Introduction

Acute leukemia is the most common pediatric cancer and accounts for approximately one third of childhood cancers, and roughly 80% are lymphoid (acute lymphoblastic leukemia (ALL)) and 20% are myeloid (acute myeloid leukemia (AML)) [1]. While the majority of childhood leukemia cases occur in the absence of any known predisposing factor, a small proportion are truly familial or caused by known hereditary cancer syndromes [2]. Both ALL and AML can be seen with a variety of hereditary cancer syndromes. According to the magnificent past review articles, the categories of leukemia-associated hereditary cancer syndrome can be divided into some groups in several ways [3,4,5,6]. Moreover, recent advances in high-throughput genotyping technology enable comprehensive screens of genetic variation, revealing new genetic variations which are prone to develop leukemia [7]. This review focuses on describing the clinical features of multiple germline syndromes that confer an increased risk of leukemia, and will specifically highlight the process of leukemia development in Down syndrome.

Hereditary cancer syndrome associated with leukemia

Inherited cancer syndromes associated with leukemia like familial myelodysplastic syndromes (MDSs) and AML syndromes were once considered rare, but are more frequent than previously expected. The number of genes involved in inherited MDS/AML has grown recently with the advance of genomic sequencing technologies, and the known hereditary cancer syndromes associated with leukemia risk account for more than 60 different genes. Each syndrome has a different risk for ALL or AML, sometimes bearing other hematological abnormalities [6]. Table 1 is a modified comprehensive list of the leukemia-associated syndromes from the excellent reviews by Seif [2], Malkin [3], and Stieglitz [4]. To discuss the leukemia-associated inherited cancer syndromes, we divide these syndromes into the following eight main categories based on clinical features, biological functions, and affected pathways: (1) genetic instability/DNA repair syndromes, (2) RAS pathway dysfunction, (3) bone marrow failure syndromes, (4) telomeropathies, (5) immunodeficiency, (6) transcription factor abnormalities, (7) pure familial leukemia, and (8) aneuploidy. Although some of these syndromes belong to multiple categories, we have classified each syndrome to the most representative category by its clinical features. For the purpose of this review, we will briefly describe each category, pick up some hereditary syndromes, and discuss about the process of leukemia development in some of these hereditary cancer syndromes.

Table 1 Hereditary syndromes associated with hematological malignancies [2,3,4]
Full size table

DNA repair/genetic instability

Genomic instability is a characteristic of most cancers and leukemia [29]. In hereditary cancer syndromes, genomic instability results from mutations in DNA repair genes and drives malignancy development. A most representative syndrome in this category is Li-Fraumeni syndrome (LFS) [30, 31], caused by p53 mutation [32, 33]. The p53 protein is a transcription factor, upregulating the transcription of target genes involved in cell cycle arrest, DNA repair, apoptosis, and senescence, in response to DNA damage. Mutations in p53 can accumulate additional genetic mutations in hematopoietic progenitor cells. Patients with LFS develop multiple tumors, and also develop leukemia (mostly ALL, and to a lesser extent AML) in 4% of affected mutation carriers in an excellent past review which also found a median age of onset of 12 years [8•]. The newly described constitutional mismatch repair-deficiency syndrome (CMMRD) is caused by bi-allelic (homozygous) alterations in the mismatch repair (MMR) genes, including MLH1, MSH2, MSH6, and PMS2. Some of the manifestations of the syndrome resemble those seen in neurofibromatosis type I, and are associated with multiple café-au-lait spots, pediatric brain tumors, colorectal cancers, and pediatric hematologic malignancies including both ALL and AML [9]. A third of the patients develop hematological malignancies in CMMRD [34]. These MMR gene mutations also cause the adult-onset autosomal dominant Lynch syndrome, previously referred to as hereditary non-polyposis colorectal cancer (HNPCC) [35]. The category of DNA repair/genetic instability also includes Werner syndrome (caused by WRN), Rothmund-Thomson syndrome (RECQL4), Bloom syndrome (BLM), Fanconi anemia (FANC genes), ataxia telangiectasia (ATM), and Nijmegen breakage syndrome (NBS1). Fanconi anemia (FA) is an autosomal recessive (AR) disorder that leads to increased chromosomal breakage through defects in DNA repair [36]. Most FA patients exhibit developmental abnormalities, developing bone marrow failure, AML, and solid malignancies. Bone marrow failure often occurs in childhood, more than a third will develop leukemia, and nearly half will develop MDS [13]. In recent decades, 19 human genes have been found in the cause of FA [14]. These genes code for a group of associated FA, which function cooperatively in a DNA damage recognition and repair. Dysfunction of these protein leads to genomic instability and results in development of MDS and AML.

RAS pathway dysfunction

This category of hereditary syndromes includes Noonan syndrome (PTPN11, K-RAS, N-RAS, etc.), CBL syndrome (CBL), and nurofibromatosis type 1 (NF1). Cancer and leukemia are diseases of uncontrolled cell division and usually linked to a series of dysfunctions in the activity of cell cycle regulators. RAS signaling affects many cellular functions, which includes cell proliferation and differentiation in both normal and malignant cells [37]. Juvenile myelomonocytic leukemia (JMML) is a unique hematopoietic disorder of infancy caused by excessive proliferation of cells of mono and granulocytic lineages [38]. Approximately 90% of patients with JMML carry either somatic or germline mutations of PTPN-11, K-RAS, N-RAS, CBL, or NF1 in their leukemic cells. Allogeneic hematopoietic stem cell transplantation (HSCT) remains the therapy of choice for most patients with JMML. A recent report recommends that HSCT be promptly offered to any child with PTPN-11, K-RAS, or NF1-mutated JMML and to the majority of those with N-RAS mutations but not CBL mutations [39]. Because JMML patients with CBL mutations and few of those with N-RAS mutations may have spontaneous resolution of hematologic abnormalities, the decision to proceed to HSCT in these patients must be weighed carefully.

Bone marrow failure

This category contains a variety of clinical syndromes with many different mutated genes, including Diamond Blackfan anemia (DBA), Shwachman-Diamond syndrome (SDS), familial aplastic anemia with SRP72 mutation, congenital amegakaryocytic thrombocytopenia (CAMT), thrombocytopenia and absent radii (TAR) syndrome, thrombocytopenia with ANKRD26 mutation, and Kostmann syndrome. FA also can be included this category. These bone marrow failure syndromes are considered as a risk factor for clonal evolution [40]. In these syndromes, MDS and AML seem to be the most common hematological malignancies with risks ranging from 5 to 25% [18, 41,42,43,44,45,46]. Careful observation of any leukemia-related signs or symptoms has been offered as management for these syndromes. Recently, RPS29, RPS27, and RPL27 genes are reported to be corresponding for DBA by whole exome sequencing [47, 48].

Telomeropathies and immunodeficiency

Table 1 shows a list of telomeropathies and immunodeficiency syndromes associated with leukemia risk. Dyskeratosis congenita (DC) is the representative disease of telomeropathies caused by dysfunction in TERT and TERC, leading to abnormal telomere maintenance [20]. Patients with DC are at increased risk for bone marrow failure (BMF), MDS or AML, solid tumors (usually squamous cell carcinoma of the head/neck or anogenital cancer), and pulmonary fibrosis. To date, ACD, CTC1, DKC1, NHP2, NOP10, PARN, RTEL1, TERC, TERT, TINF2, and WRAP53 are the genes in which pathogenic variants are known to cause DC. HSCT is the only curative treatment for hematological malignancies but has had poor long-term efficacy in DC patients, because of their high incidence of treatment-related toxicities. Therefore, reduced-intensity preparative regimens being studied in a few institutions may improve long-term outcomes [49].

Transcription factors

Hereditary syndromes in this category are caused by mutations in genes that code for hematopoeitic transcription factors. Alterations in the associated genes (RUNX1, GATA2, PAX5, SH2B3, and ETV6) are also known to occur in de novo or sporadic leukemia cells. Inherited or sporadic GATA2 gene mutations induce MonoMAC syndrome (monocytopenia with atypical mycobacterial infection), DCML deficiency (loss of dendritic cells, monocytes, and natural killer and B lymphoid cells), Emberger syndrome (lymphedema with MDS), and familial AML/MDS [22•]. Familial AML/MDS caused by GATA2 deficiency is categorized the next group “pure familial MDS/AML” because of without other clinical symptoms. Various types of heterozygous mutations, including substitutions, indel, and deletions, have been detected, scattering among five translated exons of the GATA2 gene [22•]. As a result of these alterations, either a mutant GATA2 protein or no GATA2 protein is produced. Reduced expression of the GATA2 gene in HSCs caused by a heterozygous mutation/deletion appears to play a role in the onset of GATA2-associated hematologic disease [50]. More than half of patients with GATA2 deficiency develop MDS, most of which acquired unfavorable chromosomal abnormalities, and finally progressed to AML [22•]. Recently, we reported a family of GATA2 deficiency treated by reduced intensity stem cell transplantation (RIST) [51]. Incidence of MDS/AML gradually increases with age in GATA2 deficiency, and AML with GATA2 deficiency results in poor outcome unless successfully transplanted. Allowing for these clinical findings, RIST could be a promising treatment option for patients with GATA2 deficiency before progression of advanced MDS and AML.

Other pure familial MDS/AML

We defined familial MDS/AML without other systemic clinical features as “pure familial MDS/AML” in this review. The first genes involved in this category, CEBPA, were identified a decade ago [23]. There have been more genes in this category, exemplified by multiple autosomal dominant, highly penetrant inherited cancer syndromes resulting from germline mutations in GATA2, DDX41 ATG2B, and GSKIP [22•, 24, 52]. Most of these syndromes are inherited in an autosomal dominant manner and maybe grouped by their clinical presentations. Familial mosaic monosomy 7 is also included in this category. Of note, individuals with a family history of monosomy 7 may initially have a normal karyotype in peripheral blood and/or bone marrow and later transition to mosaic monosomy 7 [25].

Aneuploidy

Acquired somatic chromosomal aneuploidies are the most common genetic aberrations in sporadic leukemia. Constitutional aneuploidies also associate with the risk for specific hematological malignancies. Trisomy 21 and trisomy 8 are representatives of constitutional aneuploidies of hereditary cancer syndrome associated with leukemia. Constitutional trisomy 8 is lesser than trisomy 21 [53,54,55], and can be seen mostly as a mosaic in the blood or the skin [28]. Meanwhile, trisomy 8 is the most common among sole cytogenetic abnormalities in both AML and MDS with respective incidences of 6 and 11% [56]; it is sometimes difficult to distinguish constitutional or not [57].

Down syndrome (DS) results from trisomy 21 and occurs in 1 in 700–1000 births [54, 55]. Patients with DS show a 10- to 20-fold higher risk of acute leukemia [58, 59]. The most marked increase in incidence in DS infants is with acute megakaryoblastic leukemia (AMKL). The relative risk of developing AMKL is estimated to be 500 times higher in children with DS than in the general population [60]. In most cases, DS-AMKL is preceded by a temporary form of megakaryoblastic leukemia unique to newborns with DS and known as transient abnormal myelopoiesis (TAM). TAM presents in the fetus or a few days after birth, and in most cases resolves spontaneously within 3 months [61, 62]. After spontaneous remission, 20% of TAM patients develop MDS and DS-AMKL within 4 years [60, 63].

Multistep leukemogenesis in DS-related myeloid disorders

TAM has been considered a preleukemic state and is a suitable pathological condition to analyze the evolutionary process of leukemia. GATA1 is a hematopoietic transcription factor, and GATA1 somatic mutations cause both TAM and subsequent DS-AMKL [64,65,66,67,68]. GATA1 mutation has been reported to be ∼30% of neonates of DS [61, 69], and disappear when TAM/ML-DS enters the remission phase [64]. Although TAM is most commonly seen in infants with DS but barely in those with non-DS, suggesting trisomy 21 also a requirement for development of TAM [70]. Recently, next-generation sequencing (NGS) methodology showed the precise frequency of GATA1 mutation in DS [69]. Besides, NGS studies has also showed that the additional genomic mutations/deletions in the genes coding cohesion components, CTCF, and other epigenetic regulators including EZH2 frequently preceded other driver mutations during progression form TAM to DS-AMKL [26••, 71]. These findings of mutations in epigenetic regulators in DS-AMKL suggest that epigenetics also play a role in the development of DS-AMKL. Mutations are also observed in members of signaling pathways, such as the JAK family of kinases, MPL, SH2B3, and multiple RAS pathway genes [26••, 71]. Importantly, these genomic analyses of DS-AMKL confirm that it evolves from the cells responsible for TAM. Collectively, in the setting of trisomy 21, GATA1 mutations cause TAM (Fig. 1). Although most TAM resolves spontaneously, further mutation in cohesin, CTCF, EZH2, or other epigenetic regulators in residual TAM cells occurs with or without mutation of signal-transducing molecules, leading to AMKL. Considering the age of leukemia onset ranges from childhood to late adulthood in other hereditary leukemia syndromes (often higher than DS-AMKL) [8•, 18, 22•, 23, 72], the fact that most of all DS-AMKL patients develop leukemia within 4 years after birth may indicate that trisomy 21 is a driving event in leukemogenesis by itself [60, 63]. Actually, gain or amplification of chromosome 21 is sometimes observed in poor prognostic childhood leukemia [73, 74], and there are several candidate genes likely to contribute leukemogenesis in chromosome 21 such as RUNX1, APP, ETS2, Dyrk1a, and ERG [75, 76]. Our experimental TAM xenograft model has revealed the existence of multiple genetically distinct subclones in TAM phase [77]. It also enabled the observation of clonal selection and expansion of minor mutant TAM clones, demonstrating the striking genetic heterogeneity and the propagating potential of minor clones in a preleukemic phase [77].

Fig. 1
figure 1

Proposed model of ML-DS pathogenesis. Reprinted from Saida S: Evolution of myeloid leukemia in children with Down syndrome. Int J Hematol 2016, 103:365–372, with permission from the Japanese Society of Hematology. Trisomy 21. vertical lines; GATA1 mutation, black and white triangles; cohesin, CTCF, and other epigenetic regulator mutation, black circle; kinase-signaling molecule mutation, black star.

Full size image

Conclusions and future directions

Recently, multiple hereditary predisposition syndromes to leukemia have been discovered owing to progression of new genetic technologies including next-generation sequencing, and more genes are likely to be identified in the future. Hereby, patients with these hereditary syndromes are likely not as rare as previously considered. Most of these patients with leukemia need to be treated chemotherapy and/or allogeneic HSCT, and they also need to be accommodated with their own hereditary syndromes because of their fragileness. These issues become especially important when planning a HSCT, because the patients with some of these syndromes are also at increased risk of developing complications by treatments. These genetic abnormalities are also seen as somatic mutations frequently in MDS/AML cells. Therefore, it is important to discriminate between somatic mutations and germline mutations when evaluating an individual for the syndromes associated with germline alterations in these genes. For that reason, physicians should be familiar with these hereditary predisposition syndromes and obtain every clinical symptom and detailed family histories for patients. The identification of these syndromes is a critical step toward individualized follow-up and treatment. The insight provided by these leukemia-associated syndromes will also have diagnostic and possibly therapeutic implication for all patients diagnosed with sporadic leukemia or other hematologic malignancies. Recent advances in genetic research will much more contribute to diagnosis and identification of causative genes of these syndromes [78].

The pathogenesis for leukemia in these syndromes is now beginning to be understood in terms of accumulation of genomic alterations. There would be different mechanisms between each syndrome in the course of leukemia development. In these hereditary syndromes, especially DS-associated myeloid malignancies (TAM/DS-AMKL) are an attractive model to investigate multistep leukemogenesis. The recent findings of mutations in epigenetic regulators in DS-AMKL suggest that epigenetics also play a role in the development of leukemia [27]. Our ongoing research aims to unravel the molecular mechanisms by which these mutations lead to malignancy.