurred. 
The differentiation of AML from acute lymphocytic leukemia has important therapeutic implications. Histochemical stains and cell surface antigen determinations aid in discrimination.
Other PDQ summaries containing information related to acute myeloid leukemia include the following:
The World Health Organization (WHO) classification of acute myeloid leukemia (AML) incorporates and interrelates morphology, cytogenetics, molecular genetics, and immunologic markers in an attempt to construct a classification that is universally applicable and prognostically valid.  In the older French-American-British (FAB) criteria, the classification of AML is solely based upon morphology as determined by the degree of differentiation along different cell lines and the extent of cell maturation.  
Under the WHO classification, the category “acute myeloid leukemia not otherwise categorized” is morphology-based and reflects the FAB classification with a few significant modifications.   The most significant difference between the WHO and FAB classifications is the WHO recommendation that the requisite blast percentage for the diagnosis of AML be at least 20% blasts in the blood or bone marrow. The FAB scheme required the blast percentage in the blood or bone marrow to be at least 30%. This threshold value for blast percentage eliminated the category “refractory anemia with excess blasts in transformation” (RAEB-t) found in the FAB classification of myelodysplastic syndromes (MDS), where RAEB-t is defined by a marrow blast percentage between 20% and 29%. In the WHO classification, RAEB-t is no longer considered a distinct clinical entity and is instead included within the broader category “AML with multilineage dysplasia” as “AML with multilineage dysplasia following a myelodysplastic syndrome.” 
Although this lowering of the blast threshold has been met with some criticism, several studies indicate that survival patterns for cases with 20% to 29% blasts are similar to survival patterns for cases with 30% or more blasts in the bone marrow.      The diagnosis of AML in itself does not represent a therapeutic mandate. The decision to treat should be based on other factors including patient age, previous history of MDS, clinical findings, disease progression, in addition to the blast percentage, and most importantly, patient preference.
Several groups have begun to investigate the use of gene expression profiling (GEP) using microarrays to augment current diagnostic and prognostic studies for AML. Distinct subsets can be identified using GEP that correspond to known cytogenetic and molecular abnormalities. The positive predictive value appears to be sufficiently powerful to be clinically useful only for patients with the t(8;21) and inv(16) (now referred to as core-binding factor [CBF] leukemias) and acute promyelocytic leukemia (APL) with the t(15;17). GEP identified several cases of CBF leukemias that were not diagnosed using conventional cytogenetics.   
Next-generation sequencing of AML genomes has identified an average of 13 mutations per case. Mutated genes include transcription-factor fusions, nucleophosmin-1, tumor-suppressor, DNA-methylation-related, signaling, chromatin-modifying, myeloid transcription-factor, cohesion-complex, and spliceosome-complex. 
In the following outline and discussion, the older FAB classifications are noted where appropriate.
This category is characterized by characteristic genetic abnormalities and frequently high rates of remission and favorable prognoses with the notable exception of those with 11q23 abnormalities.  The reciprocal translocations t(8; 21), inv(16) or t(16;16), t(15; 17), and translocations involving the 11q23 breakpoint are the most commonly identified genetic abnormalities. These structural chromosome rearrangements result in the formation of fusion genes that encode chimeric proteins that may contribute to the initiation or progression of leukemogenesis. Many of these translocations are detected by reverse transcriptase–polymerase chain reaction (RT–PCR) or fluorescence in situ hybridization (FISH), which has a higher sensitivity than cytogenetics. Other recurring cytogenetic abnormalities are less common and described below in AML not otherwise categorized.
AML with the translocation t(8; 21)(q22; q22) (occurring most commonly in FAB classification M2) is one of the most common genetic aberrations in AML and accounts for 5% to 12% of cases of AML and 33% of karyotypically abnormal cases of acute myeloblastic leukemia with maturation.  Myeloid sarcomas (chloromas) may be present and may be associated with a bone marrow blast percentage of less than 20%.
Common morphologic features include the following:
AML with maturation (FAB classification M2) is the most common morphologic type correlating with t(8; 21). Rarely, AML with this translocation presents with a bone marrow blast percentage less than 20%. 
The translocation t(8; 21)(q22; q22) involves the AML1 gene, also known as RUNX1, which encodes CBF-alpha, and the ETO (eight-twenty-one) gene.   The AML1/ETO fusion transcript is consistently detected in patients with t(8; 21) AML. This type of AML is usually associated with a good response to chemotherapy and a high complete remission (CR) rate with long-term survival when treated with high-dose cytarabine in the postremission phase as in the Cancer and Leukemia Group B (CLB-9022 and CLB-8525) trials.     Additional chromosome abnormalities are common, for example, loss of a sex chromosome and del(9)(q22). Expression of the neural-cell adhesion molecule (CD56) appears to be an adverse prognostic indicator.  
AML with inv(16)(p13; q22) or t(16; 16)(p13; q22) is found in approximately 10% to 12% of all cases of AML, predominantly in younger patients.   Morphologically, this type of AML is associated with acute myelomonocytic leukemia (FAB classification M4) with abnormal eosinophils (AMML Eo). Myeloid sarcomas may be present at initial diagnosis or at relapse.
Common morphologic features include the following:
Most cases with this genetic abnormality have been identified as AMML Eo, but occasional cases have been reported to lack eosinophilia. As is found in rare cases of AML with t(8; 21), the bone marrow blast percentage in this AML is occasionally less than 20%.
Both inv(16)(p13; q22) and t(16; 16)(p13; q22) result in the fusion of the CBF-beta (CBFβ) gene at 16q22 to the smooth muscle myosin heavy chain (MYH11) gene at 16p13, thereby forming the fusion gene CBFβ/MYH11.  The use of FISH and RT–PCR methods may be necessary to document this fusion gene because its presence cannot be reliably documented by traditional cytogenetics banding techniques.  Patients with this type of AML may achieve higher CR rates when treated with high-dose cytarabine in the postremission phase.   
APL AML with t(15; 17)(q22; q12) is an AML in which promyelocytes predominate. APL exists as two types, hypergranular or typical APL and microgranular (hypogranular) APL. APL comprises 5% to 8% of cases of AML and occurs predominately in adults in midlife.  Both typical and microgranular APL are commonly associated with disseminated intravascular coagulation (DIC).   In microgranular APL, unlike typical APL, the leukocyte count is very high with a rapid doubling time. 
Common morphologic features of typical APL include the following:
Common morphologic features of microgranular APL include the following:
In APL, the RARA gene on 17q12 fuses with a nuclear regulatory factor on 15q22 (promyelocytic leukemia or PML gene) resulting in a PML/RARA gene fusion transcript.    Rare cases of cryptic or masked t(15;17) lack typical cytogenetic findings and involve complex variant translocations or submicroscopic insertion of the RARA gene into PML gene leading to the expression of the PML/RARA fusion transcript.  FISH and/or RT–PCR methods may be required to unmask these cryptic genetic rearrangements.  
APL has a specific sensitivity to treatment with all-trans retinoic acid (ATRA, tretinoin), which acts as a differentiating agent.    High CR rates in APL may be obtained by combining ATRA treatment with chemotherapy.  In approximately 1% of the cases of APL, variant chromosomal aberrations may be found in which the RARA gene is fused with other genes.  Variant translocations involving the RARA gene include: t(11;17)(q23; q21), t(5;17)(q32; q12), and t(11; 17)(q13; q21). 
AML with 11q23 abnormalities comprises 5% to 6% of cases of AML and is typically associated with monocytic features. This AML is more common in children. Two clinical subgroups of patients have a high frequency of AML with 11q23 abnormalities: AML in infants and therapy-related AML, usually occurring after treatment with DNA topoisomerase inhibitors. Patients may present with DIC and extramedullary monocytic sarcomas and/or tissue infiltration (gingiva, skin). 
Common morphologic features of this AML include the following:
11q23 abnormalities are associated frequently with acute myelomonocytic, monoblastic, and monocytic leukemias (FAB classifications M4, M5a and M5b, respectively) and occasionally with AML with and without maturation (FAB classifications M2 and M1, respectively). 
The MLL gene on 11q23, a developmental regulator, is involved in translocations with approximately 22 different partner chromosomes.   Genes other than MLL may be involved in 11q23 abnormalities.  FISH may be required to detect genetic abnormalities involving MLL.    In general, risk categories and prognoses for individual 11q23 translocations are difficult to determine because of the lack of studies involving significant numbers of patients; however, patients with t(11; 19)(q23; p13.1) are reported to have poor outcomes. 
Activating mutations of FLT3 (FMS-like tyrosine kinase-3), present at diagnosis in 20% to 30% of de novo AML, represent the most frequent molecular abnormality in this disease.   The most common type of mutation (23%) is an internal tandem duplication mutation (FLT3/internal tandem duplication [ITD]) localized to the juxtamembrane region of the receptor, while point mutations in the kinase domain are less common (7%). Common clinical features of patients with FLT3/ITD AML are:
Patients with FLT3/ITD mutations, and possibly those with FLT3 point mutations, are consistently reported to have an increased relapse rate and reduced overall survival (OS).   The CR rate for patients with FLT3-mutant AML is generally reported to be no different than that for patients with AML with nonmutant FLT3, but most studies examining this clinical parameter used results from patients treated with intensive chemotherapy regimens, and some data are available to suggest that the conventional 7+3 regimen leads to a reduced remission rate in this group of patients.  [Level of evidence: 3iiiDiv]
One study from the German-Austrian Acute Myeloid Leukemia Study Group examined data on 872 patients with cytogenetically normal AML treated with intensive induction and postremission regimens over an 11-year period.  [Level of evidence: 3iiiA] The study group found that patients with a mutant cytosine-cytosine-adenosine-adenosine-thymidine (CCAAT)-enhancer binding-protein alpha (CEBPA) or NPM1 without fms-related tyrosine kinase 3-internal tandem duplication (FLT3-ITD) had higher complete response rates, disease-free survival (DFS) rates, and OS rates (with a 4-year OS rate of 62% and 60%, respectively) than other cytogenetically normal AML patients (who had a 4-year OS rate of between 25% and 30%). As yet, no clear strategy exists for improving patient outcome in FLT3-mutant AML, or in patients with abnormalities other than CEBPA or the NPM1 without the FLT3-ITD, but small molecule FLT3 inhibitors are in development, and the role of allogeneic transplant is being considered.
In the WHO classification, refractory anemia with excess blasts in transformation (RAEB-t) is no longer considered a distinct clinical entity and is instead included within the broader category “AML with multilineage dysplasia” as one of the following:
AML with multilineage dysplasia is characterized by 20% or more blasts in the blood or bone marrow and dysplasia in two or more myeloid cell lines, generally including megakaryocytes.  To make the diagnosis, dysplasia must be present in 50% or more of the cells of at least two lineages and must be present in a pretreatment bone marrow specimen.   AML with multilineage dysplasia may occur de novo or following MDS or a myelodysplastic and myeloproliferative disorder (MDS and MPD). (Refer to the PDQ summaries on Myelodysplastic Syndromes Treatment and Myelodysplastic/ Myeloproliferative Neoplasms for more information.) The diagnostic terminology “AML with multilineage dysplasia evolving from a myelodysplastic syndrome” should be used when an MDS precedes AML. 
Common morphologic features include the following:
The differential diagnosis of AML with multilineage dysplasia includes acute erythroid-myeloid leukemia and acute myeloblastic leukemia with maturation (FAB classifications M6a and M2). Some cases may overlap two morphologic types. 
As evidenced in several Southwest Oncology Group studies, such as SWOG-8600 and NCT00023777, the numerous chromosome abnormalities observed in AML with multilineage dysplasia were similar to those found in MDS and frequently involved gain or loss of major segments of certain chromosomes, predominately chromosomes 5 and/or 7.     The probability of achieving a CR has been reported to be affected adversely by a diagnosis of AML with multilineage dysplasia.   
This category includes AML and MDS that arise secondary to cytotoxic chemotherapy and/or radiation therapy.  The therapy-related (or secondary) MDS are included because of their close clinicopathologic relationships to therapy-related AML. Although these therapy-related disorders are distinguished by the specific mutagenic agents involved, a recent study suggests this distinction may be difficult to make because of the frequent overlapping use of multiple potentially mutagenic agents in treating cancer. 
The alkylating agent/radiation-related acute leukemias and myelodysplastic syndromes typically occur 5 to 6 years following exposure to the mutagenic agent, with a reported range of approximately 10 to 192 months.   The risk for occurrence is related to both the total cumulative dose of the alkylating agent and the age of the patient. Clinically, the disorder commonly presents initially as an MDS with evidence of bone marrow failure. This stage is followed by dysplastic features in multiple cell lineages with a blast percentage that is usually less than 5%. In the MDS phase, approximately 66% of cases satisfy the criteria for refractory cytopenia with multilineage dysplasia (RCMD), with approximately 33% of these cases exhibiting ringed sideroblasts in excess of 15% (RCMD-RS).  (Refer to the PDQ summary on Myelodysplastic Syndromes Treatment for more information.) Another 25% of cases satisfy the criteria for refractory anemia with excess blasts 1 or 2 (RAEB-1; RAEB-2). The MDS phase may evolve to a higher grade MDS or AML. Although a minority of patients may present with acute leukemia, a substantial number of patients succumb to the disorder in the MDS phase. 
Common morphologic features include the following:
Cases may correspond morphologically to AML with maturation, acute monocytic leukemia, AMML, erythroleukemia, or acute megakaryoblastic leukemia (FAB classifications M2, M5b, M4, M6a, and M7, respectively).
Cytogenetic abnormalities have been observed in more than 90% of cases of therapy-related AML or MDS and commonly include chromosomes 5 and/or 7.    Complex chromosomal abnormalities (≥3 distinct abnormalities) are the most common finding.     Therapy-related AML is usually refractory to antileukemia therapy. Median survival after diagnosis of these disorders is approximately 7 to 8 months.  
This type of AML occurs in patients treated with topoisomerase II inhibitors. The agents implicated are the epipodophyllotoxins etoposide and teniposide and the anthracyclines doxorubicin and 4-epi-doxorubicin.  The mean latency period from the time of institution of the causative therapy to the development of AML is approximately 2 years.  Morphologically, there is a significant monocytic component. Most cases are categorized as acute monoblastic or myelomonocytic leukemia. Other morphologies reported include APL, myelodysplastic syndromes, and acute megakaryoblastic leukemia. 
As with alkylating agent/radiation-related acute leukemias and myelodysplastic syndromes, the cytogenetic abnormalities are often complex.     The predominant cytogenetic finding involves chromosome 11q23 and the MLL gene.   Current data are insufficient to predict survival times.
Cases of AML that do not fulfill the criteria for AML with recurrent genetic abnormalities, AML with multilineage dysplasia, or AML and MDS, therapy-related, fall within this category. Classification within this category is based on leukemic cell features of morphology, cytochemistry, and maturation. 
This AML shows no evidence of myeloid differentiation by morphology and light microscopy cytochemistry.  The myeloid nature of the blasts is demonstrated by immunophenotyping and/or ultrastructural studies.  Immunophenotyping studies must be performed to distinguish this acute leukemia from acute lymphoblastic leukemia (ALL).  AML, minimally differentiated, comprise approximately 5% of cases of AML. Patients with this AML typically present with evidence of marrow failure, thrombocytopenia, and neutropenia. 
Morphologic and cytochemical features include the following:
Immunophenotyping reveals blast cells that express one or more panmyeloid antigens (CD13, CD33, and CD117) and are negative for B and T lymphoid-restricted antigens. Most cases express primitive hematopoietic-associated antigens (CD34, CD38, and HLA-DR). The differential diagnosis includes ALL, acute megakaryoblastic leukemia, biphenotypic/mixed lineage acute leukemia, and, rarely, the leukemic phase of large cell lymphoma. Immunophenotyping studies are required to distinguish these disorders. 
Although no specific chromosomal abnormalities have been found in AML, minimally differentiated point mutations of the AML1 gene have been observed in approximately 25% of cases. This mutation appears to correlate clinically with a higher white blood cell count and greater marrow blast involvement.   Mutation of FLT3, a receptor tyrosine kinase gene, occurs in approximately 25% of cases and has been associated with short survival.   The median OS is approximately 10 months. 
AML without maturation is characterized by a high percentage of bone marrow blasts with little evidence of maturation to mature neutrophils and comprises approximately 10% of cases of AML.  Most patients are adults. Patients usually present with anemia, thrombocytopenia, and neutropenia. (Refer to the PDQ summary on Fatigue for more information on anemia.)
Common morphologic and cytochemical features include the following:
Immunophenotyping reveals blasts that express at least two myelomonocytic antigens (CD13, CD33, CD117) and/or MPO. CD34 is often positive. The differential diagnosis includes ALL in cases of AML without maturation with no granules and a low percentage of MPO positive blasts, and AML with maturation in cases of AML with maturation with a high percentage of blasts.
Although no specific chromosomal abnormality has been identified for AML without maturation, mutation of the FLT3 gene has been associated with leukocytosis, a high percentage of bone marrow blast cells, and a worse prognosis.   
AML with maturation is characterized by 20% or more myeloblasts in the blood or bone marrow and 10% or more neutrophils at different stages of maturation. Monocytes constitute less than 20% of bone marrow cells.  This AML comprises approximately 30% to 45% of cases of AML. While it occurs in all age groups, 20% of patients are younger than 25 years and 40% of patients are aged 60 years or older.  Patients frequently present with anemia, thrombocytopenia, and neutropenia. (Refer to the PDQ summary on Fatigue for more information on anemia.)
Morphologic features include the following:
With immunophenotyping, the blasts typically express one or more myeloid-associated antigens (CD13, CD33, and CD15). The differential diagnosis includes: RAEB in cases with a low blast percentage, AML without maturation when the blast percentage is high, and AMML in cases with increased monocytes.
Approximately 33% of karyotypically abnormal cases of AML with maturation are associated with t(8; 21)(q22;q22). (Refer to the Acute myeloid leukemia with characteristic genetic abnormalities section of the Classification section of this summary for more information.)  Such cases have a favorable prognosis. Rare cases with t(6; 9)(q23; q34) are reported to have a poor prognosis.  
(Refer to the Acute promyelocytic leukemia (FAB Classification M3) section of the Acute Myeloid Leukemia With Characteristic Genetic Abnormalities section of this summary for more information.)
Acute myelomonocytic leukemia (AMML) is characterized by the proliferation of neutrophil and monocyte precursors. Patients usually present with anemia and thrombocytopenia. (Refer to the PDQ summary on Fatigue for more information on anemia.) This classification of AML comprises approximately 15% to 25% of cases of AML, and some patients have a previous history of chronic myelomonocytic leukemia (CMML). (Refer to the PDQ summary on Myelodysplastic/ Myeloproliferative Neoplasms for more information.) This type of AML occurs more commonly in older individuals. 
Morphologic and cytochemical features include the following:
Immunophenotyping generally reveals monocytic differentiation markers (CD14, CD4, CD11b, CD11c, CD64, and CD36) and lysozyme. The differential diagnosis includes AML with maturation and acute monocytic leukemia.
Most cases of AMML exhibit nonspecific cytogenetic abnormalities.  Some cases may have a 11q23 genetic abnormality. Cases with increased abnormal eosinophils in the bone marrow associated with a chromosome 16 abnormality have a favorable prognosis. (Refer to the Acute myeloid leukemia with characteristic genetic abnormalities section of the Classification section of this summary for more information.)
Acute monoblastic and acute monocytic leukemia are AMLs in which 80% or more of the leukemic cells are of a monocytic lineage. These cells include monoblasts, promonocytes, and monocytes. These two leukemias are distinguished by the relative proportions of monoblasts and promonocytes. In acute monoblastic leukemia, most monocytic cells are monoblasts (usually ≥80%). In acute monocytic leukemia, most of the monocytic cells are promonocytes.  Acute monoblastic leukemia comprises 5% to 8% of cases of AML and occurs most commonly in young individuals. Acute monocytic leukemia comprises 3% to 6% of cases and is more common in adults.  Common clinical features for both acute leukemias include bleeding disorders, extramedullary masses, cutaneous and gingival infiltration, and central nervous system involvement.
Morphologic and cytochemical features of acute monoblastic leukemia include the following:
Morphologic and cytochemical features of acute monocytic leukemia include the following:
The extramedullary lesions of these leukemias may be predominantly monoblastic or monocytic or an admixture of the two cell types. Immunophenotyping of these leukemias may reveal expression of the myeloid antigens CD13, CD33, CD117, CD14 ( + ), CD4, CD36, CD 11b, CD11c, CD64, and CD68.  The differential diagnosis of acute monoblastic leukemia includes AML without maturation, minimally differentiated AML, and acute megakaryoblastic leukemia. The differential diagnosis of acute monocytic leukemia includes AMML and microgranular APL.
An abnormal karyotype has been observed in approximately 75% of cases of acute monoblastic leukemia while approximately 30% of cases of acute monocytic leukemia are associated with an abnormal karyotype. Almost 30% of cases of acute monoblastic leukemia and 12% of cases of acute monocytic leukemia are associated with 11q23 genetic abnormalities involving the MLL gene. (Refer to the Acute myeloid leukemia with characteristic genetic abnormalities section of the Classification section of this summary for more information.) Mutation of FLT3, a receptor tyrosine kinase gene, has been observed in about 30% of cases of acute monocytic leukemia (approximately 7% in acute monoblastic leukemia).  The translocation t(8;16)(p11; p13) (strongly associated with acute monocytic leukemia, hemophagocytosis by leukemic cells, and a poor response to chemotherapy) fuses the MOZ gene (8p11) with the CBP gene (16p13).  Median actuarial DFS for acute monocytic leukemia has been reported to be approximately 21 months. 
The two subtypes of the acute erythroid leukemias, erythroleukemia and pure erythroid leukemia, are characterized by a predominant erythroid population and, in the case of erythroleukemia, the presence of a significant myeloid component. Erythroleukemia (erythroid/myeloid; M6a) is predominantly a disease of adults, comprising approximately 5% to 6% of cases of AML.  Pure erythroid leukemia (M6b) is rare and occurs in all age groups. Occasional cases of chronic myeloid leukemia (CML) may evolve to one of the acute erythroid leukemias.  Erythroleukemia may present de novo or evolve from an MDS, either RAEB or RCMD-RS or RCMD. (Refer to the PDQ summary on Myelodysplastic Syndromes Treatment for more information.) The clinical features of these acute leukemias include profound anemia and normoblastemia. (Refer to the PDQ summary on Fatigue for more information.)
Morphologic and cytochemical features of erythroleukemia include the following: 
Morphologic and cytochemical features of pure erythroid leukemia include the following:
Immunophenotyping in erythroleukemia reveals erythroblasts that react with antibodies to glycophorin A and hemoglobin A and myeloblasts that express a variety of myeloid-associated antigens (CD13, CD33, CD117, c-kit, and MPO). Immunophenotyping in acute erythroid leukemia reveals expression of glycophorin A and hemoglobin A in differentiated forms. Markers such as carbonic anhydrase 1, Gero antibody against the Gerbich blood group, or CD36 are usually positive. The differential diagnosis for erythroleukemia includes RAEB and AML with maturation with increased erythroid precursors and AML with multilineage dysplasia (involving ≥50% of myeloid or megakaryocyte-lineage cells). If erythroid precursors are 50% or more and the nonerythroid component is 20% or more, the diagnosis is erythroleukemia, whereas, if the nonerythroid component is less than 20%, the diagnosis is RAEB. The differential diagnosis for pure erythroid leukemia includes megaloblastic anemia secondary to vitamin B12 or folate deficiency, acute megakaryocytic leukemia, and ALL or lymphoma. 
No specific chromosome abnormalities are described for these AMLs. Complex karyotypes with multiple structural abnormalities are common. Chromosomes 5 and 7 appear to be affected frequently.    One study indicates that abnormalities of chromosomes 5 and/or 7 correlate with significantly shorter survival times. 
Acute megakaryoblastic leukemia, in which 50% or more of blasts are of the megakaryocyte lineage, occurs in all age groups and comprises approximately 3% to 5% of cases of AML.  Clinical features include cytopenias; dysplastic changes in neutrophils and platelets; rare organomegaly, except in children with t(1; 22); lytic bone lesions in children; and association with mediastinal germ cell tumors in young adult males.   
Immunophenotyping reveals megakaryoblast expression of one or more platelet glycoproteins: CD41 (glycoprotein IIb/IIIa) and/or CD61 (glycoprotein IIIa). Myeloid markers CD13 and CD33 may be positive; CD36 is typically positive. Blasts are negative with the anti-MPO antibody and other markers of myeloid differentiation. In bone marrow biopsies, megakaryocytes and megakaryoblasts may react positively to antibodies for Factor VIII.  The differential diagnosis includes minimally differentiated AML, acute panmyelosis with myelofibrosis, ALL, pure erythroid leukemia, and blastic transformation of chronic myeloid leukemia or idiopathic myelofibrosis and metastatic tumors in the bone marrow (particularly in children). (Refer to the PDQ summary on Chronic Myeloproliferative Neoplasms Treatment for more information on chronic myeloid leukemia or idiopathic myelofibrosis).
No unique chromosomal abnormalities are associated with acute megakaryoblastic leukemia in adults.   In children, particularly infants, a distinct clinical presentation may be associated with t(1:22)(p13; q13).   The prognosis for this type of acute leukemia is poor.  
Individuals with Down syndrome (trisomy 21) have an increased disposition to acute leukemia, primarily the myeloid type.   The primary subtype appears to be acute megakaryoblastic leukemia. In cases in which the leukemia remits spontaneously, the process is referred to as transient myeloproliferative disorder or transient leukemia. Clinical features include presentation in the neonatal period (10% of newborn infants with Down syndrome), marked leukocytosis, blast percentage in the blood greater than 30% to 50%, and extramedullary involvement.
Morphologic and cytochemical features include the following:
Immunophenotyping reveals markers that are generally similar to those of other cases of childhood acute megakaryoblastic leukemia.
In addition to trisomy 21, some cases may show other clonal abnormalities, particularly trisomy 8.   Spontaneous remission occurs within 1 to 3 months in transient cases. Recurrence followed by a second spontaneous remission or persistent disease may occur. Treatment outcomes for pediatric patients with Down syndrome and persistent disease may be better than those for pediatric patients with acute leukemia in the absence of trisomy 21. 
Acute basophilic leukemia is an AML that exhibits a primary differentiation to basophils. This acute leukemia is relatively rare, comprising less than 1% of all cases of AML.  Clinical features include bone marrow failure, circulating blasts, cutaneous involvement, organomegaly, occasional osseous lytic lesions, and symptoms secondary to hyperhistaminemia.
Morphologic and cytochemical features include the following:
Immunophenotypically, the blasts express the myeloid markers CD13 and CD33 and the early hematopoietic markers CD34 and class-II HLA-DR. The differential diagnosis includes: blast crisis of CML, other AML subtypes with basophilia such as AML with maturation (M2) associated with abnormalities of 12p or t(6;9), acute eosinophilic leukemia, and, rarely, a subtype of ALL with prominent coarse granules. 
No consistent chromosome abnormality has been identified for acute basophilic leukemia.  Because of its rare incidence, little information regarding survival is available.
Acute panmyelosis with myelofibrosis (also known as acute myelofibrosis, acute myelosclerosis, and acute myelodysplasia with myelofibrosis) is an acute panmyeloid proliferation associated with fibrosis of the bone marrow. This disorder is very rare and occurs in all age groups.  The disorder may occur de novo or after treatment with alkylating-agent chemotherapy and/or radiation. (Refer to the Acute myeloid leukemias and myelodysplastic syndromes, therapy related section of this summary for more information). Clinical features include constitutional symptoms such as weakness and fatigue. (Refer to the PDQ summary on Fatigue for more information.)
Morphologic and cytochemical features include the following:
Immunophenotypically, blasts may express one or more myeloid-associated antigens (CD13, CD33, CD117, and MPO). Some cells may express erythroid or megakaryocytic antigens. The major differential diagnosis includes acute megakaryoblastic leukemia, acute leukemias with associated marrow fibrosis, metastatic tumor with a desmoplastic reaction, and chronic idiopathic myelofibrosis.  (Refer to the PDQ summary on Chronic Myeloproliferative Neoplasms Treatment for more information.)
No specific chromosomal abnormalities are associated with acute panmyelosis with myelofibrosis. This AML is reported to respond poorly to chemotherapy and to be associated with a short survival. 
Myeloid sarcoma (also known as extramedullary myeloid tumor, granulocytic sarcoma, and chloroma) is a tumor mass that consists of myeloblasts or immature myeloid cells, occurring in an extramedullary site;  development in 2% to 8% of patients with AML has been reported.  Clinical features include occurrence common in subperiosteal bone structures of the skull, paranasal sinuses, sternum, ribs, vertebrae, and pelvis; lymph nodes, skin, mediastinum, small intestine, and the epidural space; and occurrence de novo or concomitant with AML or a myeloproliferative disorder.  
Morphologic and cytochemical features include the following:
Immunophenotyping with antibodies to MPO, lysozyme, and chloroacetate are critical to the diagnosis of these lesions.  The myeloblasts in granulocytic sarcomas express myeloid-associated antigens (CD13, CD33, CD117, and MPO). The monoblasts in monoblastic sarcomas express acute monoblastic leukemia antigens (CD14, CD116, and CD11c) and usually react with antibodies to lysozyme and CD68. The main differential diagnosis includes non-Hodgkin lymphoma of the lymphoblastic type, Burkitt lymphoma, large-cell lymphoma, and small, round-cell tumors, especially in children (e.g., neuroblastoma, rhabdomyosarcoma, Ewing/primitive neuroectodermal tumors, and medulloblastoma).
No unique chromosomal abnormalities are associated with myeloid sarcoma.   AML with maturation and t(8; 21)(q22; q22) and AMML Eo with-in (16)(p13; q22) or t(16;16)(p13; q22) may be observed and monoblastic sarcoma may be associated with translocations involving 11q23.  The presence of myeloid sarcoma in patients with the otherwise good-risk t(8; 21) AML may be associated with a lower CR rate and decreased remission duration.  Myeloid sarcoma occurring in the setting of MDS or MPD is equivalent to blast transformation. In the case of AML, the prognosis is that of the underlying leukemia.  Although the initial presentation of myeloid sarcoma may appear to be isolated, several reports indicate that isolated myeloid sarcoma is a partial manifestation of a systemic disease and should be treated with intensive chemotherapy.   
Acute leukemias of ambiguous lineage (also known as acute leukemias of undetermined lineage, mixed phenotype acute leukemias, mixed lineage acute leukemias, and hybrid acute leukemias) are types of acute leukemia in which the morphologic, cytochemical, and immunophenotypic features of the blast population do not allow classification in myeloid or lymphoid categories; or the types have morphologic and/or immunophenotypic features of both myeloid and lymphoid cells or both B and T lineages (i.e., acute bilineal leukemia and acute biphenotypic leukemia).      These rare leukemias account for less than 4% of all cases of acute leukemia and occur in all age groups but are more frequent in adults.  Clinical features include symptoms and complications caused by cytopenias, i.e., fatigue, infections, and bleeding disorders. (Refer to the PDQ summary on Fatigue for more information.)
The differential diagnosis includes myeloid antigen-positive ALL or lymphoid-positive AML (from which biphenotypic acute leukemia should be distinguished) and minimally differentiated AML (from which undifferentiated acute leukemia must be distinguished).
Cytogenetic abnormalities are observed in a high percentage of bilineal and biphenotypic leukemias.     Approximately 33% of cases have the Philadelphia chromosome, and some cases are associated with t(4; 11)(q21; q23) or other 11q23 abnormalities. In general, the prognosis appears to be unfavorable, particularly in adults; the occurrence of the translocation t(4; 11) or the Philadelphia chromosome are especially unfavorable prognostic indicators.   
There is no clear-cut staging system for this disease.
Untreated adult acute myeloid leukemia (AML) is defined as newly diagnosed leukemia with no previous treatment. The patient exhibits the following features: abnormal bone marrow with at least 20% blasts and signs and symptoms of the disease, usually accompanied by an abnormal white blood cell count and differential, an abnormal hematocrit/hemoglobin count, and an abnormal platelet count.
AML in remission is defined as a normal peripheral blood cell count (absolute neutrophil count >1,000/mm3 and platelet count >100,000/mm3)  and normocellular marrow with less than 5% blasts in the marrow and no signs or symptoms of the disease. In addition, no signs or symptoms are evident of central nervous system leukemia or other extramedullary infiltration. Because the vast majority of AML patients meeting these criteria for remission have residual leukemia, modifications to the definition of complete remission have been suggested, including cytogenetic remission, in which a previously abnormal karyotype reverts to normal, and molecular remission, in which interphase fluorescence in situ hybridization (FISH) or multiparameter flow cytometry are used to detect minimal residual disease. Immunophenotyping and interphase FISH have greater prognostic significance than the conventional criteria for remission.  
Check the list of NCI-supported cancer clinical trials that are now accepting patients with adult acute myeloid leukemia. The list of clinical trials can be further narrowed by location, drug, intervention, and other criteria.
General information about clinical trials is also available from the NCI website.
Successful treatment of acute myeloid leukemia (AML) requires the control of bone marrow and systemic disease and specific treatment of central nervous system (CNS) disease, if present. The cornerstone of this strategy includes systemically administered combination chemotherapy. Because only 5% of patients with AML develop CNS disease, prophylactic treatment is not indicated.   
Treatment is divided into two phases: remission induction (to attain remission) and postremission (to maintain remission). Maintenance therapy for AML was previously administered for several years but is not included in most current treatment clinical trials in the United States, other than for acute promyelocytic leukemia. (Refer to the Adult Acute Myeloid Leukemia in Remission section of this summary for more information.) Other studies have used more intensive postremission therapy administered for a shorter duration of time after which treatment is discontinued.  Postremission therapy appears to be effective when given immediately after remission is achieved. 
Since myelosuppression is an anticipated consequence of both the leukemia and its treatment with chemotherapy, patients must be closely monitored during therapy. Facilities must be available for hematologic support with multiple blood fractions including platelet transfusions and for the treatment of related infectious complications.  Randomized trials have shown similar outcomes for patients who received prophylactic platelet transfusions at a level of 10,000/mm3 rather than 20,000/mm3.  The incidence of platelet alloimmunization was similar among groups randomly assigned to receive pooled platelet concentrates from random donors; filtered, pooled platelet concentrates from random donors; ultraviolet B-irradiated, pooled platelet concentrates from random donors; or filtered platelets obtained by apheresis from single random donors.  Colony-stimulating factors, for example, granulocyte colony–stimulating factor (G-CSF) and granulocyte-macrophage colony–stimulating factor (GM-CSF), have been studied in an effort to shorten the period of granulocytopenia associated with leukemia treatment.  If used, these agents are administered after completion of induction therapy. GM-CSF was shown to improve survival in a randomized trial of AML in patients aged 55 to 70 years (median survival was 10.6 months vs. 4.8 months). In this Eastern Cooperative Oncology Group (ECOG) (EST-1490) trial, patients were randomly assigned to receive GM-CSF or placebo following demonstration of leukemic clearance of the bone marrow;  however, GM-CSF did not show benefit in a separate similar randomized trial in patients older than 60 years.  In the latter study, clearance of the marrow was not required before initiating cytokine therapy. In a Southwest Oncology Group (NCT00023777) randomized trial of G-CSF given following induction therapy to patients older than 65 years, complete response was higher in patients who received G-CSF because of a decreased incidence of primary leukemic resistance. Growth factor administration did not impact on mortality or on survival.   Because the majority of randomized clinical trials have not shown an impact of growth factors on survival, their use is not routinely recommended in the remission induction setting.
The administration of GM-CSF or other myeloid growth factors before and during induction therapy, to augment the effects of cytotoxic therapy through the recruitment of leukemic blasts into cell cycle (growth factor priming), has been an area of active clinical research. Evidence from randomized studies of GM-CSF priming have come to opposite conclusions. A randomized study of GM-CSF priming during conventional induction and postremission therapy showed no difference in outcomes between patients who received GM-CSF and those who did not receive growth factor priming.   [Level of evidence: 1iiA] In contrast, a similar randomized placebo-controlled study of GM-CSF priming in patients with AML aged 55 to 75 years showed improved disease-free survival (DFS) in the group receiving GM-CSF (median DFS for patients who achieved complete remission was 23 months vs. 11 months; 2-year DFS was 48% vs. 21%), with a trend towards improvement in overall survival (2-year survival was 39% vs. 27%, P = .082) for patients aged 55 to 64 years.  [Level of evidence: 1iiDii]
The two-drug regimen of cytarabine given in conjunction with daunorubicin (the so-called 7+3 induction therapy) will result in a complete response rate of approximately 65%. Some physicians opt to add a third drug, thioguanine, to this regimen, although little evidence is available to conclude that this three-drug regimen is better therapy. One study suggested that the addition of etoposide during induction therapy may improve response duration.  The choice of anthracycline and the dose-intensity of anthracycline may influence the survival of patients with acute myeloid leukemia (AML). Idarubicin appeared to be more effective than daunorubicin, particularly in younger adults, although the doses of idarubicin and daunorubicin may not have been equivalent.     No significant survival difference between daunorubicin and mitoxantrone has been reported. 
In patients aged 60 years and younger, outcomes for those receiving daunorubicin (90 mg/m2/dose, total induction dosing at 270 mg/m2) were superior to those receiving more traditional dosing (45 mg/m2/dose; total dose = 135 mg/m2). Complete remission (CR) rate was 71% versus 57% (P < .001), and median survival was 24 months versus 16 months (P = .003).  No randomized comparison data between daunorubicin at 270 mg/m2 and daunorubicin at 180 mg/m2, nor between daunorubicin at 270 mg/m2 and idarubicin, are available. However, two studies examined when idarubicin (36 mg/m2) versus daunorubicin (180 mg/m2 or 240 mg/m2) were administered to elderly patients. While overall survival (OS) was not impacted by the choice of anthracycline, the percentage of long-term disease-free survivors in a mixed-cure model did appear to be impacted (hazard ratio [HR], 0.8; 0.65–0.98).  The addition of the CD33-directed immunotoxin gemtuzumab ozogamicin to cytarabine plus anthracycline or clofarabine plus anthracycline in patients aged 51 to 79 years led to a small increase in median survival (25% vs. 20%; HR, 0.87; 95% confidence interval [CI], 0.76–1.00; P = < .05).  In contrast, gemtuzumab did not improve the 1-year survival rate of elderly patients receiving low-dose cytarabine, although the CR rate increased from 17% to 30% (odds ratio [OR], 0.48 (0.32–0.73); P = .006). 
The role of high-dose cytarabine in induction therapy is controversial; randomized trials have shown prolongation of disease-free survival (DFS)   or no effect   compared with conventionally dosed cytarabine-based induction chemotherapy. Post hoc analyses of two negative trials suggested potential benefit for the intensified therapy in subsets of patients at high risk for treatment failure;   however, an analysis of a subset of patients with complex cytogenetic abnormalities treated in a randomized multicenter trial in Germany showed improvement in CR rate with minimal improvement in event-free survival (EFS) (CR, 56% vs. 23%; P = .04; median EFS, 1 month vs. 2 months; P = .04).  [Level of evidence: 1iiDii]
AML arising from myelodysplasia or secondary to previous cytotoxic chemotherapy has a lower rate of remission than de novo AML. A retrospective analysis of patients undergoing allogeneic bone marrow transplantation (BMT) in this setting showed that the long-term survival for such patients was identical regardless of whether or not patients had received remission induction therapy (DFS was approximately 20%). These data suggest that patients with these subsets of leukemia may be treated primarily with allogeneic BMT if their overall performance status is adequate, potentially sparing patients the added toxic effect of induction chemotherapy.  [Level of evidence: 3iiiDii]
Older adults who decline intensive remission induction therapy or are considered unfit for intensive remission induction therapy may derive benefit from low-dose cytarabine, administered twice daily for 10 days in cycles repeated every 4 to 6 weeks. The CR rate using this regimen was 18% compared with 1% for patients treated with hydroxyurea (P = .006).  Survival with low-dose cytarabine was better than survival was with hydroxyurea (OR, 0.60; 95% CI, 0.44–0.81; P = .009).  [Level of evidence: 1iiA] The hypomethylating agents decitabine and azacitidine are used commonly in this population of older adults, particularly in the United States. Although approval of the drugs by the U.S. Food and Drug Administration is for a myelodysplastic syndrome indication, the registration studies leading to approval included patients with 20% to 30% myeloblasts, or what would now be considered oligoblastic AML.  
One phase III trial randomly assigned 485 AML patients older than 65 years to receive decitabine (n=242) or their preferred choice (n=243) of either supportive care (n=28) or low-dose cytarabine (n=215). Although rates of CR + CRp (CR with incomplete platelet recovery) were more than double in the decitabine arm (17.8%) compared with the treatment-choice arm (7.8%) (P = .001), median OS was not significantly improved for patients receiving decitabine (7.7 months) compared with the treatment of choice (5.0 months) (HR for death for decitabine, 0.85; 95% CI, .69–1.04; P = .11). 
Preliminary results from a phase III trial randomly assigned AML patients older than 65 years to azacitidine compared with conventional care regimens of best supportive care, low-dose cytarabine, and 7+3 AML-type induction chemotherapy and similarly showed a nonsignificant difference in median OS for patients receiving azacitidine (10.4 months) versus conventional care (6.5 months) (HR for death for azacitidine, 0.84; 95% CI, .69–1.02; (P = .08). 
In sum, low-dose cytarabine, decitabine, azacitidine, or best supportive care can be considered equivalently effective treatment approaches for older AML patients who decline traditional, 7+3 induction chemotherapy.
Supportive care during remission induction treatment should routinely include red blood cell and platelet transfusions when appropriate.   Empiric broad spectrum antimicrobial therapy is an absolute necessity for febrile patients who are profoundly neutropenic.   Careful instruction in personal hygiene, dental care, and recognition of early signs of infection are appropriate in all patients. Elaborate isolation facilities (including filtered air, sterile food, and gut flora sterilization) are not routinely indicated but may benefit transplant patients.   Rapid marrow ablation with consequent earlier marrow regeneration decreases morbidity and mortality. Prophylactic oral antibiotics may be appropriate in patients with expected prolonged, profound granulocytopenia (<100/mm3 for 2 weeks).  Norfloxacin and ciprofloxacin have been shown to decrease the incidence of gram-negative infection and time to first fever in randomized trials. The combination of ofloxacin and rifampin has proven superior to norfloxacin in decreasing the incidence of documented granulocytopenic infection.    Serial surveillance cultures may be helpful in such patients to detect the presence or acquisition of resistant organisms.
A long-term follow-up of 30 patients who had AML that was in remission for at least 10 years has demonstrated a 13% incidence of secondary malignancies. Of 31 long-term female survivors of AML or acute lymphoblastic leukemia younger than 40 years, 26 resumed normal menstruation following completion of therapy. Among 36 live offspring of survivors, two congenital problems occurred. 
Special consideration must be given to induction therapy for acute promyelocytic leukemia (APL). Oral administration of tretinoin (all-trans-retinoic acid [ATRA]); 45 mg/mm2/day) can induce remission in 70% to 90% of patients with M3 AML. (ATRA is not effective in patients with AML that resembles M3 morphologically but does not demonstrate the t(15;17) or typical PML-RARA gene rearrangement.)        ATRA induces terminal differentiation of the leukemic cells followed by restoration of nonclonal hematopoiesis. Administration of ATRA leads to rapid resolution of coagulopathy in most patients, and heparin administration is not required in patients receiving ATRA. However, randomized trials have not shown a reduction in morbidity and mortality during ATRA induction when compared with chemotherapy. Administration of ATRA can lead to hyperleukocytosis and a syndrome of respiratory distress now known as the differentiation syndrome. Prompt recognition of the syndrome and aggressive administration of steroids can prevent severe respiratory distress.  The optimal management of ATRA-induced hyperleukocytosis has not been established; neither has the optimal postremission management of patients who receive ATRA induction. However, two large cooperative group trials have demonstrated a statistically significant relapse-free and OS advantage to patients with M3 AML who receive ATRA at some point during their antileukemic management.  
The C9710 (NCT00003934) trial randomly assigned patients receiving ATRA and anthracyclines to two cycles of consolidation with or without arsenic trioxide (ATO). Event-free survival (EFS), the primary endpoint, was significantly better for patients assigned to receive ATO consolidation, with an 80% EFS rate compared with a 63% EFS rate at 3 years (stratified log-rank test, P < .0001). The secondary endpoint, survival, was better in the ATO arm, with an 86% survival rate compared with an 81% survival rate at 3 years (P = .059) The inclusion of ATO led to outcomes for higher-risk patients that were equivalent to the outcomes for lower-risk patients.  A phase II study showed that incorporation of ATO in the primary management of APL patients could reduce the total amount of therapy administered. 
Investigators at the University of Texas MD Anderson Cancer Center used an ATO-based regimen, which included gemtuzumab ozogamicin (GO) as the only cytotoxic drug.  Patients received ATRA plus ATO induction; patients also received a dose of GO if the WBC was greater than 10,000/mm3 on presentation or rose to over 30,000/mm3 during induction. Patients in remission received alternating months of ATO and ATRA for a total of seven cycles; GO was substituted if either ATO or ATRA were discontinued as a result of toxicity. Eighty-two patients were treated; seven patients died during induction, the remainder achieved remission. Three patients relapsed and four patients died during remission; thus EFS was approximately 76%.
This approach was investigated in a randomized, noninferiority trial that compared ATO plus ATRA with an ATRA-anthracycline-based regimen in patients with lower-risk APL. With median follow-up of 34.4 months, 2-year EFS rates were 97% in the ATRA-ATO group and 86% in the ATRA-chemotherapy group (95% CI for the difference, 2%–22%). OS was also better with ATRA-ATO (P = .02). 
Most current regimens for the treatment of APL include some form of maintenance therapy, particularly for patients with higher-risk APL. A meta-analysis of randomized trials has indicated that maintenance clearly improves DFS but not OS; however, these studies did not include ATO-containing trials.
Check the list of NCI-supported cancer clinical trials that are now accepting patients with untreated adult acute myeloid leukemia. The list of clinical trials can be further narrowed by location, drug, intervention, and other criteria.
General information about clinical trials is also available from the NCI website.
Although individual patients have been reported to have long disease-free survival (DFS) or cure with a single cycle of chemotherapy,  postremission therapy is always indicated in therapy that is planned with curative intent. In a small randomized study conducted by the Eastern Cooperative Oncology Group (ECOG), all patients who did not receive postremission therapy experienced a relapse after a short median complete remission (CR) duration.  Current approaches to postremission therapy include short-term, relatively intensive chemotherapy with cytarabine-based regimens similar to standard induction clinical trials (postremission chemotherapy), postremission chemotherapy with more dose-intensive cytarabine-based treatment, high-dose chemotherapy or chemoradiation therapy with autologous bone marrow rescue, and high-dose marrow-ablative therapy with allogeneic bone marrow rescue. While older studies have included longer-term therapy at lower doses (maintenance), no convincing evidence is available with acute myeloid leukemia (AML) that maintenance therapy provides prolonged DFS beyond shorter-term, more dose-intensive approaches, and few current treatment clinical trials include maintenance therapy.
Nontransplant postremission therapy using cytarabine-containing regimens has treatment-related death rates that are usually less than 10% to 20% and have yielded reported long-term DFS rates from 20% to 50%.     A large, randomized trial that compared three different cytarabine-containing postremission therapy regimens showed a clear benefit in survival to patients younger than 60 years who received high-dose cytarabine.  Intensification of cytarabine dose or duration of postremission chemotherapy with conventionally dosed cytarabine did not improve DFS or OS in patients aged 60 years or older, as evidenced in the Medical Research Council (MRC-LEUK-AML11) trial.   The duration of postremission therapy has ranged from one cycle   to four or more cycles.   The standard postremission therapy for AML patients in remission is high-dose cytarabine; however, there exists some controversy about whether it benefits all younger AML patients in first complete response versus selected subgroups, such as those with core-binding factor abnormalities.      The optimal doses, schedules, and duration of postremission chemotherapy have not been determined. Therefore, to address these issues, patients with AML should be included in clinical trials at institutions that treat large numbers of such patients.
Dose-intensive cytarabine-based chemotherapy can be complicated by severe neurologic  and/or pulmonary toxic effects  and should be administered by physicians experienced in these regimens at centers that are equipped to deal with potential complications. In a retrospective analysis of 256 patients who received high-dose bolus cytarabine at a single institution, the most powerful predictor of cytarabine neurotoxicity was renal insufficiency. The incidence of neurotoxicity was significantly greater in patients treated with twice daily doses of 3 g/m2/dose when compared with 2 g/m2/dose.
Allogeneic bone marrow transplantation (BMT) results in the lowest incidence of leukemic relapse, even when compared with BMT from an identical twin (syngeneic BMT). This has led to the concept of an immunologic graft-versus-leukemia effect, similar to (and related to) graft-versus-host disease. The improvement in freedom from relapse using allogeneic BMT as the primary postremission therapy is offset, at least in part, by the increased morbidity and mortality caused by graft-versus-host disease, veno-occlusive disease of the liver, and interstitial pneumonitis. The DFS rates using allogeneic transplantation in first complete remission (CR) have ranged from 45% to 60%.    The use of allogeneic BMT as primary postremission therapy is limited by the need for a human leukocyte antigen (HLA)-matched sibling donor and the increased mortality from allogeneic BMT of patients who are older than 50 years. The mortality from allogeneic BMT that uses an HLA-matched sibling donor ranges from 20% to 40%, depending on the series. The use of matched, unrelated donors for allogeneic BMT is being evaluated at many centers but has a very substantial rate of treatment-related mortality, with DFS rates less than 35%.  Retrospective analysis of data from the International Bone Marrow Transplant Registry suggests that postremission chemotherapy does not lead to an improvement in DFS or OS for patients in first remission undergoing allogeneic BMT from an HLA-identical sibling.  [Level of evidence: 3iiiA]
A common clinical trial design used to evaluate the benefit of allogeneic transplant as consolidation therapy for AML in first remission is the so-called donor-no donor comparison. In this design, newly diagnosed AML patients who achieve a CR have one or more siblings, and are deemed medically eligible for allogeneic transplant, undergo HLA typing. If a sibling donor is identified, the patient is allocated to the transplantation arm. Analysis of outcome is by intention to treat; that is, patients assigned to the donor arm who do not receive a transplant are grouped in the analysis with the patients who did actually receive a transplant. Relapse-free survival (RFS) is the usual endpoint for this type of trial. Overall survival (OS) from the time of diagnosis is less frequently reported in these trials. Results of these trials have been mixed, with some trials showing a clear benefit across all cytogenetic subgroups, and others showing no benefit.
Investigators attempted to address this issue with a meta-analysis using data from 18 separate prospective trials of AML patients using the donor-no donor design, with data from an additional six trials included for sensitivity analysis.  The trials included in this meta-analysis enrolled adult patients aged 60 and younger during the years 1982 to 2006. Median follow-up ranged from 42 months to 142 months. Preparative regimens were similar among the different trials. Allogeneic transplant was compared with autologous transplant (6 trials) or with a variety of consolidation chemotherapy regimens, with high-dose cytarabine being the most common.
Treatment-related mortality ranged from 5% to 42% in the donor groups compared with 3% to 27% in the no-donor group. Of 18 trials reporting RFS across all cytogenetic risk groups, the combined hazard ratio (HR) for overall RFS benefit with allogeneic transplant was 0.80, indicating a statistically significant reduction in death or relapse in a first CR. Of the 15 trials reporting OS across all cytogenetic risk groups, the combined HR for OS was 0.90, again indicating a statistically significant reduction in death or relapse in a first CR.
In subgroup analysis according to cytogenetic risk category, there was no RFS or OS benefit of allogeneic transplant for patients with good-risk AML (RFS: HR, 1.07; 95% confidence interval [CI], 0.83–1.38; P = .59; OS: HR, 1.06; 95% CI, 0.64–1.76; P = .81). However, a transplant benefit was seen for patients with intermediate (RFS: HR, 0.83; 95% CI, 0.74–0.93; P < .01; OS: HR, 0.84; 95% CI, 0.71–0.99; P = .03) or poor-risk cytogenetics (RFS: HR, 0.73; 95% CI, 0.59–0.90; P < .01; OS: HR, 0.60; 95% CI, 0.40–0.90; P = .01). The conclusion from this meta-analysis was that allogeneic transplant from a sibling donor in a first CR is justified on the basis of improved RFS and OS for patients with intermediate- or poor-risk, but not good-risk, cytogenetics.  [Level of evidence: 2A]
An important caveat to this analysis is that induction and postremission strategies for AML among studies included in the meta-analysis were not uniform; nor were definitions of cytogenetic risk groups uniform. This may have resulted in inferior survival rates among chemotherapy-only treated patients. Most U.S. leukemia physicians agree that transplantation should be offered to AML patients in first CR in the setting of poor-risk cytogenetics and should not be offered to patients in first CR with good-risk cytogenetics.
The use of matched, unrelated donors for allogeneic BMT is being evaluated at many centers but has a very substantial rate of treatment-related mortality, with DFS rates less than 35%.  Retrospective analysis of data from the International Bone Marrow Transplant Registry suggests that postremission chemotherapy does not lead to an improvement in DFS or OS for patients in first remission undergoing allogeneic BMT from an HLA-identical sibling.  [Level of evidence: 3iiiA]
Autologous BMT yielded DFS rates between 35% and 50% in patients with AML in first remission. Autologous BMT has also cured a smaller proportion of patients in second remission.        Treatment-related mortality rates of patients who have had autologous peripheral blood or marrow transplantation range from 10% to 20%. Ongoing controversies include the optimum timing of autologous stem cell transplantation, whether it should be preceded by postremission chemotherapy, and the role of ex vivo treatment of the graft with chemotherapy, such as 4-hydroperoxycyclophosphamide (4-HC)  or mafosphamide,  or monoclonal antibodies, such as anti-CD33.  Purged marrows have demonstrated delayed hematopoietic recovery; however, most studies that use unpurged marrow grafts have included several cycles of postremission chemotherapy and may have included patients who were already cured of their leukemia.
In a prospective trial of patients with AML in first remission, City of Hope investigators treated patients with one course of high-dose cytarabine postremission therapy, followed by unpurged autologous BMT following preparative therapy of total-body radiation therapy, etoposide, and cyclophosphamide. In an intent-to-treat analysis, actuarial DFS was approximately 50%, which is comparable to other reports of high-dose postremission therapy or purged autologous transplantation.  [Level of evidence: 3iiDii]
A randomized trial by ECOG and the Southwest Oncology Group (SWOG) compared autologous BMT using 4-HC-purged bone marrow with high-dose cytarabine postremission therapy.  No difference in DFS was found between patients treated with high-dose cytarabine, autologous BMT, or allogeneic BMT; however, OS was superior for patients treated with cytarabine compared with those who received BMT.  [Level of evidence: 1iiA]
A randomized trial has compared the use of autologous BMT in first CR with postremission chemotherapy, with the latter group eligible for autologous BMT in second CR. The two arms of the study had equivalent survival.  Two randomized trials in pediatric AML have shown no advantage of autologous transplantation following busulfan/cyclophosphamide preparative therapy and 4HC-purged graft when compared with postremission chemotherapy, including high-dose cytarabine.   An additional randomized Groupe Ouest Est d'etude des Leucemies et Autres Maladies du Sang trial (NCT01074086) of autologous BMT versus intensive postremission chemotherapy in adult AML, using unpurged bone marrow, showed no advantage to receiving autologous BMT in first remission.  Certain subsets of AML may specifically benefit from autologous BMT in first remission. In a retrospective analysis of 999 patients with de novo AML treated with allogeneic or autologous BMT in first remission in whom cytogenetic analysis at diagnosis was available, patients with poor-risk cytogenetics (abnormalities of chromosomes 5, 7, 11q, or hypodiploidy) had less favorable outcomes following allogeneic BMT than patients with normal karyotypes or other cytogenetic abnormalities. Leukemia-free survival for the patients in the poor-risk groups was approximately 20%.  [Level of evidence: 3iiiDii]
An analysis of the SWOG/ECOG (E-3489) randomized trial of postremission therapy according to cytogenetic subgroups suggested that in patients with unfavorable cytogenetics, allogeneic BMT was associated with an improved relative risk of death, whereas in the favorable cytogenetics group, autologous transplantation was superior. These data were based on analysis of small subsets of patients and were not statistically significant.  While secondary myelodysplastic syndromes have been reported following autologous BMT, the development of new clonal cytogenetic abnormalities following autologous BMT does not necessarily portend the development of secondary myelodysplastic syndromes or AML.  [Level of evidence: 3iiiDiv] Whenever possible, patients should be entered on clinical trials of postremission management.
Because BMT can cure about 30% of patients who experience relapse following chemotherapy, some investigators suggested that allogeneic BMT can be reserved for early first relapse or second CR without compromising the number of patients who are ultimately cured;  however, clinical and cytogenetic information can define certain subsets of patients with predictable better or worse prognoses in those using postremission chemotherapy.  Good-risk factors include t(8; 21), inv(16) associated with M4 AML with eosinophilia, normal karyotype with NPM1 mutation (in absence of flt-3 mutation), and normal karyotype with double cytosine-cytosine-adenosine-adenosine-thymidine (CCAAT)-enhancer binding protein (C/EBP)-alpha mutations. Poor-risk factors include deletion of 5q and 7q, trisomy 8, t(6; 9), t(9; 22), most translocations involving chromosome 11q23, and mutations of the MLL gene, a history of myelodysplasia or antecedent hematologic disorder, and normal karyotype with flt-3 mutation. Patients in the good-risk group have a reasonable chance of cure with intensive postremission therapy, and it may be reasonable to defer transplantation in that group until early first relapse. The poor-risk group is unlikely to be cured with postremission chemotherapy, and allogeneic BMT in first CR is a reasonable option for patients with an HLA-identical sibling donor. However, even with allogeneic stem cell transplantation, the outcome for patients with high-risk AML is poor (5-year DFS of 8% to 30% for patients with treatment-related leukemia or myelodysplasia).  The efficacy of autologous stem cell transplantation in the poor-risk group has not been reported to date but is the subject of active clinical trials. Patients with normal cytogenetics are in an intermediate-risk group, and postremission management should be individualized or, ideally, managed according to a clinical trial.
The rapid engraftment kinetics of peripheral blood progenitor cells demonstrated in trials of high-dose therapy for epithelial neoplasms has led to interest in the alternative use of autologous and allogeneic peripheral blood progenitor cells as rescue for myeloablative therapy for the treatment of AML. One pilot trial of the use of autologous transplantation with unpurged peripheral blood progenitor cells in first remission had a 3-year DFS rate of 35%; detailed prognostic factors for these patients were not provided.  This result appears inferior to the best results of chemotherapy or autologous BMT and suggests that the use of peripheral blood progenitor cells be limited to clinical trials.
Allogeneic stem cell transplantation can be performed using stem cells obtained from a bone marrow harvest or a peripheral blood progenitor cell harvest. In a randomized trial of 175 patients undergoing allogeneic stem cell transplantation, with either bone marrow or peripheral blood stem cells, for a variety of hematologic malignancies using methotrexate and cyclosporine to prevent graft-versus-host disease, the use of peripheral blood progenitor cells led to earlier engraftment (median neutrophil engraftment, 16 vs. 21 days; median platelet engraftment, 13 vs. 19 days).  The use of peripheral blood progenitor cells was associated with a trend toward increased graft-versus-host disease but comparable transplant-related death. The relapse rate at 2 years appeared lower in patients receiving peripheral blood progenitor cells (hazard ratio [HR], 0.49; 95% CI, 0.24–1.00); however, OS was not significantly increased (HR for death within 2 years, 0.62; 95% CI, 0.38–1.02). 
Check the list of NCI-supported cancer clinical trials that are now accepting patients with adult acute myeloid leukemia in remission. The list of clinical trials can be further narrowed by location, drug, intervention, and other criteria.
General information about clinical trials is also available from the NCI website.
No standard regimen exists for the treatment of patients with relapsed acute myeloid leukemia (AML), particularly in patients with a first remission duration of less than 1 year. 
A number of agents have activity in recurrent AML.   A combination of mitoxantrone and cytarabine was successful in 50% to 60% of patients who experienced relapse after initially obtaining a complete remission (CR).  Other studies using idarubicin and cytarabine or high-dose etoposide and cyclophosphamide reported similar results.     Mitoxantrone, etoposide, and cytarabine (MEC) demonstrated a CR induction rate of 55% in a population including 30 patients with relapsed AML, 28 patients with primary refractory AML, and 16 patients with secondary AML.  [Level of evidence: 3iiiDiv] However, in a phase III Eastern Cooperative Oncology Group (ECOG) (E-2995) trial of MEC with or without PSC388, a multidrug resistance modulator, complete response (CR) was only 17% to 25% in a population including relapse at less than 6 months after first complete remission (CR), relapse after allogeneic or autologous bone marrow transplantation (BMT), second or greater relapse, primary induction failures, secondary AML, and high-risk myelodysplastic syndromes.  [Level of evidence: 1iiDiv] Thus, treatments with new agents under clinical evaluation remain appropriate in eligible patients with recurrent AML. 
The immunotoxin gemtuzumab ozogamicin has been reported to have a 30% response rate in patients with relapsed AML expressing CD33. This included 16% of patients who achieved CRs and 13% of patients who achieved a CRp, a new response criteria defined for this trial. CRp refers to clearance of leukemic blasts from the marrow, with adequate myeloid and erythroid recovery but with incomplete platelet recovery (although platelet transfusion independence for at least 1 week was required). Unclear is whether the inadequate platelet recovery is the result of megakaryocyte toxic effects of gemtuzumab or subclinical residual leukemia. The long-term outcomes of patients who achieve CRp following gemtuzumab are not yet known. Gemtuzumab induces profound bone marrow aplasia similar to leukemia induction chemotherapy and also has substantial hepatic toxic effects, including hepatic venoocclusive disease.   The farnesyltransferase inhibitor tipifarnib (R115777) demonstrated a 32% response rate in a phase I study in patients with relapsed and refractory acute leukemia (two CRs and six partial responses in 24 patients treated) and has entered phase II trials.  Clofarabine, a novel purine nucleoside analogue, induced CR in 8 out of 19 patients in first relapse as a single agent  and in seven out of 29 patients when administered in combination with intermediate-dose cytarabine.  [Level of evidence: 3iiiDiv]
A subset of relapsed patients treated aggressively may have extended disease-free survival (DFS); however, cures in patients following a relapse are thought to be more commonly achieved using BMT.  [Level of evidence: 3iDii] A retrospective study from the International Bone Marrow Transplant Registry compared adults younger than 50 years with AML in second CR who received HLA-matched sibling transplantation versus a variety of postremission approaches.  The chemotherapy approaches were heterogeneous; some patients received no postremission therapy. The transplantation regimens were similarly diverse. Leukemia-free survival appeared to be superior for patients receiving BMTs for two groups: patients older than 30 years whose first remission was less than 1 year; and patients younger than 30 years whose first remission was longer than 1 year.  [Level of evidence: 3iDii]
Allogeneic BMT from an HLA-matched donor in early first relapse or in second CR provides a DFS rate of approximately 30%.  [Level of evidence: 3iiiA] Transplantation in early first relapse potentially avoids the toxic effects of reinduction chemotherapy.    Allogeneic BMT can salvage some patients whose disease fails to go into remission with intensive chemotherapy (primary refractory leukemia). Nine of 21 patients with primary refractory AML were alive and disease free at 10 years following allogeneic BMT.  [Level of evidence: 3iiiA] Randomized trials testing the efficacy of this approach are not available. Autologous BMT is an option for patients in second CR, offering a DFS that may be comparable to autografting in first CR.   
Patients who relapse following an allogeneic BMT may undergo an infusion of lymphocytes from the donor (donor lymphocyte infusion or DLI), similar to the therapy patients with relapsing chronic myelogenous leukemia (CML) undergo. (Refer to the Relapsing Chronic Myelogenous Leukemia section of the PDQ summary on Chronic Myelogenous Leukemia Treatment for more information.) There are no published studies of any prospective trials examining the role of DLI for patients with AML who relapsed following allogeneic BMT. A retrospective study of European patients found that, out of 399 patients who relapsed after an allogeneic BMT, 171 patients received DLI as part of their salvage therapy.  A multivariate analysis of survival showed a significant advantage for the 171 DLI recipients, who achieved a 2-year overall survival from the time of relapse of 21%, compared to 9% for the 228 patients who did not receive DLI (P < .04; RR, 0.8; 95% confidence interval, 0.64–0.99).  [Level of evidence: 3iiiA] The strength of this finding is limited by the retrospective nature of the study, and the possibility that much of the survival advantage could have been the result of selection bias. Furthermore, the remission rate of 34% reported in this study was considerably less than the 67% to 91% reported for CML.  Therefore, even if the survival advantage conferred by DLI is real, the fraction of relapsed AML patients who might benefit from this therapy appears to be quite limited.
Arsenic trioxide, an agent with both differentiation-inducing and apoptosis-inducing properties against acute promyelocytic leukemia (APL) cells, has a high rate of successful remission induction in patients with relapsed APL. Clinical CRs have been reported in 85% of patients induced with arsenic trioxide, with a median time to clinical CR of 59 days. Eighty-six percent of evaluable patients tested negative for the presence of PML-RARA transcript after induction or postremission therapy with arsenic trioxide. Actuarial 18-month relapse-free survival was 56%. Induction with arsenic trioxide may be complicated by APL differentiation syndrome (identical to ATRA syndrome), prolongation of QT interval, and neuropathy.   Arsenic trioxide is now being incorporated into the postremission treatment strategy of de novo APL patients in clinical trials.
Some patients induced into second remissions with ATO have experienced long-term DFS following autologous stem cell transplantation. 
Check the list of NCI-supported cancer clinical trials that are now accepting patients with recurrent adult acute myeloid leukemia. The list of clinical trials can be further narrowed by location, drug, intervention, and other criteria.
General information about clinical trials is also available from the NCI website.
The PDQ cancer information summaries are reviewed regularly and updated as new information becomes available. This section describes the latest changes made to this summary as of the date above.
Updated statistics with estimated new cases and deaths for 2017 (cited American Cancer Society as reference 1).
This summary is written and maintained by the PDQ Adult Treatment Editorial Board, which is editorially independent of NCI. The summary reflects an independent review of the literature and does not represent a policy statement of NCI or NIH. More information about summary policies and the role of the PDQ Editorial Boards in maintaining the PDQ summaries can be found on the About This PDQ Summary and PDQ® - NCI's Comprehensive Cancer Database pages.
This PDQ cancer information summary for health professionals provides comprehensive, peer-reviewed, evidence-based information about the treatment of acute myeloid leukemia. It is intended as a resource to inform and assist clinicians who care for cancer patients. It does not provide formal guidelines or recommendations for making health care decisions.
This summary is reviewed regularly and updated as necessary by the PDQ Adult Treatment Editorial Board, which is editorially independent of the National Cancer Institute (NCI). The summary reflects an independent review of the literature and does not represent a policy statement of NCI or the National Institutes of Health (NIH).
Board members review recently published articles each month to determine whether an article should:
Changes to the summaries are made through a consensus process in which Board members evaluate the strength of the evidence in the published articles and determine how the article should be included in the summary.
The lead reviewers for Adult Acute Myeloid Leukemia Treatment are:
Any comments or questions about the summary content should be submitted to Cancer.gov through the NCI website's Email Us. Do not contact the individual Board Members with questions or comments about the summaries. Board members will not respond to individual inquiries.
Some of the reference citations in this summary are accompanied by a level-of-evidence designation. These designations are intended to help readers assess the strength of the evidence supporting the use of specific interventions or approaches. The PDQ Adult Treatment Editorial Board uses a formal evidence ranking system in developing its level-of-evidence designations.
PDQ is a registered trademark. Although the content of PDQ documents can be used freely as text, it cannot be identified as an NCI PDQ cancer information summary unless it is presented in its entirety and is regularly updated. However, an author would be permitted to write a sentence such as “NCI’s PDQ cancer information summary about breast cancer prevention states the risks succinctly: [include excerpt from the summary].”
The preferred citation for this PDQ summary is:
PDQ® Adult Treatment Editorial Board. PDQ Adult Acute Myeloid Leukemia Treatment. Bethesda, MD: National Cancer Institute. Updated <MM/DD/YYYY>. Available at: http://www.cancer.gov/types/leukemia/hp/adult-aml-treatment-pdq. Accessed <MM/DD/YYYY>. [PMID: 26389432]
Images in this summary are used with permission of the author(s), artist, and/or publisher for use within the PDQ summaries only. Permission to use images outside the context of PDQ information must be obtained from the owner(s) and cannot be granted by the National Cancer Institute. Information about using the illustrations in this summary, along with many other cancer-related images, is available in Visuals Online, a collection of over 2,000 scientific images.
Based on the strength of the available evidence, treatment options may be described as either “standard” or “under clinical evaluation.” These classifications should not be used as a basis for insurance reimbursement determinations. More information on insurance coverage is available on Cancer.gov on the Managing Cancer Care page.
More information about contacting us or receiving help with the Cancer.gov website can be found on our Contact Us for Help page. Questions can also be submitted to Cancer.gov through the website’s Email Us.