Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Inherited Bone Marrow Failure Syndrome

1.1 Diamond-Blackfan Anemia

1.1.1 Introduction

Diamond-Blackfan anemia (DBA) is a congenital pure red cell aplasia characterized by normochromic macrocytic anemia and reticulocytopenia in the peripheral blood and the loss of erythroblasts in the bone marrow. It generally presents in the first year of life. Approximately 40% of DBA patients have variable malformations [1,2,3]. In addition, the patients have a predisposition for malignancies [4]. DBA was first reported by Dr. Josephs in 1936 and refined as a distinct clinical entity by Drs. Diamond and Blackfan in 1938 [5, 6].

DBA has been associated with heterozygous mutations or large deletions in ribosomal protein (RP) genes in more than 50% of patients [7,8,9,10,11,12,13,14,15,16,17], suggesting that DBA is caused by ribosome dysfunction [18]. Corticosteroids and transfusion are the mainstays of treatment. Approximately 80% of patients respond to an initial course of steroids; however, only 60–70% of patients ultimately become transfusion-independent [3]. Bone marrow transplantation is the only curative treatment, but it requires an HLA-matched sibling and is primarily reserved for patients with severe complications [19].

1.1.2 Epidemiology

Ten to 20% of DBA families have more than one affected individual, and the majority of these cases appear to be dominantly inherited. The remaining cases are sporadic or familial with other inheritance patterns. The incidence is estimated to be 5–10 per million live births without ethnic predilection, with both sexes equally affected [2, 19,20,21,22]. In the Registration of the Japanese Society of Pediatric Hematology and Oncology, 175 patients with DBA, including patients diagnosed with idiopathic pure red cell anemia, were reported from 1988 to 2011 [23].

1.1.3 Pathogenesis

Following the observation that a DBA patient had an X;19 chromosomal translocation, a major DBA locus was mapped to chromosomal location 19q13, and the breakpoint was shown to occur in the RPS19 gene, which encodes one of 80 ribosomal proteins [9]. Subsequent large-scale studies established that RPS19 is mutated in approximately 25% of DBA patients [24]. To date, 15 RP genes including RPS7, RPS10, RPS15A, RPS17, RPS19, RPS24, RPS26, RPS27, RPS28, RPS29, RPL5, RPL11, RPL26, RPL27, and RPL35A have been reported to be responsible for DBA [7,8,9,10,11,12,13,14,15,16,17, 136]. A sizable fraction of DBA patients have a large deletion of RP genes that are not detected by sequencing [25, 26]. Although 10–20% of patients have inherited DBA, most patients are sporadic and have de novo mutations. DBA has been associated with heterozygous mutations or large deletions in RP genes in up to 60% of patients, except for rare germline GATA1 mutation reported in three X-linked DBA families [137].

The ribosome is an intracellular organelle and is the site of protein synthesis and translation of messenger RNA into continuous chains of amino acids. The mammalian ribosome (80S) consists of a large subunit (60S) and a small subunit (40S), and each subunit is composed of ribosomal RNAs and RPs. RPs of the small and large subunits are called RPS and RPL, respectively. RPS19, RPS24, RPS10, and RPS26 play important roles in maturation of the 18S rRNA and assembly of the 40S ribosomal subunit [8, 27,28,29,30]. RPL35A, RPL5, and RPL11 play important roles in the maturation of the 28S and 5.8S rRNAs and assembly of the 60S ribosomal subunit [10, 12]. Therefore, deficiency in RPs leads to a relative lack of the 40S or 60S ribosomal subunits and a decline in translation initiation. The fact that all of the causative genes for DBA are ribosomal proteins, except for GATA1, suggests that insufficiency in ribosomal function may be the underlying cause of red cell aplasia in patients with DBA. Although the mechanism whereby mutations in the ribosomal protein genes cause specific defects in red cell maturation is not fully understood, many lines of evidence indicate that p53 activation caused by ribosomal dysfunction may be central to DBA pathogenesis (Fig. 7.1) [31].

Fig. 7.1
figure 1

Model of p53 stabilization in response to impaired ribosome biogenesis in DBA. Normal erythroblasts produce large numbers of ribosomes for protein synthesis. Levels of p53 remain low via a feedback loop whereby MDM2, a transcriptional p53 target, ubiquitinates p53 to promote its degradation by proteasomes. However, haploinsufficiency for specific RPs causes accumulation of other RPs, which bind to MDM2, thereby inhibiting its ability to promote p53 degradation. Consequently, p53 accumulates and triggers cell cycle arrest and apoptosis

Full size image

1.1.4 Clinical Management

1.1.4.1 Clinical Presentation
1.1.4.1.1 Anemia

DBA is a congenital pure red cell aplasia characterized by normochromic macrocytic anemia and reticulocytopenia in the peripheral blood and the loss of erythroblast from bone marrow. Approximately 90% of affected individuals present in infancy or early childhood, although a “nonclassical” mild phenotype may not be diagnosed until later in life [2, 3].

1.1.4.1.2 Congenital Anomalies

Clinical manifestations of DBA are variable. In about 40% of patients, DBA is associated with physical anomalies and growth retardation, but in some patients, no congenital anomalies are found [3]. Table 7.1 summarizes the range of congenital abnormalities found in DBA patients.

Table 7.1 Congenital abnormalities observed in Diamond-Blackfan anemia (DBA)
Full size table
1.1.4.1.3 Cancer Predisposition

Twenty-nine cases were reported among more than 700 patients in the literature [3]. A recent prospective study showed that among 608 patients, 15 solid tumors, two acute myeloid leukemias (AML), and two cases of myelodysplastic syndrome (MDS) were diagnosed at a median age of 41 years in patients who had not received a bone marrow transplant. The rate of development of solid tumors or leukemia was 5.4-fold higher in DBA patients than expected in a demographically matched comparison with the general population. Furthermore, specific tumors had significantly elevated incidence ratios for MDS, AML, adenocarcinoma of the colon, osteogenic sarcoma, and female genital cancer [138].

1.1.4.2 Diagnostic Procedures
1.1.4.2.1 Diagnostic Criteria

The diagnostic criteria for “classical” DBA include macrocytic (or normocytic) anemia with no other significant cytopenia presenting prior to the first birthday, reticulocytopenia, and normal marrow cellularity with a paucity of erythroid precursors, and it is accompanied by congenital anomalies in some cases. However, diagnosis of DBA is often difficult due to the incompleteness of the phenotypes and the wide variability in clinical expression. Even mutations in individual RP genes lead to widely variable phenotypic expression; family members with the same mutation in an RP gene can present with clinical differences. For example, RPS19 mutations are found in some first-degree relatives presenting only with isolated high erythrocyte adenosine deaminase activity and/or macrocytosis [24]. Therefore, it is very difficult to make a diagnosis on the basis of clinical phenotype alone. Molecular diagnosis enables the detection of carriers and the avoidance of hematopoietic stem cell transplantation from sibling donors with these mutations. However, about 40% have no known pathogenic mutations. The diagnostic and supporting criteria for diagnosis of DBA, including mild type, are described in Table 7.2. We have modified the criteria from the international clinical consensus conference [3] based on our results for the novel DBA biomarker reduced glutathione (GSH) [32].

Table 7.2 Diagnostic criteria for patients with Diamond-Blackfan anemia (DBA)
Full size table
1.1.4.3 Differential Diagnosis

In children, transient erythrocytopenia in childhood (TEC) should be a major consideration [3]. TEC usually presents in children older than 1 year of age following viral infection. Most cases resolve spontaneously within 1–2 months. Patients with TEC present with normocytic anemia, and, in contrast to those with DBA, their HbF levels and erythrocyte adenosine deaminase (eADA) activities are normal. Congenital bone marrow failure syndromes accompanied by physical anomalies include dyskeratosis congenita, Shwachman-Diamond syndrome, congenital amegakaryocytic thrombocytopenia, and Pearson syndrome. Although all of these diseases are extremely rare, it is possible to discriminate them from specific clinical findings. Recently, because the causative genes for all of these diseases have been identified, the molecular etiology may be clarified in the near future, and genetic diagnosis may be possible in many patients.

1.1.4.4 Treatment
1.1.4.4.1 Transfusion Therapy

Patients who do not respond to steroids may require chronic transfusion therapy every ~4–8 weeks. Although maintaining hemoglobin levels above 8 g/dL is recommended, a disadvantage of chronic transfusion therapy is hemosiderosis by iron overload. To avoid liver dysfunction, diabetes, and myocardiopathy due to iron deposition, chelation therapy with deferasirox or deferoxamine is recommended. However, methods for oral iron chelation therapy with deferasirox in neonates have not been established.

1.1.4.4.2 Drug Therapy

Approximately 80% of patients respond to steroids. The recommended starting dose is 2 mg/kg/day of prednisone at initial treatment. In over 20% of DBA patients, steroids eventually are stopped completely [3]. It should be mentioned that important steroid side effects include growth retardation, osteoporosis, obesity, hypertension, diabetes mellitus, cataract, and glaucoma. The therapy is not generally recommended in babies under 6 months of age. Alternative treatments, including cyclosporine, metoclopramide, EPO and a combination of prednisolone and cyclosporine, have been mentioned, but evaluations have not been provided.

1.1.4.4.3 Hematopoietic Stem Cell Transplantation

Chronic transfusion dependence with refractoriness to steroids is considered an acceptable indication for hematopoietic stem cell transplantation (HSCT). The outcomes of HSCT for DBA in Japan are better than those reported in other countries. To date, 19 patients have received transplants of allogeneic hematopoietic stem cells (HSCs). Thirteen patients underwent BMT (six cases with HLA-matched siblings and seven cases with unrelated donors), and all have disease-free survival [33]. Five patients underwent cord blood transplantation (CBT), and two of these with transplants from siblings have survived. However, in three patients treated with unrelated donor cord blood, two were not successfully engrafted, and the third died of lymphoproliferative disease, despite confirmed engraftment. Consequently, bone marrow should be selected as a source of HSCs for transplantation at present. Better therapeutic outcomes have been observed with a conditioning therapy combining busulfan (oral administration, 16 mg/kg or 560 mg/m2) and cyclophosphamide (120–200 mg/kg). Although the reported number of patients is few, a conditioning therapy using a half dose of busulfan leads to good outcomes when using unrelated HSCs transplantation. When busulfan was omitted from pretreatment, the number of cases with engraftment failure was modestly higher. There is not sufficient evidence supporting the value of bone marrow nondestructive pretreatment therapy for HSCT in DBA [34].

1.1.5 Future Challenge

Recent studies have provided fascinating insights into the pathogenesis of DBA. However, several important questions remain to be answered. For example, it is not clear how haploinsufficiency of ribosomal proteins leads to a phenotype of defective erythropoiesis and why mutations in some ribosomal protein genes such as RPL5 have more profound impact on fetal development than mutations in other RP genes. Although corticosteroids remain the mainstay of treatment of DBA more than half a century after the original report of their efficacy, their mechanism of action is still unknown. DBA will necessarily lead us toward a further understanding of bone marrow failure syndrome and potentially to novel ways of treatment.

Recently, using DBA models in zebra fish and mice, it has been reported that administration of l-leucine, an essential amino acid, provided relief from anemia [35, 36]. Clinical trials to evaluate the therapeutic effects of l-leucine are underway in the United States. About half of DBA patients have notbeen identifiedcausative mutations. Comprehensive genetic analysis using next-generation sequencing in the remaining DBA patients would be valuable in the identification of other causative gene mutations.

1.2 Fanconi’s Anemia

1.2.1 Introduction

Fanconi’s anemia (FA) is characterized by progressive bone marrow failure, congenital abnormalities, and a pronounced susceptibility for malignancy, including myelodysplastic syndrome (MDS), leukemia, and solid tumors such as squamous cell carcinoma. FA was first reported by Dr. Fanconi in 1927 [37]. Lymphocytes from FA patients exhibit chromosomal instability [38], a characteristic that was increased by mitomycin C treatment [39]. Mutations in 19 different genes (FA genes) have been identified, the products of which cooperate with other gene products in a common pathway regulating DNA repair [40]. Stem cell transplantation is the only curative treatment for FA. In recent years, HSCT from alternative donors has become more popular and has been successfully employed [41, 42].

1.2.2 Epidemiology

About 5–10 cases are diagnosed with FA, and the incidence in Japan is estimated to be 5 in 1,000,000 newborns. The carrier frequency for FA in Japan may be 1 in 200–300 people [23].

1.2.3 Pathogenesis

FA is a multigenic disorder with 19 genes currently identified (A, B, C, D1, D2, E, F, G, I, J, L, M, N, O, P, Q, R, S, and T) [40]. Among them, FANCT was the first FA-causative gene discovered by a Japanese group [43]. FANCD1, FANCJ, and FANCN were identified as BRCA2, BRIP1, and PALB2, respectively. These three FA genes are associated with breast and ovarian cancer in heterozygotes. With the exception of the X-linked FANCB gene, the remaining 18 FA genes are autosomal recessive. The most common gene in FA is FANCA (60–70%), and more than 80% of FA patients have mutations inFANCA or FANCG in Japan. These FA proteins cooperate with other gene products, in a common pathway (FA-BRCA pathway) that regulates DNA repair (Fig. 7.2). The FA pathway plays an important role in the proliferation of hematopoietic stem cells and tumor suppression. The FANCD1 and FANCN subtypes initiate a malignant tumor in early childhood, and their prognosis is very poor. In contrast, clonal expansion of hematopoietic cells with reversion mosaicism can restore normal hematopoiesis [44].

Fig. 7.2
figure 2

Schematic of the Fanconi’s anemia DNA repair pathway. After DNA damage, the FA core complex is activated, which then functions as an E3 ubiquitin ligase and monoubiquitinates FANCD2 and FANCI. The monoubiquitinated FANCD2/FANCI complex (ID complex) is then targeted to chromatin where it forms a complex with additional FA proteins and other DNA repair proteins

Full size image

Recent studies in mice have suggested that the FA proteins might counteract aldehyde-induced genotoxicity in hematopoietic stem cells [139, 140]. Acetaldehyde dehydrogenase 2 (ALDH2) primarily catalyzes the conversion of acetaldehyde to acetic acid; it also converts a range of endogenous toxic aldehydes. ALDH2 deficiency resulting from a Glu504Lys substitution (rs671, hereinafter referred to as the A allele) is highly prevalent in East Asian populations, including Japanese. The A allele (Lys504) acts as a dominant negative, since the variant form can suppress the activity of the Glu504 form (G allele) in GA heterozygotes by the formation of heterotetramers [45]. Individuals with the A variant experience flushing when drinking alcohol and have an elevated risk of esophageal cancer with habitual drinking. Compared to individuals with GG homozygotes, enzymatic activity is reduced by 60–80% or nearly abolished (~4%) in GA or AA individuals, respectively [46]. Recently, Hira et al. showed that the presence of the A allele strongly accelerated progression of BMF in Japanese FA patients. Although ALDH2 status did not influence the number of developmental abnormalities, malformations at some specific anatomic locations were observed more frequently in ALDH2-deficient patients [47]. These results indicate that the level of ALDH2 activity impacts the pathogenesis of FA.

1.2.4 Clinical Management

1.2.4.1 Clinical Presentation
1.2.4.1.1 Pancytopenia

Pancytopenia was the most common presentation, particularly when the red cell mean cell volume (MCV) and fetal hemoglobin (HbF) were elevated for age. Bone marrow biopsy examination most often showed hypocellularity for age, due to decreased numbers of hematopoietic precursors with normal morphology. The median age at diagnosis is 6.5 years, ranging from birth to adult. The diagnostic age in the reported cases was similar in both sexes [48].

1.2.4.1.2 Congenital Anomalies

Clinical manifestations of FA are variable. In about 60% of patients, FA is associated with physical anomalies and growth retardation, but in some patients, no congenital anomalies are found [48]. The most common characteristics at birth are skin hyperpigmentation, café au lait spots, short stature, abnormal thumbs, and radii [49].

1.2.4.1.3 Cancer Predisposition

Patients with FA are at a particularly high risk of developing MDS/AML and specific solid tumors at unusually young ages, including head, neck, esophageal, and gynecological squamous cell carcinoma. In the literature, about 15–20% of FA patients had hematologic neoplasms, and 5–10% had nonhematologic neoplasms [50, 51]. Of these neoplasms in Japanese FA patients, 33% were hematologic and 10.4% were nonhematologic [49]. Solid tumors were observed both after and before HSCT, and they developed between 1 and 39 years of age.

1.2.4.2 Diagnostic Procedures
1.2.4.2.1 Diagnostic Evaluation

FA is an autosomal recessive disorder associated with a high frequency of bone marrow failure, leukemia, and solid tumors. FA is a complex disease that can affect many systems of the body. Initially, patients are recognized when they have the combination of aplastic anemia and birth defects. However, current diagnostic criteria are more extensive and rely on demonstration of chromosomal aberrations in cells cultured with DNA cross-linking agents. The suspected diagnosis is usually confirmed by demonstration of chromosomal aberrations in blood lymphocytes cultured with a DNA cross-linking agent such as diepoxybutane (DEB) or mitomycin C (MMC) [48]. Following DNA damage, the complex of upstream FA gene products (A, B, C, E, F, G, I, L, M) leads to ubiquitination of the product of FANCD2, detected with a Western blot with a D2-specific antibody [52].

However, some FA patients have hematopoietic somatic mosaicism. These cases have corrected one mutated allele in a bone marrow stem cell, leading to an acquired heterozygosity in the blood cells. If diagnostic test results of blood are not conclusive and there is a high probability of FA, skin or bone marrow fibroblast cultures are required to demonstrate sensitivity to DNA cross-linking agents. Finally, FA mutation analysis determines the initial complementation group. Target next-generation sequencing can be performed to determine the relevant mutations.

1.2.4.3 Differential Diagnosis

Chromosome instability syndrome is a group of inherited conditions associated with chromosomal instability and breakage, and the following chromosome instability syndromes are known: xeroderma pigmentosum, ataxia-telangiectasia, Bloom syndrome, and Nijmegen syndrome.

1.2.4.4 Treatment
1.2.4.4.1 Transfusion Therapy

FA patients can be transfused to maintain minimal trough hemoglobin of 6 g/dL and minimal trough platelet of 5000/μL.

1.2.4.4.2 Drug Therapy

Patients who choose not to pursue transplant for severe cytopenias may benefit from treatment with androgens. Androgens can improve cytopenias in all three lineages, erythroid, myeloid, and platelets, but the effects are typically most pronounced for the erythroid lineage [48]. The effects of androgens were recognized in half of FA patients, but its effect seems to have lasted only temporarily [53]. Androgen may exacerbate the risk of liver tumors, and prior exposure to androgens is the one of the risk factors affecting survival after unrelated transplant [54]. Metenolone acetate is used in Japan. A clinical trial of danazol, which has less virilizing side effects, is currently under way. The use of steroid hormones should be avoided.

1.2.4.4.3 Hematopoietic Stem Cell Transplantation

Currently, the only cure for the hematological abnormalities of FA remains HSCT. The optimal timing of transplantation is challenging because the best outcomes are achieved prior to the development of complications such as infections from chronic severe neutropenia, high transfusion burden to treat anemia/thrombocytopenia, and the development of MDS or AML. Patients with FA are exquisitely sensitive to genotoxic agents such as cyclophosphamide, busulfan, and ionizing radiation. FA patients are also susceptible to the damaging inflammatory side effects of graft-versus-host disease. The use of high-dose cyclophosphamide (CY) and radiation in preparative regimens for FA patients often resulted in excessive organ toxicity and death. Therefore, efforts have focused on reducing the doses required for transplant preparative regimens, choosing nongenotoxic regimens to prevent graft-versus-host disease and using alternative conditioning regimens [48]. The use of low-dose CY (20–40 mg/kg) combined with 4–6 Gy of thoracoabdominal irradiation (TAI) or total body irradiation (TBI) reduced procedure toxicity and improved the outcome for FA patients transplanted from HLA-matched sibling donors [55]. Non-radiation regimens have been increasingly used for FA patients to reduce the secondary malignancies associated with radiation [56, 57].

A significant proportion of FA patients undergoing HSCT can now be dramatically cured, even in the absence of an HLA-identical sibling, especially if the conditioning regimen includes fludarabine [42, 58]. In Japanese FA patients transplanted using a fludarabine regimen, the 3-year estimate of overall survival (OS) for eight patients was 100% when the donor was an HLA-matched sibling [59], and 26 patients out of 27 patients transplanted from an alternative donor are alive [42]. It is essential to test all potential sibling donors for FA regardless of clinical findings since the phenotypic variation even within a given family is broad.

HSCT does not correct the nonhematological manifestations of Fanconi’s anemia. Solid tumor risk, particularly head and neck squamous cell carcinoma, continues to increase after transplant, particularly in FA patients experiencing severe graft-versus-host disease [60]. A retrospective study comparing solid tumor risks in transplanted versus non-transplanted FA patients reported a 4.4-fold higher age-specific hazard rate of squamous cell carcinoma in patients treated by transplantation. The tumors appeared at an earlier age in the transplanted cohort [61]. Data on solid tumor risks in patients transplanted with the newer regimens are as yet limited.

1.2.5 Future Challenges

Patients with FA usually are diagnosed in childhood, and their registrations have been managed in the Japanese Society of Pediatric Hematology/Oncology in Japan. Although the adult FA population is small, it is important to understand transplanted/non-transplanted adult FA patients, who have or have not developed bone marrow failure, hematological malignancy, and solid tumors. Fludarabine-based preconditioning regimen can be used satisfactorily in alternative HSCT for FA. However, long-term observation of secondary cancers and other late effects will be required to determine the therapeutic utility of this approach. A recent report showed endogenously produced aldehydes impact the severity of FA, suggesting the possibility of a novel therapeutic approach [47]. For example, Alda-1 can stimulate the enzymatic activity of both the normal and variant ALDH2 [62]. ALDH2 agonists such as Alda-1 might be a novel protective drug against BMF in FA patients.

2 TAM

2.1 Introduction

Trisomy 21, the genetic hallmark of Down syndrome (DS), is the most frequent human chromosomal abnormality. In Japan, the incidence in the general population is 1 in 600–800 live births. Neonates with DS are at a high risk of developing a hematologic disorder referred to as transient abnormal myelopoiesis (TAM), which is characterized by the rapid growth of abnormal blast cells expressing megakaryocytic markers [63]. It has been estimated that 5–10% of infants with DS exhibit the disorder, and in most cases, it resolves spontaneously within 3 months. However, approximately 20% of severe cases are subject to fatal complications, and 20–30% of patients who escape from early death develop AMKL referred to as myeloid leukemia of DS (ML-DS) within the first 4 years of life [63,64,65,66].

Human tumors have been shown to progress by the accumulation of genetic abnormalities [67]. The malignant progression from TAM to AMLK offers a unique model to study the stepwise developments of cancer pathogenesis. The blasts in TAM are indistinguishable from true leukemic cells in both morphology and expressed surface markers. Blast cells in most patients with TAM as well as those with ML-DS have mutations in the gene encoding the transcription factor GATA1 [68,69,70,71,72,73,74], which is essential for normal development of erythroid and megakaryocytic cells [75,76,77]. Recent reports showed that TAM is caused by a single GATA1 mutation and constitutive trisomy 21, and subsequent AMKL evolves from a preexisting TAM clone through the acquisition of additional mutations in genes coding for cohesin components, CCCTC-binding factor (CTCF), epigenetic regulators, as well as tyrosine kinases/Ras pathway genes (Fig. 7.3) [78]. In this chapter, advances in the pathobiology and clinical management of TAM will be described.

Fig. 7.3
figure 3

Natural history and pathogenesis of TAM and ML-DS. Trisomy 21, the genetic hallmark of Down syndrome (DS), is the most frequent human chromosomal abnormality. Five to 10% of neonates with DS develop transient abnormal myelopoiesis (TAM), which is characterized by rapid growth of abnormal megakaryoblastic cells. Almost all cases of TAM have mutations in GATA1. In most cases, it resolves spontaneously within 3 months. However, early death occurs in about 20% of the cases. Furthermore, about 20% of patients develop myeloid leukemia of DS (ML-DS) within 4 years of life after acquisition of additional mutations in the genes encoding multiple cohesion components, CTCF, and epigenetic regulations such as EZH2, as well as genes encoding common signaling pathways, such as JAK family kinases and Ras pathway genes. FL-HSCs: Hematopoietic stem cells in fetal livers, MEPs: Megakaryocyte-erythroid progenitors

Full size image

2.2 Epidemiology

The actual incidence of TAM has been difficult to define. Estimates of the frequency of TAM vary from 3.8 to 30% of neonates with DS, depending on the diagnostic and eligibility criteria used in each study and the sensitivity of the methodologies used for screening [66, 79,80,81]. Zipursky et al. screened a small series of DS neonates by microscopic examination of blood films and found that 10% had blasts in the peripheral blood [66]. It is possible that some cases with relatively few blasts would have been missed. Pine et al. screened Guthrie cards from 590 DS infants for GATA1 mutations by single-strand conformation polymorphisms (SSCP) and direct sequencing of PCR products and showed that 3.8% of the infants harbored GATA1 mutations [80]. A recent prospective study of 200 DS neonates showed an 8.5% frequency of GATA1 mutations, using a combination of standard Sanger sequencing with direct high-performance liquid chromatography (Ss/HPLC) [81]. All of these patients had more that 10% blasts in the peripheral blood. Furthermore, low abundant minor GATA1-mutant clones were detected in 20.4% of DS neonates without detectable GATA1 mutant clones with Ss/HPLC, by more sensitive targeted next-generation sequencing that can detect as few as 0.3% TAM blasts. They referred to these patients as “silent TAM.” This study indicates approximately 30% of DS neonates have GATA1 mutations [81].

2.3 Pathogenesis

2.3.1 GATA1 Mutations in TAM and ML-DS

Cell differentiation is controlled in part by cell lineage-restricted transcription factors. The transcription factor GATA1 is expressed within the hematopoietic hierarchy in erythroid, megakaryocytic, eosinophilic, and mast cell lineages [82,83,84,85], as well as in Sertoli cells of the testis [86, 87]. Gene targeting experiments have revealed that GATA1 is required for the terminal differentiation of definitive erythroid and megakaryocytic cells [88,89,90].

GATA1 mRNAs are abundantly expressed in blast cells from TAM patients, as well as ML-DS patients [91]. Almost all patients with TAM and ML-DS have GATA1 mutations [68,69,70,71,72,73,74]. The mutations lead to a loss of expression of the full-length GATA1 protein and the expression of a truncated GATA1 protein (GATA1s) that lacks the N-terminal transactivation domain. GATA1s, which is translated from the second methionine at codon 84, retains intact zinc finger regions, binds appropriately to the GATA consensus sequence, and interacts normally with its essential cofactor FOG-1.

Several cell culture-based rescue experiments have reported that the N-terminal activation domain appears to be dispensable during erythroid and megakaryocytic cell differentiation [92,93,94,95]. Furthermore, distinct regions in the N-terminus of GATA1, which regulates the proliferation of immature megakaryocytic progenitors, are required for terminal megakaryocyte differentiation and controlling the growth of immature precursors [96, 97]. GATA1, but not GATA1s, interacts physically with the transcription factor E2F, which can trigger reentry into the cell cycle. That observation suggests that the failure of GATA1s to repress E2F activity in cellular transformation of fetal progenitors may be caused by this failure in direct or indirect interaction [98]. Supporting this hypothesis, Toki et al. recently discovered novel GATA1 mutants in TAM that lack just the E2F interaction domain in the N-terminus of GATA1 [99].

On the other hand, mice harboring a heterozygous GATA1 knockdown allele frequently develop erythroblastic leukemia [100]. These observations indicate that the expression level of GATA1 is crucial for the proper development of erythroid and megakaryocytic cells and that compromised GATA1 expression is a causal factor in leukemia [101]. Recently, Kanezaki et al. showed that the spectrum of transcripts derived from the mutant genes affects GATA1s protein expression. Mutations resulting in low GATA1s levels are significantly associated with a risk of progression to ML-DS [102]. However, a subsequent report contended that the GATA1 mutational spectrum did not differ between TAM and AMKL and that the type of GATA1 sequence mutation is not a reliable tool and is not prognostic of which patients with TAM are likely to develop ML-DS [103]. To resolve this problem, a prospective study with a large series of TAM patients is necessary.

2.3.2 The Target Cells of TAM and ML-DS

TAM is a disorder found mainly in patients with DS during the newborn period and is sometimes associated with liver fibrosis, which usually takes a lethal course. Based on pathological observation of TAM cases with fatal liver fibrosis, Miyauchi et al. have proposed a hypothesis that TAM blasts originate from fetal liver hematopoietic progenitors [104]. Analyses of GATA1 mutations provide direct evidence for this hypothesis. The presence of identical GATA1 mutations in the blasts of both TAM and ML-DS in an identical twin suggested that GATA1 mutations occur early during prenatal hematopoiesis [105]. GATA1 mutations were detected in hematopoietic tissues from DS fetuses that had no pathological evidence of leukemia [106] and in Guthrie newborn screening cards at birth from DS infants who later developed ML-DS [107].

In addition, mice expressing GATA1s presented with hyperproliferation of yolk sac and fetal liver progenitors [108]. Consistent with this finding, GATA1-deficient mice rescued with transgenic expression of GATA1s exhibited hyper-megakaryopoiesis during a limited embryonic and postnatal period, resembling the phenotype in human TAM cases [109]. These data from mouse experiments also indicated the possibility that fetal megakaryocytic progenitors are the target cells of TAM and ML-DS.

2.3.3 The Essential Roles of Trisomy 21 in Leukemogenesis of DS

For the following reasons, trisomy 21 is thought to be the first genetic event in the development of ML-DS. First, TAM occurs almost exclusively in patients with DS [66]. Second, when TAM occurs in patients with mosaic trisomy 21, the TAM clones involve only the cells with trisomy 21 [110]. Third, an inherited mutation in humans leading to production of only GATA1s is associated with impaired erythropoiesis and granulopoiesis, but does not promote leukemia in the absence of trisomy 21 [111]. Fourth, trisomy 21 itself is associated with enhanced expansion of human fetal erythroid and megakaryocytic precursors, independent of GATA1 mutations [112, 113].

The mechanisms by which trisomy 21 predisposes individuals for the development of acute leukemia are thought to involve increased expression of a gene, or genes, on chromosome 21 that stimulate abnormal proliferation of hematopoietic stem cells in infancy [114]. A recent genotype-phenotype study in 30 DS individuals with rare segmental trisomy 21 led to the identification of a critical region of 8.35 Mb that is likely to contribute to increased risk of developing TAM and AMKL [115]. This region includes the RUNX1, DYRK1A, ERG, and ETS genes.

2.3.3.1 ERG and ETS2

Both ERG and ETS2 transcription factors belong to the ETS family. ERG is expressed in TAM and AMKL in DS, and its forced expression in erythroleukemia cells causes a phenotypic shift toward the megakaryocytic lineage. ERG binds the hematopoietic enhancer of SCL/TAL1 that regulates its expression in hematopoietic stem cells [116]. SCL1/TAL1 overexpression is known to force progenitor cells toward the megakaryocytic lineage. Ectopic expression of ERG in fetal hematopoietic progenitors promotes megakaryopoiesis, and ERG alone acts as a potent oncogene in vivo leading to rapid onset of leukemia in mice. Furthermore, ERG also strongly cooperates with GATA1s to immortalize megakaryocytic progenitors [117]. Overexpression of ETS2 and ERG immortalizes GATA1 knockdown and GATA1s knockin, but not wild-type fetal liver progenitors. Immortalization is accompanied by activation of the JAK-STAT pathway, commonly seen in megakaryocytic malignancies [118]. These results indicate that there is a specific synergy between loss of full-length GATA1 and overexpression of ETS family members in the control of self-renewal of megakaryocyte progenitors.

2.3.3.2 RUNX1

Both GATA1 and GATA1s are associated with the RUNX1 transcription factor [119], which cooperates with GATA1 during megakaryocyte differentiation [120]. Analysis using a transchromosomic system, in which mouse embryonic stem cells (ESCs) carried an extra human chromosome 21, showed that trisomy 21 causes hyperproduction of multipotential immature hematopoietic precursors, accompanied by increased expression of GATA2, c-Kit, and TIE-2. A panel of partial trisomy 21 ESCs, which was mapped by 3.5 kbp-resolution tiling arrays, helped to identify two different, nonoverlapping regions related to abnormal hematopoiesis. A human-specific siRNA silencing experiment revealed that an extra copy of RUNX1, but not ERG or ETS2, raised the levels of TIE-2/c-Kit [121]. Interestingly, the expression levels of c-KIT and SCF/KIT signaling are significantly increased in primary TAM blasts [122].

A recent study with induced pluripotent stem cell (iPSC) lines demonstrated synergistic interaction of trisomy 21 and GATA1 mutations in hematopoietic abnormalities. Banno et al. established systematic TAM models using human iPSCs and genome/chromosome-editing technologies. They found that trisomy 21 promoted expansion of early hematopoiesis. Furthermore, GATA1s is upregulated by trisomy 21, leading to aberrant megakaryopoiesis [123]. Assays for expression of chromosome 21 genes in trisomic and disomic iPSC-derived CD34+CD38− cells identified a group of genes with relatively higher expression levels in trisomic iPSCs. These genes included RUNX1, ETS2, and ERG, which are clustered within a 4 Mb region on chromosome 21. Notable, trisomy of the 4 Mb region is critical to the perturbation of megakaryocyte development via GATA1s upregulation. Knockout of either RUNX1, ETS2, or ERG genes in a single chromosome revealed that the loss of one copy of the RUNX1 gene canceled the accelerating effect on early hematopoiesis, and the increased gene dosage of RUNX1, ETS2, and ERG perturbed megakaryocyte differentiation synergistically via GATA1s upregulation.

2.3.3.3 DIRK1A

Malinge et al. recently used a mouse model of three oncogenic events to show that trisomy of 33 gene orthologs of human chromosome 21, a GATA1 mutation, and an MPL mutation were sufficient to induce ML-DS in vivo. Furthermore, functional screening of the trisomic genes identified DYRK1A, which encodes dual-specificity tyrosine-phosphorylation-regulated kinase 1A, as a potent megakaryoblastic tumor-promoting gene that contributes to leukemogenesis through dysregulation of nuclear factor of activated T cells (NFAT) activation [124].

2.3.4 Additional Genetic Events During the Progression from TAM to AMKL

It has been proposed that other genetic alterations must be acquired in addition to the GATA1 mutation for progression from TAM to ML-DS. Available evidence indicates that acute leukemia could arise from cooperation between one class of mutations that interferes with differentiation, such as loss-of-function mutations in hematopoietic transcription factors, and a second class of mutations that confers a proliferative advantage to cells, such as activating mutations in the hematopoietic tyrosine kinases [125]. Indeed, mutations in JAK1, JAK2, JAK3, FLT3, and TP53 were reported in some ML-DS patients [65, 126,127,128,129,130]. However, the possible presence of additional mutations in ML-DS remains unknown in most cases.

Recently, Yoshida et al. reported genomic profiling of 41 TAM, 49 ML-DS, and 19 non-DS-AMKL samples, including whole-genome and/or exome-sequencing of 15 TAM and 14 ML-DS samples. Non-silent mutations in TAM blasts are primarily limited to the GATA1 gene. In contrast, ML-DS blasts carry a higher burden of mutations, with frequent mutations of cohesin components (53%), CTCF (20%), EZH2, KANSL1, and other epigenetic regulators (45%), as well as common signaling pathways, such as the JAK family kinases, MLP, SH2B3 (LNK), and multiple RAS pathway genes (47%) (Fig. 7.2) [78].

Cohesin is a multiprotein complex consisting of four core components, including the SMC1, SMC3, RAD21, and STAG proteins. In concert with functionally associated proteins such as NIPBL, cohesin is engaged in sister chromatids cohesion, post-replicative DNA repair, and transcriptional regulation. CTCF is a zinc finger protein implicated in long-range regulation of gene expression in collaboration with cohesin. Importantly, all mutations and deletions in different cohesin components were mutually exclusive, suggesting that cohesin function was the common target of these mutations.

Genes commonly observed with chromatin modification were frequently mutated in ML-DS. Especially, EZH2, which encodes a catalytic subunit of the polycomb repressive complex 2 (PRC2), was another recurrent mutational target in ML-DS. The frequency of the mutations was 33%, which is much higher than that in other hematological malignancies. In erythroid cells, PRC2 is involved in epigenetic silencing of a subset of GATA1-repressed genes, such as KIT and GATA2.

Collectively, these findings support a multistep leukemogenesis model whereby TAM is caused by a single GATA1 mutation and constitutive trisomy 21, and subsequent AMKL evolve from a preexisting TAM clone through the acquisition of additional mutations (Fig. 7.3).

2.4 Clinical Management

2.4.1 Clinical Presentation

TAM presents in a wide variety of ways, from asymptomatic alterations in the blood counts to fulminant hepatic fibrosis. Recent prospective and retrospective studies revealed a median age at diagnosis of TAM is 3–7 days [131,132,133]. About 10–25% of TAM patients are asymptomatic at presentation. In symptomatic infants, hepatomegaly, splenomegaly, pleural/pericardial effusions, ascites, jaundice, and bleeding diatheses (bruising, petechiae, or bleeding) are most common findings, whereas hepatic fibrosis, hydrops fetalis, renal dysfunction/failure, and rash are less common. Patients may develop liver fibrosis due to blast cell infiltration that can cause fulminant liver failure and early death. Smrcek et al. reported that 11 of 79 (14%) patients with DS had fetal hydrops, and three of these patients (27%) presented with hepatosplenomegaly and TAM in the second and third trimesters [134].

2.4.2 Diagnostic Procedures

Generally accepted diagnostic criteria for TAM have not been established. Each study group uses its own criteria [79, 131]. For example, according to the Japan Pediatric Leukemia/Lymphoma Study Group (JPLSG), diagnosis of TAM is made if infants with trisomy 21 or mosaic trisomy 21 are up to 90 days at presentation with blasts in the peripheral blood or if infants with normal karyotype are up to 90 days at presentation with TAM-like blasts in the peripheral blood, which have trisomy 21 and GATA1 mutations.

Myeloid blasts in the peripheral blood with megakaryoblastic features are diagnostic hallmarks of TAM. The blasts express variable combinations of CD33, CD41, CD42b, CD117, CD7, CD34, CD45, and CD56. However, a recent study showed that about 98% of DS neonates have circulating blasts [81], suggesting that it is very difficult to diagnose TAM morphologically. The most reliable method to identify TAM blasts is the detection of a GATA1 mutation. We recommend direct sequencing analysis by using complementary DNA (cDNA) prepared from total RNA extracted from peripheral blood cells, and then confirmation of the results using genomic DNA. Sequence analysis of cDNA for the GATA1 mutation is more sensitive than that of genomic DNA because only TAM blasts express GATA1 in the peripheral blood. Recently, Roberts showed that targeted next-generation sequencing (NGS) can detect GATA1 mutation with a sensitivity of 0.3% and that the estimated frequency of TAM is around 30% in DS neonates [81].

2.4.3 Treatment

Most neonates with TAM do not need treatment. The TAM neonates without severe disease can be safely monitored without treatment since their outcome is favorable [79]. The Pediatric Oncology Group conducted the first prospective study of TAM. Three children with life-threatening disease received low-dose cytosine arabinoside (LDCA) (0.4–1.5 mg/kg every 12 h for 5 or 7 days). The disease resolved in all of the patients [135]. In the BFM trial, chemotherapy with LDCA has a beneficial effect on the outcome of those children with risk factors for early death. LDCA (0.5–1.5 mg/kg for 3–12 days) improved the 5-year event-free survival from 28 to 52%. The main aim of treating high risk TAM patients is to immediately improve life-threatening features of TAM. Treatment of TAM did not alter the risk of developing subsequent ML-DS [131].

2.4.4 Prognostic Factors

Prospective and retrospective studies from the United States, Germany, and Japan confirmed that TAM is not a benign disease, as early death was reported in 15–20% of patients [131,132,133]. Risk factors for early death include high white blood cell (WBC) count, preterm delivery, bleeding diatheses, failure of spontaneous remission, increased bilirubin, and liver enzymes. Recent prospective study (TAM-10) of the Japan Pediatric Leukemia/Lymphoma Study Group (JPLSG) confirmed that a high WBC count and effusion are risk factors for early death (unpublished data).

2.5 Future Challenges

Recent studies have provided fascinating insights into the pathogenesis of TAM and AMKL in DS. Most important is the discovery of acquired mutations in GATA1 in TAM and additional genetic events during TAM to AMKL progression. Furthermore, several candidate genes on chromosome 21 that may cooperate with GATA1s have been identified. However, there are still many important questions to be answered. For example, why do GATA1s mutations occur to such a high degree exclusively in children with DS? What is the true incidence of TAM? What are the epigenetic events that occur subsequent to GATA1 mutations that cause malignant transformation from TAM to ML-DS?

Population-based study with more sensitive methods for GATA1 mutation screening will clarify the true incidence of TAM in DS and risk factors that predict malignant transformation from TAM to AMKL. Studies of ML-DS have now begun to lead us toward a fuller understanding of acute myeloid leukemia in children and adults and potentially to novel ways of prevention and treatment of this devastating disorder.