Prenatal origin of childhood AML occurs less frequently than in childhood ALL

Mechanistic insights into the developmental origin of pediatric hematologic disorders

Embryonic and fetal hematopoietic stem and progenitor cells differ in some key properties from cells that are part of the adult hematopoietic system. These include higher proliferation and self-renewal capacity, different globin gene usage, and differing lineage biases. Although these evolved to cover specific requirements of embryonic development, they can have serious consequences for the pathogenesis of hematologic malignancies that initiate prebirth in fetal blood cells and may result in a particularly aggressive disease that does not respond well to treatments that have been designed for adult leukemias. This indicates that these infant/pediatric leukemias should be considered developmental diseases, where a thorough understanding of their unique biology is essential for designing more effective therapies. In this review, we will highlight some of these unique fetal properties and detail the underlying molecular drivers of these phenotypes. We specifically focus on those that are pertinent to disease pathogenesis and that may therefore reveal disease vulnerabilities. We have also included an extensive description of the origins, phenotypes, and key molecular drivers of the main infant and pediatric leukemias that have a known prenatal origin. Importantly, successes in recent years in generating faithful models of these malignancies in which cellular origins, key drivers, and potential vulnerabilities can be investigated have resulted in uncovering potential, new therapeutic avenues.

Presence of leukemic clone-specific immunoglobulin heavy chain rearrangements in neonatal blood spots of children with B-cell precursor acute lymphoblastic leukemia

INTRODUCTION

Childhood B-cell precursor acute lymphoblastic leukemia (BCP-ALL) can be traced back to birth using leukemic clone-specific immunoglobulin heavy chain (IGH) rearrangements, implying prenatal origin of this disease.

METHODS

We retrospectively analyzed neonatal blood spots (Guthrie cards) of 24 patients with childhood BCP-ALL aged 1-9.6 years (median 3.1 years) for the presence of clonotypic IGH rearrangements identified in diagnostic bone marrow samples. Based on the sequences of IGH rearrangements, 2 patient-specific primers were designed for each patient and used in semi-nested polymerase chain reaction for the detection of preleukemic clones at birth.

RESULTS

Clonotypic IGH rearrangements were detected in neonatal blood spots of 54.2% of patients (13/24). In two cases with double IGH rearrangements detected at diagnosis, only one rearrangement was present at birth, while in the third case both leukemic rearrangements were detected in neonatal blood. Guthrie card-positive findings were significantly more frequent in children ≤5 years of age than in older children (p = 0.011). Regarding patients' characteristics at birth and at diagnosis, Guthrie card-positivity was not associated with sex, birth weight and mother's age, as well as with white blood cell count, percentage of bone marrow blasts, immunophenotype and the presence of ETV6/RUNX1 and TCF3/PBX1 fusion genes at diagnosis.

CONCLUSION

Our study confirms that a large proportion of childhood BCP-ALL originates in utero, regardless of the molecular subtype defined by chromosomal aberrations. The observed trend toward younger age at diagnosis in Guthrie card-positive versus Guthrie card-negative patients implies that the age at diagnosis depends on the presence of preleukemic clone at birth, as well as on the timing of postnatal transforming genetic events.

The Landscape of Secondary Genetic Rearrangements in Pediatric Patients with B-Cell Acute Lymphoblastic Leukemia with t(12;21)

The most frequent chromosomal rearrangement in childhood B-cell acute lymphoblastic leukemia (B-ALL) is translocation t(12;21)(p13;q22). It results in the fusion of the ETV6::RUNX1 gene, which is active in the regulation of multiple crucial cellular pathways. Recent studies hypothesize that many translocations are influenced by RAG-initiated deletions, as well as defects in the RAS and NRAS pathways. According to a "two-hit" model for the molecular pathogenesis of pediatric ETV6::RUNX1-positive B-ALL, the t(12;21) translocation requires leukemia-causing secondary mutations. Patients with ETV6::RUNX1 express up to 60 different aberrations, which highlights the heterogeneity of this B-ALL subtype and is reflected in differences in patient response to treatment and chances of relapse. Most studies of secondary genetic changes have concentrated on deletions of the normal, non-rearranged ETV6 allele. Other predominant structural changes included deletions of chromosomes 6q and 9p, loss of entire chromosomes X, 8, and 13, duplications of chromosome 4q, or trisomy of chromosomes 21 and 16, but the impact of these changes on overall survival remains unclarified. An equally genetically diverse group is the recently identified new B-ALL subtype ETV6::RUNX1-like ALL. In our review, we provide a comprehensive description of recurrent secondary mutations in pediatric B-ALL with t(12;21) to emphasize the value of investigating detailed molecular mechanisms in ETV6::RUNX1-positive B-ALL, both for our understanding of the etiology of the disease and for future clinical advances in patient treatment and management.

The Prenatal Origin of Childhood Leukemia: Potential Applications for Epidemiology and Newborn Screening

Childhood leukemias are heterogeneous diseases with widely differing incident rates worldwide. As circulating tumors, childhood acute leukemias are uniquely accessible, and their natural history has been described in greater detail than for solid tumors. For several decades, it has been apparent that most cases of childhood acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML) initiate in utero. Circumstantial evidence in support of this contention includes the young age of onset and high rate of concordance among identical twins. "Backtracking" of leukemic somatic mutations, particularly gene translocations, to cord blood and dried blood spots collected during the perinatal period has provided molecular proof of prenatal leukemogenesis. Detection of a patient's leukemia translocation in easily accessible birth samples, such as dried blood spots, is straightforward with the knowledge of their idiosyncratic breakpoints. However, to translate these findings into population-based screening and leukemia prevention requires novel methods able to detect translocations at all possible breakpoints when present in a low frequency of cells. Several studies have attempted to screen for leukemic translocations, mainly the common ETV6-RUNX1 translocation, in cord blood samples from healthy children. Most studies have reported finding translocations in healthy children, but estimates of prevalence have varied widely and greatly exceed the incidence of leukemia, leading to concerns that technical artifact or contamination produced an artificially inflated estimate of translocation prevalence at birth. New generation techniques that capture the presence of these translocations at birth have the potential to vastly increase our understanding of the epidemiology of acute leukemias. For instance, if leukemic translocations are present at birth in a far higher proportion of children than eventually develop acute leukemia, what are the exposures and somatic molecular events that lead to disease? And could children with translocations present at birth be targeted for prevention of disease? These questions must be answered before large-scale newborn screening for leukemia can occur as a public health initiative. Here, we review the literature regarding backtracking of acute leukemias and the prevalence of leukemic translocations at birth. We further suggest an agenda for epidemiologic research using new tools for population screening of leukemic translocations.

Insights into the prenatal origin of childhood acute lymphoblastic leukemia

Pediatric acute lymphoblastic leukemia (ALL) is defined by recurrent chromosomal aberrations including hyperdiploidy and chromosomal translocations. Many of these aberrations originate in utero and the cells transform in early childhood through acquired secondary mutations. In this review, we will discuss the most common prenatal lesions that can lead to childhood ALL, with a special emphasis on the most common translocation in childhood ALL, t(12;21), which results in the ETV6-RUNX1 gene fusion. The ETV6-RUNX1 fusion arises prenatally and at a 500-fold higher frequency than the corresponding ALL. Even though the findings regarding the frequency of ETV6-RUNX1 were originally challenged, newer studies have confirmed the higher frequency. The prenatal origin has also been proven for other gene fusions, including KMT2A, the translocations t(1;19) and t(9;22) leading to TCF3-PBX1 and BCR-ABL1, respectively, as well as high hyperdiploidy. For most of these aberrations, there is evidence for more frequent occurrence than the corresponding leukemia incidences. We will briefly discuss what is known about the cells of origin, the mechanisms of leukemic transformation through lack of immunosurveillance, and why only a part of the carriers develops ALL.

Backtracked analysis of preleukemic fusion genes and DNA repair foci in umbilical cord blood of children with acute leukemia

The first event in origination of many childhood leukemias is a specific preleukemic fusion gene (PFG) that arises, often in utero, in hematopoietic stem/progenitor cells (HSPC) from misrepaired DNA double strand break (DSB). An immanently elevated level of DSB and impaired apoptosis may contribute to origination and persistence of PFG and donor cell-derived leukemia in recipients of allogeneic transplantation of umbilical cord blood (UCB). We investigated DSB, apoptosis and PFG in the backtracked UCB cells of leukemic patients. RNA from UCB of three patients with acute lymphoblastic leukemia, patient with acute megakaryoblastic leukemia and Down syndrome, and four healthy children was screened for common PFG by RT-qPCR. Presence of PFG was validated by sequencing. Endogenous γH2AX and 53BP1 DNA repair foci, cell populations, and apoptosis were analyzed in UCB CD34+/- cells with imaging and standard flow cytometry. We found MLL₂-AF4 and BCR-ABL (p190) fusion genes in UCB of two out from four pediatric patients, apparently not detected at diagnosis, while UCB cells of TEL-AML1+ ALL patient were tested negative for this PFG and no PFG were detected in UCB cells of healthy children. No significant difference in DNA damage and apoptosis between UCB CD34+/- cells from healthy children and leukemic patients was observed, while Down syndrome trisomy increased DNA damage and resulted in distribution of cell populations resembling transient abnormal myelopoiesis. Our findings indicate increased genetic instability in UCB HSPC of leukemic patients and may be potentially used for diagnostics and exclusion of possibly affected UCB from transplantation.

Long-term remission of therapy-related acute myeloid leukemia with a new t(11;18)(q23;q21.2) translocation and KMT2A-ME2 (MLL-ME2) fusion gene

We describe a unique case of a woman with acute myeloid leukemia with a new, previously undescribed translocation, t(11;18)(q23;q21.2), affecting the KMT2A (MLL) gene and resulting in an KMT2A(MLL)-ME2 fusion. This disease occurred secondarily following chemotherapy for a different acute myeloid leukemia with the recurrent genetic abnormality inv(16)(p13.1;q22). The secondary leukemia was treated with intensive chemotherapy without allogeneic hematopoietic cell transplantation. Complete remission lasting more than 10 years has been achieved with concurrent and sustained remission of the primary leukemia.

Different aspects of stem cell procedures in children with poor responding AML: when is HSCT the best answer?

Acute myeloid leukemia in children is a heterogeneous disease with different morphological and cytogenetic features. New diagnostic tools and treatments, improved supportive care and the use of genomic tissue typing in selecting donors for hematopoietic stem cell transplantation (HSCT) adds to increased survival rates. Candidates to HSCT in first complete remission are patients with cytogenetic or molecular unfavorable prognostic markers, or blasts >15% after first induction. The use of minimal residual disease can also identify children benefiting from HSCT in first complete remission and the patients post HSCT with signs of relapse. The outcome and cure rate of acute myeloid leukemia, still remains poor and new diagnostic tools and treatments strategies need to be evaluated. In this management perspective, future management of novel minimal residual disease tools are discussed, conditioning therapies, as well as different transplantation procedures including haplo-transplantation and haplo-identical natural killer cell transplantation, but also altered graft-versus-host-disease treatments.

Global characteristics of childhood acute promyelocytic leukemia

Acute promyelocytic leukemia (APL) comprises approximately 5-10% of childhood acute myeloid leukemia (AML) cases in the US. While variation in this percentage among other populations was noted previously, global patterns of childhood APL have not been thoroughly characterized. In this comprehensive review of childhood APL, we examined its geographic pattern and the potential contribution of environmental factors to observed variation. In 142 studies (spanning >60 countries) identified, variation was apparent-de novo APL represented from 2% (Switzerland) to >50% (Nicaragua) of childhood AML in different geographic regions. Because a limited number of previous studies addressed specific environmental exposures that potentially underlie childhood APL development, we gathered 28 childhood cases of therapy-related APL, which exemplified associations between prior exposures to chemotherapeutic drugs/radiation and APL diagnosis. Future population-based studies examining childhood APL patterns and the potential association with specific environmental exposures and other risk factors are needed.

Germinal and Somatic Trisomy 21 Mosaicism: How Common is it, What are the Implications for Individual Carriers and How Does it Come About?

It is well known that varying degrees of mosaicism for Trisomy 21, primarily a combination of normal and Trisomy 21 cells within individual tissues, may exist in the human population. This involves both Trisomy 21 mosaicism occurring in the germ line and Trisomy 21 mosaicism documented in different somatic tissues, or indeed a combination of both in the same subjects. Information on the incidence of Trisomy 21 mosaicism in different tissue samples from people with clinical features of Down syndrome as well as in the general population is, however, still limited. One of the main reasons for this lack of detailed knowledge is the technological problem of its identification, where in particular low grade/cryptic Trisomy 21 mosaicism, i.e. occurring in less than 3-5% of the respective tissues, can only be ascertained by fluorescence in situ hybridization (FISH) methods on large cell populations from the different tissue samples.In this review we summarize current knowledge in this field with special reference to the question on the likely incidence of germinal and somatic Trisomy 21 mosaicism in the general population and its mechanisms of origin. We also highlight the reproductive and clinical implications of this type of aneuploidy mosaicism for individual carriers. We conclude that the risk of begetting a child with Trisomy 21 Down syndrome most likely is related to the incidence of Trisomy 21 cells in the germ line of any carrier parent. The clinical implications for individual carriers may likewise be dependent on the incidence of Trisomy 21 in the relevant somatic tissues. Remarkably, for example, there are indications that Trisomy 21 mosaicism will predispose carriers to conditions such as childhood leukemia and Alzheimer's Disease but there is on the other hand a possibility that the risk of solid cancers may be substantially reduced.

Residential pesticides and childhood leukemia: a systematic review and meta-analysis

It is a systematic review and meta-analysis of previous observational epidemiologic studies examining the relationship between residential pesticide exposures during critical exposure time windows (preconception, pregnancy, and childhood) and childhood leukemia. Searches of Medline and other electronic databases were performed (1950-2009). Study selection, data abstraction, and quality assessment were performed by two independent reviewers. Random effects models were used to obtain summary odds ratios (ORs) and 95% confidence intervals (Cis). Of the 17 identified studies, 15 were included in the meta-analysis. Exposures during pregnancy to unspecified residential pesticides insecticides, and herbicides were positively associated with childhood leukemia. Exposures during childhood to unspecified residential pesticides and insecticides were also positively associated with childhood leukemia, but there was no association with herbicides. Positive associations were observed between childhood leukemia and residential pesticide exposures. Further work is needed to confirm previous findings based on self-report, to examine potential exposure-response relationships, and to assess specific pesticides and toxicologically related subgroups of pesticides in more detail.

Acute leukemias with ETV6/ABL1 (TEL/ABL) fusion: poor prognosis and prenatal origin

The ETV6/ABL1 (TEL/ABL) fusion gene is a rare aberration in malignant disorders. Only 19 cases of ETV6/ABL1-positive hematological malignancy have been published, diagnosed with chronic myeloid leukemia, other types of chronic myeloproliferative neoplasm, acute myeloid leukemia or acute lymphoblastic leukemia (ALL). This study reports three new cases (aged 8 months, 5 years, and 33 years) of ALL with the ETV6/ABL1 fusion found by screening 392 newly diagnosed ALL patients (335 children and 57 adults). A thorough review of the literature and an analysis of all published data, including the three new cases, suggest poor prognosis of ETV6/ABL1-positive acute leukemias. The course of the disease in the two pediatric patients is characterized by minimal residual disease monitoring, using quantification of both the ETV6/ABL1 transcript and immunoreceptor gene rearrangements. Eosinophilia could not be confirmed as a hallmark of the ETV6/ABL1-positive disease. Studies of neonatal blood spots demonstrated that, in the child diagnosed at five years, the ETV6/ABL1 fusion initiating the ALL originated prenatally.

FLT3 mutation incidence and timing of origin in a population case series of pediatric leukemia

Background

Mutations in FLT3 result in activated tyrosine kinase activity, cell growth stimulation, and a poor prognosis among various subtypes of leukemia. The causes and timing of the mutations are not currently known. We evaluated the prevalence and timing of origin of FLT3 mutations in a population series of childhood leukemia patients from Northern California.

Methods

We screened and sequenced FLT3 mutations (point mutations and internal tandem duplications, ITDs) among 517 childhood leukemia patients, and assessed whether these mutations occurred before or after birth using sensitive "backtracking" methods.

Results

We determined a mutation prevalence of 9 of 73 acute myeloid leukemias (AMLs, 12%) and 9 of 441 acute lymphocytic leukemias (ALLs, 2%). Among AMLs, FLT3 mutations were more common in older patients, and among ALLs, FLT3 mutations were more common in patients with high hyperdiploidy (3.7%) than those without this cytogenetic feature (1.4%). Five FLT3 ITDs, one deletion mutation, and 3 point mutations were assessed for their presence in neonatal Guthrie spots using sensitive real-time PCR techniques, and no patients were found to harbor FLT3 mutations at birth.

Conclusions

FLT3 mutations were not common in our population-based patient series in California, and patients who harbor FLT3 mutations most likely acquire them after they are born.

Residential pesticides and childhood leukemia: a systematic review and meta-analysis

OBJECTIVE

We conducted a systematic review and meta-analysis of previous observational epidemiologic studies examining the relationship between residential pesticide exposures during critical exposure time windows (preconception, pregnancy, and childhood) and childhood leukemia.

DATA SOURCES

Searches of MEDLINE and other electronic databases were performed (1950-2009). Reports were included if they were original epidemiologic studies of childhood leukemia, followed a case-control or cohort design, and assessed at least one index of residential/household pesticide exposure/use. No language criteria were applied.

DATA EXTRACTION

Study selection, data abstraction, and quality assessment were performed by two independent reviewers. Random effects models were used to obtain summary odds ratios (ORs) and 95% confidence intervals (CIs).

DATA SYNTHESIS

Of the 17 identified studies, 15 were included in the meta-analysis. Exposures during pregnancy to unspecified residential pesticides (summary OR = 1.54; 95% CI, 1.13-2.11; I2 = 66%), insecticides (OR = 2.05; 95% CI, 1.80-2.32; I2 = 0%), and herbicides (OR = 1.61; 95% CI, 1.20-2.16; I2 = 0%) were positively associated with childhood leukemia. Exposures during childhood to unspecified residential pesticides (OR = 1.38; 95% CI, 1.12-1.70; I2 = 4%) and insecticides (OR = 1.61; 95% CI, 1.33-1.95; I2 = 0%) were also positively associated with childhood leukemia, but there was no association with herbicides.

CONCLUSIONS

Positive associations were observed between childhood leukemia and residential pesticide exposures. Further work is needed to confirm previous findings based on self-report, to examine potential exposure-response relationships, and to assess specific pesticides and toxicologically related subgroups of pesticides in more detail.

Covert preleukemia driven by MLL gene fusion

Acute leukemia is considered to be a two- or multiple-step process. Although there is a considerable knowledge regarding the character of the "first hit," the nature of the "second hit" remains unanswered in most of the cases including leukemias with MLL gene rearrangement. We demonstrate here a striking sequence of events, which include a covert, protracted preleukemic phase characterized by a dominant MLL/FOXO3A clone with intact myeloid differentiation and the subsequent acquisition of a secondary genetic abnormality, leading to overt lymphoblastic leukemia. Backtracking of the secondary acute lymphoblastic leukemia (sALL) with the MLL rearrangement showed no blasts in the bone marrow (BM) during the protracted preleukemic phase. However, at the same time (more than 1 year before the sALL diagnosis) the MLL/FOXO3A was present in up to 90% of BM cells including myeloid lineage, suggesting that the fusion arose in a multipotent progenitor. To identify potential "second hit" precipitating sALL we compared DNA in preleukemic versus fully leukemic samples. The analysis revealed a 10 Mb gain on 19q13.32 in the sALL, absent in the preleukemic specimen. These data provide insight into the dynamics of leukemogenesis in secondary leukemia with MLL rearrangement.

Childhood cancer among twins and higher order multiples

Although several studies have found no change or a decreased risk of childhood cancer in twins, few have controlled for potential confounders such as birth weight. We examined the association of birth plurality and childhood cancer in pooled data from five U.S. states (California, Minnesota, New York, Texas, and Washington) using linked birth-cancer registry data. The data, excluding children with Down syndrome or who died before 28 days of life, included 17,672 cases diagnosed from 1980 to 2004 at ages 28 days to 14 years and 57,966 controls with all cases and controls born from 1970 to 2004. Analyses were restricted to children weighing Collapse

Key Words

• childhood cancer
• twins
• multiple birth
• case-control
• pooled data

Collapse

MESH Headings

• Adolescent
• Case-Control Studies
• Child
• Child, Preschool
• Diseases in Twins/epidemiology
• Female
• Humans
• Infant
• Infant, Newborn
• Kidney Neoplasms/epidemiology
• Logistic Models
• Male
• Multiple Birth Offspring
• Neoplasms/epidemiology
• Registries
• Retinal Neoplasms/epidemiology
• Retinoblastoma/epidemiology
• Risk Factors
• United States/epidemiology
• Wilms Tumor/epidemiology

Collapse

Grants

• R01CA717450 NCI NIH HHS
• R01 CA092670-04 NCI NIH HHS
• U58 DP000783 NCCDPHP CDC HHS
• R01CA92670 NCI NIH HHS
• R01 CA092670 NCI NIH HHS
• T32 CA099936-04 NCI NIH HHS
• N01 CN005230 NCI NIH HHS
• R01 CA071745-04 NCI NIH HHS
• T32 CA099936 NCI NIH HHS

Collapse

Affiliation(s)

Susan E. Puumala Division of Epidemiology/Clinical Research, Department of Pediatrics, University of Minnesota
Susan E. Carozza Department of Epidemiology and Biostatistics, Texas A&M University
Eric J. Chow Fred Hutchinson Cancer Research Center
Erin E. Fox Cancer Epidemiology and Surveillance Branch, Texas Department of State Health Services
Scott Horel Department of Epidemiology and Biostatistics, Texas A&M University
Kimberly J. Johnson Division of Epidemiology/Clinical Research, Department of Pediatrics, University of Minnesota
Colleen McLaughlin New York State Cancer Registry, New York Department of Health
Beth A. Mueller Fred Hutchinson Cancer Research Center
Peggy Reynolds Northern California Cancer Center
Julie Von Behren Northern California Cancer Center
Logan G. Spector Division of Epidemiology/Clinical Research, Department of Pediatrics, University of Minnesota Masonic Cancer Center, University of Minnesota

Collapse

Chromosome 12p deletions in TEL-AML1 childhood acute lymphoblastic leukemia are associated with retrotransposon elements and occur postnatally

TEL-AML1 (ETV6-RUNX1) is the most common translocation in the childhood leukemias, and is a prenatal mutation in most children. This translocation has been detected at a high rate among newborns ( approximately 1%); therefore, the rate-limiting event for leukemia seems to be secondary mutations. One such frequent mutation in this subtype is partial deletion of chromosome 12p, trans from the translocation. Nine del(12p) breakpoints within six leukemia cases were sequenced to explore the etiology of this genetic event, and most involved cryptic sterile translocations. Twelve of 18 del(12p) parent sequences involved in these breakpoints were located in repeat regions (8 of these in long interspersed nuclear elements). This stands in contrast with TEL-AML1, in which only 21 of 110 previously assessed breakpoints (19%) occur in DNA repeats (P=0.0001). An exploratory assessment of archived neonatal blood cards revealed significantly more long interspersed nuclear element CpG methylations in individuals at birth who were later diagnosed with TEL-AML1 leukemia, compared with individuals who did not contract leukemia (P=0.01). Nontemplate nucleotides were also more frequent in del(12p) than in TEL-AML1 junctions (P=0.004), suggesting formation by terminal deoxynucleotidyl transferase. Assessment of six archived neonatal blood cards indicated that no del(12p) rearrangements backtracked to birth, although two of these patients were previously positive for TEL-AML1 using the same assay with comparable sensitivity. These data are compatible with a two-stage natural history: TEL-AML1 occurs prenatally, and del(12p) occurs postnatally in more mature cells with a structure that suggests the involvement of retrotransposon instability.

p53 signaling in response to increased DNA damage sensitizes AML1-ETO cells to stress-induced death

Chromosomal translocation (8;21) is present in 10% to 15% of patients with acute myeloid leukemia. Expression of the AML1-ETO (AE) fusion protein alone is not sufficient to induce leukemia, but the nature of the additional genetic alterations is unknown. It is unclear whether AE facilitates acquisition of these cooperating events. We show that AE down-regulates genes involved in multiple DNA repair pathways, potentially through a mechanism involving direct binding at promoter elements, and increases the mutation frequency in vivo. AE cells display increased DNA damage in vitro and have an activated p53 pathway. This results in increased basal apoptosis and enhanced sensitivity to DNA damaging agents. Intriguingly, microarray data indicate that t(8;21) patient samples exhibit decreased expression of DNA repair genes and increased expression of p53 response genes compared with other acute myeloid leukemia (AML) patient samples. Inhibition of the p53 pathway by RNAi increases the resistance of AE cells to DNA damage. We thus speculate that AML1-ETO may facilitate accumulation of genetic alterations by suppressing endogenous DNA repair. It is possible that the superior outcome of t(8;21) patients is partly due to an activated p53 pathway, and that loss of the p53 response pathway is associated with disease progression.