Acute myeloid leukemia: Pathogenesis

INTRODUCTION — 

Acute myeloid leukemia (AML) develops as the consequence of a series of genetic changes in a hematopoietic precursor cell. These changes alter normal hematopoietic growth and differentiation, resulting in an accumulation of large numbers of abnormal, immature myeloid cells in the bone marrow and peripheral blood. These cells are capable of dividing and proliferating, but cannot differentiate into mature hematopoietic cells (ie, neutrophils).

This topic reviews the pathogenesis of AML.

Related topics include:

●(See "Acute myeloid leukemia: Molecular genetics".)

●(See "Acute myeloid leukemia: Cytogenetic abnormalities".)

●(See "Acute myeloid leukemia: Risk factors and prognosis".)

●(See "Familial disorders of acute leukemia and myelodysplastic syndromes".)

CELL OF ORIGIN

Normal counterpart — AML refers to a heterogeneous group of diseases characterized by clonal cells that exhibit maturation defects that correspond to stages in hematopoietic differentiation.

Hematopoietic stem cells are multipotent and have the capacity to differentiate into the cells of all blood lineages: erythrocytes, platelets, neutrophils, eosinophils, basophils, monocytes, T and B lymphocytes, natural killer cells, and dendritic cells. In order to sustain hematopoiesis, stem cells are part of a developmental hierarchy capable of three basic functions:

●Maintenance in a noncycling state (ie, not actively progressing through the cell cycle)

●Self-renewal, allowing the production of additional stem cells

●Production of committed progenitor cells

These progenitor cells commit to subsets of myeloid and lymphoid lineages, and ultimately to single developmental pathways, resulting in the expression of the terminally differentiated stage of each cell type (figure 1) [1,2]. (See "Overview of hematopoietic stem cells".)

Normal hematopoiesis is a dynamic, highly regulated process controlled by the combined effects of growth factors that permit cellular proliferation, and transcription factors that activate specific genetic programs, resulting in commitment to a specific lineage and in terminal differentiation (figure 2). Regulatory growth factors and specific transcription factors have been identified that play critical roles in lineage commitment and in the subsequent development of the mature lymphoid and myeloid (erythroid, granulocytic, monocytic, and megakaryocytic) lineages [3,4].

A number of genes encoding these transcription factors are involved in recurring chromosomal translocations seen in AML, suggesting that the AML variants arise because the translocations alter regulatory processes that control growth and differentiation programs [5]. The novel fusion genes created by these translocations are reviewed separately. (See "Acute myeloid leukemia: Molecular genetics".)

Clonality — AML is a clonal process that develops from a single transformed hematopoietic progenitor cell.

Virtually all cases of AML are thought to be preceded by a premalignant proliferative disorder characterized by clonal hematopoiesis without other evidence of a malignancy [6-8]. Such clonal hematopoiesis of indeterminate potential (CHIP) increases in association with age and most commonly involves a mutation in DNMT3A, TET2, or ASXL1 (mutations that are associated with myeloid malignancies). While clonal hematopoiesis was associated with an increased risk of hematologic cancer, the absolute risk of progression is small. (See "Clonal hematopoiesis of indeterminate potential (CHIP) and related disorders of clonal hematopoiesis".)

Leukemic stem cells — AML is a heterogeneous disease with leukemic cells of different subtypes resembling normal cells at various stages of maturation.

Nevertheless, only a limited number of cells within the bulk population of leukemia cells have the capacity to function as stem cells that maintain the potential for unlimited self-renewal [9]. These leukemic stem cells (LSC, also leukemia-initiating cells) may be more immature than the majority of circulating leukemic cells and are thought to have originated from cells with existing self-renewal capacity or from progenitors that have re-acquired this stem cell-like property.

Stage of leukemic transformation

Two models have been proposed to explain the heterogeneity of AML observed at the molecular, cytogenetic, phenotypic, and clinical level.

●Transformation to leukemia occurring at one of several developmental stages

●Transformation to leukemia occurring within primitive multipotent cells

There is no universal agreement as to which of these two models best reflects the truth.

Transformation at one of several developmental stages — This model proposes that any cell type within the stem cell/progenitor cell hierarchy, from primitive multipotent stem cell to lineage-committed progenitor cell, is susceptible to leukemic transformation, resulting in the expansion of abnormal cells that exhibit different stages of differentiation. For AML, this model predicts that the phenotype of the leukemic stem cells restricted to the granulocytic-monocytic series differs from that of cells in the erythroid or megakaryocytic lineages (figure 1).

The correlation between specific cytogenetic and molecular genetic aberrations and the morphologic appearance of leukemic cells might suggest that the transforming event occurs at different stages of myeloid differentiation. This hypothesis is underscored by the classification for AML, which distinguishes some subtypes of AML based upon the stage of apparent differentiation (eg, AML with minimal differentiation). Support for this model includes flow cytometric/molecular analyses of the leukemic cell in acute promyelocytic leukemia (APL) that suggest that the leukemic cell arises in a committed lineage-restricted, CD34+/CD38+ progenitor cell [10].

Transformation within primitive multipotent cells — A second model proposes that mutations responsible for leukemic transformation and progression occur only in primitive multipotent stem cells, with disease heterogeneity resulting from a variable ability of these primitive stem cells to differentiate and acquire specific phenotypic lineage markers [11,12].

Hematopoietic stem cells express a characteristic cell surface antigen (CD34), and can be further subdivided by the expression of additional cell surface antigens, including CD38 and HLA-DR [13,14]:

●CD34+/CD38-/HLA-DR- cells are multipotential hematopoietic stem cells, give rise to mixed-lineage granulocytic-erythroid-megakaryocytic colonies in culture, can repopulate immune-deficient mice with normal hematopoietic cells in vivo, and demonstrate self-renewal capacity, as assessed by their ability to be serially transplanted into recipient mice. There are data to suggest that, in some cases of AML, the leukemic stem cell may be quite similar to normal hematopoietic stem cells [15].

●CD34+/CD38+/HLA-DR+ cells define a committed population of myeloid progenitor cells [16]. (See "Overview of hematopoietic stem cells".)

Cytogenetic and FISH (fluorescence in situ hybridization) studies of sorted stem cell compartments from patients with AML evolving from a prior myelodysplastic syndrome and patients with de novo AML have shown that the characteristic cytogenetic abnormality from both groups was present in the CD34+/CD38- multipotential stem cell compartment [17-19]. Similar findings were noted in patients with the 5q- syndrome [20] and monosomy 7 [21], myelodysplastic disorders with differing risks of leukemic transformation. (See "Clinical manifestations, diagnosis, and classification of myelodysplastic syndromes (MDS)".)

More compelling evidence comes from studies in which purified stem cell subpopulations from normal subjects and those with AML were transplanted into mice with severe combined immunodeficiency disease (SCID) [22]. These experiments have detected approximately one SCID mouse leukemia-initiating cell (SL-IC) in 10⁵ AML cells, which can repopulate immune-deficient mice with leukemic cells phenotypically identical to those of the AML patient from which they were derived [19,22,23].

Using a nonobese diabetic (NOD)/SCID mouse [24], SL-ICs were found to reside only in the CD34+/CD38- fraction [9]. This was consistent regardless of the AML subtype, lineage markers, or percentage of leukemic blast cells expressing the CD34 antigen. The SL-ICs also demonstrated self-renewal capacity, a requirement for maintenance of the leukemic clone. The uniformity of the leukemic stem cell phenotype strongly suggests that the leukemia-initiating transformation and progression-associated genetic events occur in primitive cells and not in committed progenitors. Similar conclusions about the site of the leukemia-initiating transformation have been made in acute lymphoblastic leukemia [25].

Additional evidence for the second model derives from the use of a retroviral gene transfer system to express AML1::ETO, a fusion gene linked to the pathogenesis of AML, in normal human hematopoietic stem and progenitor cells [26]. When this fusion gene was expressed in more mature progenitor cells, the result was growth arrest and abrogated colony formation in primary clonogenic assays. On the other hand, AML1::ETO expression in stem cells resulted in their preferential expansion and/or self-renewal [27]. (See "Acute myeloid leukemia: Molecular genetics", section on 'Cooperating mutations in AML'.)

Further supporting evidence comes from studies of clonal hematopoiesis in AML, which indicate that, although the majority of cells derived from the leukemic clone can undergo differentiation to cells of the granulocytic-monocytic lineage, they also may differentiate into cells of the erythroid and/or megakaryocytic pathways [28-31]. Such multi-lineage involvement has been noted more frequently in older patients, in secondary AML that arose from myelodysplastic syndrome, or following treatment for another malignancy [30,31].

Protection by stromal cells — AML cells develop in the bone marrow where interaction with the microenvironment may foster chemoresistance [32-35]. As an example, two studies in mouse models of APL demonstrated that dislodgement of leukemic cells from their stromal environment via interruption of CXCR4 signaling resulted in an increased sensitivity of the leukemic cells to chemotherapy [36,37]. Other studies have suggested that this abnormal microenvironment is at least partially due to leukemic cell growth that disrupts normal hematopoietic progenitor cell niches in the bone marrow [33,38].

Suppression of normal hematopoiesis — Pancytopenia is common among patients with AML and many patients initially present with symptoms related to complications of leukopenia, anemia, and thrombocytopenia. Historically, pancytopenia in AML has been attributed to the leukemic cells either displacing or killing normal hematopoietic stem cells in the bone marrow. However, subsequent studies suggest that, in patients with AML, the number of normal hematopoietic stem cells in the bone marrow is normal or increased [39]. Instead, the leukemic cells appear to inhibit the ability of the hematopoietic stem cells to produce more mature hematopoietic cells. When the stem cells were removed from the leukemic environment the ability to produce mature hematopoietic cells was restored.

Aberrant cellular metabolism — Altered cellular metabolism is a hallmark of cancer and may contribute to AML initiation and maintenance [40-42].

As an example, up to 15 to 20 percent of patients with AML have mutations in IDH1 or IDH2 that result in neomorphic enzymatic activity and production of the onco-metabolite, 2-hydroxyglutarate (2HG) from alpha-ketoglutarate [40]. 2HG inhibits multiple alpha-ketoglutarate-dependent dioxygenase reactions, leads to aberrant deoxyribonucleic acid (DNA) hypermethylation and a differentiation block in myeloid precursors, and promotes leukemogenesis. Allosteric inhibitors of the mutant IDH isoforms, ivosidenib (IDH1) and enasidenib (IDH2) can overcome the differentiation block and have been approved for treatment.

TWO-HIT HYPOTHESIS OF LEUKEMOGENESIS — 

Progression to acute leukemia may require a series of genetic events beginning with clonal expansion of a transformed leukemic stem cell [43-45]. The specific mutational event(s) required for this progression are not currently well defined. Single nucleotide polymorphism (SNP) studies have been performed on tumor samples from adults and children with AML as a genomic screen to locate previously unidentified genes that may play a role in the pathobiology of AML [46,47].

The "two-hit hypothesis" of leukemogenesis implies that AML is the consequence of at least two mutations, one conferring a proliferative advantage (class I mutations) and another impairing hematopoietic differentiation (class II mutations) [48]. Type I mutations include those of FLT3-ITD, K-RAS mutations, and KIT mutations, while mutations in CEBPA are type II abnormalities [49]. (See "Acute myeloid leukemia: Risk factors and prognosis".)

On average, AML clones have 8 to 13 mutations found within the coding regions of the genome. The accumulation of these lesions in a step-wise process within a hematopoietic stem cell was demonstrated in a study that analyzed gene mutations found in primary tumor-relapse pairs of de novo AML and patient-matched skin samples [50]. Identification of individual cells containing subsets of these mutations allowed for the identification of mutations that occurred early and late in the process. This finding suggests not only that several hits are required for the development of AML, but also that relapsed disease can represent the further replication of the original dominant clone, the emergence of a minor clone present at diagnosis [51,52], or further evolution of a preleukemic clone after treatment (figure 3).

●Whole-genome or whole-exome sequencing of 200 cases of de novo AML reported an average of 13 gene mutations per tumor [53]. Of these, each tumor had an average of five genes known to be one of a group of 23 genes recurrently mutated in AML that can be broadly grouped into nine categories of genes thought to be involved in leukemogenesis. Mutations were found in genes associated with transcription-factor fusions (18 percent), nucleophosmin (27 percent), tumor suppression (16 percent), DNA-methylation (44 percent), signaling (59 percent), chromatin-modification (30 percent), myeloid transcription factor (22 percent), the cohesin-complex (13 percent), and the spliceosome complex (14 percent). Some mutation pairs occurred more commonly than expected (eg, NPM1 and FLT3), suggesting synergy, while others were mutually exclusive, suggesting duplicative pathways. (See "Acute myeloid leukemia: Molecular genetics", section on 'Gene mutations'.)

●In one study of seven patients with AML that had evolved from myelodysplastic syndrome, approximately 85 percent of bone marrow cells were clonal at the time of myelodysplastic syndrome diagnosis [54]. Whole-genome sequencing of paired skin and bone marrow samples identified 11 recurrently mutated genes. Genotyping of bone marrow samples from the same patients collected at the time of AML diagnosis identified those mutations that were present at the time of myelodysplastic syndrome diagnosis (ie, NPM1, RUNX1, SMC3, STAG2, TP53, U2AF1, UMODL1, and ZSWIM4) and those that developed subsequently (ie, CDH23, PTPN11, WT1).

A retrospective study that compared mutational analyses of paired diagnostic and relapse leukemia samples from 69 children with AML reported that 38 percent of patients had a change in their mutation status between diagnosis and relapse. Patients whose AML demonstrated a type I/II mutation at relapse had a shorter time to relapse. Consistent with the two-hit theory, expression of a chimeric protein represents only one of the genetic modifications necessary for the development of cancer and leukemia, and that the affected cell requires additional mutational events in order to express the transformed phenotype. (See "Acute myeloid leukemia: Cytogenetic abnormalities", section on 't(8;21); RUNX1::RUNX1T1'.)

As an example, the presence of the RUNX1::RUNX1T1 fusion transcript may not be sufficient, in itself, to result in AML:

●Remission bone marrow samples from patients with de novo AML (FAB -M2) with t(8;21)(q22;q22) and the RUNX1::RUNX1T1 fusion transcript have been found to harbor the aberrant fusion transcript for as many as eight years following cessation of all chemotherapy [55].

●The RUNX1::RUNX1T1 fusion transcript has been detected in bone marrow samples from patients in remission following allogeneic bone marrow transplantation for AML [56].

●In five children who developed AML with t(8;21) at 3 to 12 years of age, in whom archived blood samples for metabolic studies (Guthrie cards) were available, RUNX1::RUNX1T1 sequences were detected at birth [57]. Of interest, similar observations were made in three children developing acute lymphoblastic leukemia at three to five years of age with t(12;21) [58].

MECHANISMS OF GENETIC DAMAGE — 

Genetic changes associated with leukemogenesis can occur following chemotherapy, ionizing radiation, chemical exposure, and infection with retroviruses. In addition, certain familial disorders are associated with an increased incidence of AML.

It must be emphasized that for most patients with de novo AML, the etiologic factors contributing to the development of AML remain unknown. Interestingly, in a series of 127 patients with a previous primary malignancy and secondary AML, 30 percent did not receive any chemotherapy or radiation treatment prior to the development of AML [59].

Therapy-related AML — The development of myelodysplastic syndromes/neoplasms (MDS) and AML following chemotherapy or radiation for another condition is defined as a therapy-related myeloid neoplasm (t-MN).

●Chemotherapy – Most cases of t-MNs arise after treatment with an alkylating agent. Others arise after treatment with drugs that target DNA-topoisomerase II (eg, epipodophyllotoxins, anthracyclines). The causes, latency periods, and classification of t-MNs are discussed separately. (See "Therapy-related myeloid neoplasms: Epidemiology, causes, evaluation, and diagnosis".)

●Ionizing radiation – Ionizing radiation used to treat other malignancies (eg, Hodgkin lymphoma, breast cancer, uterine cancer, lung cancer) has also been linked to the development of AML and MDS. Radiation shares with alkylating agents the ability to damage DNA, usually by inducing double-strand breaks that can cause the mutations, deletions, or translocations that may transform hematopoietic stem cells.

This risk appears to be quite low when radiation alone is used as treatment, and is associated with the age of the patient, and doses of more than 20 Gy [60,61]. Some studies suggest that the risk of development of AML is significantly increased when the two modalities are combined, other studies demonstrated that high doses of radiotherapy confined to small volumes in combination with chemotherapy did not significantly increase leukemogenic risk [62-65].

It is unknown whether ionizing radiation used for medical examinations, such as computed tomography (CT) scans, results in an increased risk of leukemia in adults. However, exposure to x-rays and gamma rays during childhood is associated with a small absolute increase in the incidence of leukemia. In a cohort study of individuals who had CT scans when they were younger than 22 years of age, compared with those who received a cumulative radiation dose <5 mGy, the risk of subsequent leukemia was tripled for those who received a cumulative radiation dose ≥30 mGy, which is equivalent to approximately 5 to 10 head CT scans [66]. Radiation risk in children is compounded with their longer lifespan following exposure so that there is a longer time over which radiation-induced cancers can occur.

Myeloid malignancies have arisen after gene therapy, but these are not considered t-MNs. Such malignancies were reported following lentiviral gene therapy for cerebral adrenoleukodystrophy and were associated with insertional mutagenesis at MECOM-EVI1 and PRDM16 loci [67]. AML has also been reported following gene therapy for sickle cell disease, but this was not clearly attributable to insertional mutagenesis [68].

Chemical exposure — Exposure to certain chemicals has been associated with the development of AML.

High levels of benzene have been associated with an increased risk of developing AML [69,70]. A potential association between formaldehyde exposure and AML has been controversial, with conflicting data from meta-analyses of epidemiologic studies [71-74]. Except for groups exposed to high levels of benzene or radiation, the reported risks associated with occupation and chemicals have generally been less than two-fold [75]. Some studies described an association between chronic acetaminophen use and the risk of AML [76,77].

Cigarette smoking — Case-control studies have shown that cigarette smoking is associated with an increase in leukemic risk, with the greatest risk for AML noted in subjects over the age of 60 [78]. A meta-analysis of 23 studies confirmed the association of cigarette smoking with an increased risk for the development of AML [79]. The relative risks for AML correlated with increased duration of smoking and intensity of smoking.

Overweight — Several studies have linked obesity with the development of AML [80]. A large meta-analysis confirmed that overweight body mass index (BMI) and obesity are associated with an increased incidence of AML [81]. In particular, epidemiologic studies have found that an elevated BMI has a strong association with an increased risk of acute promyelocytic leukemia [82,83].

Familial acute leukemia and myelodysplasia syndromes — Most cases of AML are caused by acquired, rather than inherited, mutations.

However, there are rare pedigrees with multiple cases of AML and/or MDS. Familial leukemia can occur in the context of a medical syndrome in which AML is one component of the overall disease, or it can occur as an isolated leukemia not specifically associated with comorbid conditions [84]. Familial acute leukemias and MDS are discussed in detail separately. (See "Familial disorders of acute leukemia and myelodysplastic syndromes".)

The development of myeloid malignancies has been reported following lentiviral gene therapy for cerebral adrenoleukodystrophy [67]. Insertional mutagenesis into genes is associated with the development of AML, especially MECOM-EVI1 and PRDM16 [68]. AML has also been reported following gene therapy for sickle cell disease. However, this was not clearly attributable to insertional mutagenesis.

INFORMATION FOR PATIENTS — 

UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Beyond the Basics topics (see "Patient education: Acute myeloid leukemia (AML) treatment in adults (Beyond the Basics)")

SUMMARY

●Description – Acute myeloid leukemia (AML) develops as the consequence of a series of genetic changes in a hematopoietic precursor cell that alter normal hematopoietic growth and differentiation. This results in the accumulation of abnormal, immature myeloid cells in the bone marrow and peripheral blood. These cells can divide and proliferate, but they cannot differentiate into mature hematopoietic cells (eg, neutrophils).

●Cell of origin – AML is a heterogeneous group of diseases characterized by clonal cells that exhibit maturation defects that correspond to stages in hematopoietic differentiation. (See 'Cell of origin' above.)

●Leukemic stem cells – All leukemias, including AML, appear to be maintained by a pool of self-renewing malignant cells. These leukemic stem cells (also leukemia-initiating cells) may be more immature than the majority of circulating leukemic cells and are thought to have originated from cells with existing self-renewal capacity or from progenitors that have reacquired this stem cell-like property. (See 'Leukemic stem cells' above.)

●Two-hit hypothesis – The "two-hit hypothesis" of leukemogenesis implies that AML is the consequence of at least two mutations, one conferring a proliferative advantage (class I mutations) and another impairing hematopoietic differentiation (class II mutations). (See 'Two-hit hypothesis of leukemogenesis' above.)

●Mechanisms of leukemogenesis – Most cases of AML are caused by acquired, rather than inherited, mutations. An increased risk for the development of AML can occur following treatment with chemotherapy or ionizing radiation (therapy-related myeloid neoplasms), chemical exposure, cigarette smoking, and other factors. (See 'Mechanisms of genetic damage' above.)

Rare familial clusters of AML are associated with inherited or germline pathogenic gene variants.