Childhood Cancer Genetics

At a fundamental level, cancer is caused by an accumulation of genetic changes that result in unregulated cell growth and proliferation. These genetic changes can occur by multiple mechanisms that can be inherited (found in the germline) or acquired (somatic). ^[1]  Mutations found in pediatric malignancies can be categorized broadly into driver mutations and passenger mutations. Genetic drivers are mutations or fusions that confer a survival advantage to "drive" malignant proliferation. Passenger mutations are additional mutations found in the cancers that do not contribute to cancer proliferation.

In the discussion below, we will focus on driver mutations. Driver events in pediatric malignancies can occur through loss of function in tumor suppressor genes or gain of function in proto-oncogenes. This can happen via a number of different mechanisms, including point mutations, deletions, copy number alterations, translocations, enhancer hijacking events, chromoplexy, and epigenetic modifications. Our understanding of the landscape of pediatric cancer genomics has improved markedly in the past 20 years with the advent of next generation sequencing and other high throughput genetic sequencing. ^[2]  In this section, we will review the two classic mechanisms by which driver events occur in pediatric and key examples of each mechanism.

Loss of function of tumor suppressor genes

Tumor suppressor genes encode proteins that normally provide negative control of cell proliferation. Their loss of function is a well described mechanism of malignant proliferation. We will discuss three key proteins that function as tumor suppressors implicated in the development of pediatric (and some adult) cancers: pRB, p53, and PTEN.

RB1 encodes the protein pRB and was the first tumor suppressor gene to be molecularly defined. pRB functions as a negative regulatory transcription factor during the G1 to S phase cell cycle transition. ^[3]  Loss of function mutations in this gene are implicated in pediatric retinoblastoma. Knudson first proposed the two-hit hypothesis in the context of retinoblastoma. He hypothesized that loss of function of both copies of RB1 are required for the development of retinoblastoma. ^[4] His prediction was subsequently supported by the cloning of RB1 and by functional studies of pRB. The first mutation of RB1 in cases of retinoblastoma can be either constitutional or somatic, whereas the second mutation is always somatic. In the inherited form of retinoblastoma, the first mutation is present in the germline; an early onset and a high frequency of bilateral disease characterize these cases. In contrast, both mutations in nonhereditary retinoblastoma are somatic. The two-hit hypothesis has now been adopted as a key mechanism for the loss of function of tumor suppressors leading to oncogenesis.

TP53 encodes the protein p53, which is known as the "guardian of the genome." p53 is a transcription factor whose expression is increased by DNA damage and blocks cell division at the G1 phase of the cell cycle to allow DNA repair. In addition, it can stimulate apoptosis of cells containing damaged DNA. ^[5] Targeted disruption of TP53 in the mouse leads to the development of various tumors. Germline mutation of one TP53 allele is found in patients with Li-Fraumeni syndrome. Patients with Li-Fraumeni syndrome are predisposed to sarcomas, breast cancer, brain tumors, adrenocortical cell carcinoma, and acute leukemia; they have a 50% probability of cancer development by age 30 years.

PTEN  encodes a protein kinase of the same name and functions as a tumor suppressor through regulation of cell proliferation. It specifically negatively regulates the PI3K–AKT signaling pathway to induce cell cycle arrest. ^[6]  PTEN deficiencies have been described in a variety of cancers including breast, thyroid, and endometrial cancers. Mutations that disrupt PTEN function define a condition called ​PTEN hamartoma syndrome, which also includes Cowden syndrome, Bannayan-Riley-Ruvalcaba syndrome, and Proteus/Proteus-like syndrome. Children with PTEN hamartoma syndrome are predisposed to the cancers listed above as well as intestinal polyps. ^[7]

Proto-oncogene activation

Activation of proto-oncogenes is a common theme in childhood leukemias and solid tumors. Proto-oncogenes are genes that function normally in healthy cells to promote growth and proliferation. While they are normally regulated by other genes, including tumor suppressor genes, proto-oncogenes can be activated to become oncogenes that are less sensitive to this regulation. Transcription factors, proteins that bind to the regulatory sequences of target genes, compose the largest class of oncogenes identified in pediatric tumors. The main mechanisms for activation of proto-oncogenes consist of point mutations, amplifications, and chromosomal translocations.

Activation by point mutation: Proto-oncogenes can be activated to oncogenes by point mutations resulting in gain of function. An example is RAS, the most frequently mutated proto-oncogene in human cancers. The Ras protein functions by binding GTP to signal cell growth and proliferation. RAS missense mutations causing gain of function have been described in multiple pediatric malignancies, including central nervous system (CNS) malignancies, bone and soft tissue sarcomas, neuroblastoma, and leukemias. Additionally, germline Ras and Ras pathway activating mutations have been characterized in a group of cancer predisposition syndromes, known as RASopathies. ^{[8, 9, 10]}

Amplification: Amplification, an increase in the number of copies of a gene, is another mechanism of proto-oncogene activation. Amplification of the transcription factor MYCN is a well described mechanism of oncogenesis in a number of pediatric and adult cancers.  MYCN amplification was first described in pediatric neuroblastoma. It has since been found to occur in 18-20% of neuroblastomas and 40% of high-risk neuroblastomas. Other pediatric malignancies in which ​MYCN amplification has been described include rhabdomyosarcoma, medulloblastoma, Wilms tumor, and retinoblastoma. Its amplification is typically associated with poorer prognosis regardless of tumor type. ^{[11, 12]}

Chromosomal translocation: Oncogenes can be activated through translocation events resulting in fusions. The first fusion gene described in acute lymphoblastic leukemia (ALL), BCR-abl, is defined by t(9;22)(q34;q11). This fusion, also known as the Philadelphia chromosome, moves the abl proto-oncogene from chromosome 9 into the BCR gene on chromosome 22, resulting in elevated activity of the tyrosine kinase encoded by abl. This translocation occurs in 3-5% of cases of childhood ALL and is present in almost all cases of chronic myelogenous leukemia (CML). Its prognostic significance is variable, but it has a negative impact on outcomes in pediatric ALL. ^[13]

The expansion of sequencing technologies has improved our understanding of the somatic and germline genetic underpinnings of pediatric cancers and the following features that distinguish them from adult cancers ^[14] :

• Pediatric cancers, in general, have fewer somatic mutations than adult cancers. One study reported a mutation rate 14 times lower for pediatric than for adult cancers. ^[15]  This is consistent with the principle that mutations in cells accumulate with age.
• Pediatric cancers are often characterized by a single driver mutation, in contrast to adult cancers that may harbor multiple genetic alterations that drive cancer development and progression. ^[15]
• In pediatric cancers, a larger proportion of these driver mutations are germline events. One study found that 7.6% of cancers in their pediatric cancer cohort were associated with detectable germline mutations. ^[15]
• The genes that are aberrant in pediatric cancers are different than those in adult cancers. Studies have demonstrated only 30-40% overlap with alterations observed in adult cancers in pan-cancer analyses. ^{[15, 16]}
• The types of genomic alterations in pediatric cancers differ from those observed in adult cancers. For example, single nucleotide variants (SNVs) and small insertions or deletions ( indels) are less common in pediatric than adult cancers. In contrast, pediatric cancers have a higher prevalence of structural variants, such as gene fusions and chromosomal rearrangements, and epigenetic modifiers than adult cancer counterparts. ^[17]

The distinctive features of pediatric cancers have significant implications for management. Nearly 50% of pediatric tumors that have been profiled have genomic alterations that can be therapeutically targeted, although most of these potential targets have not been functionally validated. ^{[2, 15]}  However, due to the lack of significant overlap between pediatric and adult cancer genomes, the drugs being developed against adult cancer vulnerabilities are likely to be insufficient to target pediatric cancers. ^[18]  Clinical assays that are specific to pediatric tumors also need to be developed and utilized to more effectively detect genomic alterations unique to pediatric cancers. Lastly, these findings emphasize the need to recommend genetic counseling to patients with pediatric cancers who may harbor germline mutations and benefit from proband testing, risk-adapted therapy, and cancer surveillance.

In this section we will provide an overview of genetic changes characteristic of particular pediatric malignancies. 

Genetic alterations in acute lymphoblastic leukemias

BCR-ABL fusion: The first fusion gene described in acute lymphoblastic leukemia (ALL), BCR-abl, is created by the der(22) of the t(9;22)(q34;q11), which occurs in 3-4% of childhood cases. t(9;22). This translocation is also known as Philadelphia chromosome, and leukemias that have this translocation are also referred to as Ph+ ALL. This translocation moves the abl proto-oncogene from chromosome 9 into the BCR gene on chromosome 22. Before the advent of tyrosine kinase inhibitors (TKIs), Ph+ ALL was associated with a poor prognosis. Survival has improved with the development of TKIs as a targeted therapy aimed at the fusion protein. ^[19]

E2A-PBX1 fusion: The t(1;19)(q23;p13.3) fuses the transactivation domains of the E2A gene (which encodes a basic helix-loop-helix (bHLH) transcription factor) to PBX1 (an atypical homeobox (HOX) gene). It occurs in 25% of childhood pre-B ALL cases and 5% of childhood B-ALL cases. Each of the E2A proteins (E12 and E47) contains a bHLH domain that is responsible for sequence-specific DNA binding and protein dimerization. In addition, the amino terminal portion of E2A contains 2 transcriptional transactivation domains. The E2A-PBX1 chimeric protein contains these transactivation domains fused to the homeodomain of PBX1. E2A-PBX1 binds DNA in a site-specific manner and is a strong transcriptional transactivator. It has been shown to induce malignancies including T-cell lymphomas and AML in transgenic mice. In human leukemia, E2A-PBX1 has been shown to be a coactivator for RUNX1 in the development of pre-B ALL. ^[20]

MLL fusion genes: The MLL gene, located at band 11q23, is altered in 70-80% of infant ALL cases. The gene is altered in ~10% of acute leukemias in all age groups. Additionally, it is present in 70% or more of secondary AML cases that arise in patients who have been treated with topoisomerase II inhibitors. MLL encodes an approximately 500-kd protein that has many functional domains and binding partners. The majority of MLL rearranged leukemias have fusions with one of the following partner genes: AF4 [t(4,11)], AF9 [t(9,11)], ENL [t(11,19)(q23,p13.3)], AF10 [t(10,11)], ELL [t(11,19)(q23,p13.1)], or AF6 [t(6,11)]. MLL rearrangements confer a poor prognosis on infants with ALL; long-term EFS rates are approximately 20%. A subset of these patients, however, particularly those whose disease responds well to initial chemotherapy, have a relatively favorable outcome. ^[21]

TEL-AML1 gene fusion:  The TEL-AML1 fusion gene is created by the t(12;21). In the resulting chimeric protein, the helix-loop-helix (HLH) domain of TEL (also known as ETV6) is fused to the DNA-binding and transactivation domains of AML1 (also known as RUNX1). Molecular techniques have demonstrated that it is present in approximately 25% of childhood ALL cases. TEL-AML1 expression is associated with an excellent prognosis, with event-free survival approaching 90% and additionally is associated with relatively low rates of relapse. ^{[22, 23]}

Activation of transcription factor genes in T-cell ALL: Recurring translocations in T-cell ALL often involve the transcriptionally active sites of the TCR b locus (7q34) or the TCR a and TCR d locus (14q11). These translocations lead to dysregulated expression of transcription factor genes. Transcription factor genes altered in T-cell ALL include members of the bHLH (myc, TAL1/SCL1, TAL2/SCL2, LYL1), LIM (LMO1/RBTN1/TTG1, LMO2/RBTN2/TTG2), and homeodomain (HOX11) families. More than 50% of T-cell ALL cases have mutations that involve NOTCH1; this gene encodes a transmembrane receptor that regulates normal T-cell development. Aberrant NOTCH1 signaling may entail constitutive expression of myc and cooperation with RAS. ^{[24, 25]}

Genetic alterations in acute myelogenous leukemia

Acute myelogenous leukemia (AML) is a heterogenous disease with varied morphologies and outcomes. Pediatric AML has relatively low rates of mutation among cancers that are molecularly well characterized and is a markedly different disease than adult AML. Structural cytogenetic lesions can define subtypes of pediatric AML and inform prognosis. We will describe some of these subtypes and lesions below.

Acute promyelocytic leukemia:   Acute promyelocytic leukemia (APL) is characterized by the fusion of promyelocytic leukemia (PML) gene with the retinoic acid receptor alpha (RARA), creating the PML-RARA fusion gene. This is most commonly due to the balanced translocation t(15;17)(q24.1;q21.2). RARa is a ligand-dependent transcription factor that regulates many genes. PML is a tumor suppressor that plays a role in apoptotic pathways. One mechanism by which the PML-RARA fusion causes AML is by repressing RARA target genes, which leads to blocked differentiation at the promyelocytic stage and inhibits cell death.

APL makes up 5-10% of pediatric AML cases. Prognosis for APL is generally very good, with remission rates over 95% and greater than 80% overall survival. The mainstay of therapy is all-trans-retinoic acid (ATRA), which induces APL cell differentiation. Most current pediatric APL regimens consist of ATRA plus an anthracycline.

Core binding factor AML: This subtype of AML is defined by the presence of either t(8;21)(q22;q22) or inv(16)(p13q22). Both genetic alterations have similar effects on core binding factor (CBF), and both are associated with favorable outcomes. The CBF complex is a heterodimer with a DNA binding alpha subunit and a non-DNA binding beta subunit, CBFβ, which allosterically enhances DNA binding and stability of the complex. It dimerizes with RUNX1 to activate gene expression and causes myeloid cell differentiation. Both t(8;21) and inv(16) fusions inhibit the normal function of the CBF complex, which causes maturation arrest of the myeloid cell lineage. CBF AML accounts for 20-25% of pediatric AML. Prognosis is favorable with almost 90% of patients achieving complete remission with chemotherapy. 

KMT2A rearranged AML: Rearrangements involving KMT2A (Lysine (K)-specific methyltransferase 2A, (previously MLL1) are common in childhood leukemias and are associated with poor prognosis. In AML, KMT2A rearrangements are most common in infants (35-60%) and are less common in children and adolescents (~10-15%). Normally, KMT2A functions in the multiprotein MLL1 complex to regulate gene expression. KMT2A fusions mediate the development of leukemia by activating genes regulated by KMT2A. KMT2A rearrangements typically portend a poor prognosis. 

Other fusions in AML:  Other fusions in AML include NUP98 fusions, KAT6A fusions, MNX1 rearrangements, and CBFA2T3-GLIS2 fusion. ^[26]

Gene mutations in AML:  FLT3 mutations, which involve internal tandem duplications, portend a poor prognosis in adults and children, but the incidence is appreciably lower in children. The prognosis of point mutations in FLT3 is less clear. They occur in 30-40% of acute promyelocytic leukemia. RAS and tyrosine kinase receptor mutations such as in ckit occur in 25% of AML. Prognostic significance is not clearly shown in children. GATA1 mutations are present in most if not all patients with Down syndrome with transient myeloproliferative disorder (TMD) or megakaryocytic AML. GATA1 is a transcription factor receptor for the development of erythroid, megakaryoctes, eosinophils, and mast cells, and increases the sensitivity of cytosine arabinoside. 

Genetic features of Hodgkin lymphoma

Classic Reed-Sternberg cells, the pathognomonic cells associated with Hodgkin lymphoma, typically display immunoglobulin (Ig) variable (VAR) region gene rearrangements. On the chromosomal level, various abnormalities may be found in virtually all cases, suggesting genomic instability, although no specific rearrangement has been recognized as prognostically important. A common phenotype of these cells, however, is that they are resistant to apoptosis, a key feature of many lymphomas. On the molecular level, resistance to apoptosis may be related to elevated or constitutive activity of NF-κB, a pro-survival transcription factor that renders cells resistant to CD95-induced apoptosis. ^[27]

Genetic features of non-Hodgkin lymphoma

Non-Hodgkin lymphoma (NHL) is made up of a heterogenous group of lymphoid tissue malignancies derived from B or T progenitor or mature cells. These consist of Burkitt lymphoma, lymphoblastic lymphoma, anaplastic large cel lymphoma (ALCL), diffuse large B-cell lymphoma (DLBCL), and other rarer variants. ^[28]

Burkitt lymphoma is a high-grade, monoclonal proliferation of B lymphocytes and has 2 major forms, the endemic (African) form and the nonendemic (sporadic) form. This neoplasm is typically extranodal; the endemic form most often involves the mandible or maxilla, and the nonendemic form typically involves the distal ileum, cecum, or mesentery. In approximately 85% of cases, Burkitt lymphoma involves reciprocal translocation of chromosomes 8 and 14 (t(8;14)(q24;q32)), which results in transposition of the c-MYC proto-oncogene on chromosome 8 with either the immunoglobulin heavy chain gene locus (IgH, IGH), the kappa light chain (IGK) gene locus, or the lambda light chain (IGL) locus. In 80% of cases, the translocation occurs with the IGH locus. The c-MYC gene product is a helix-loop-helix transcription factor that functions as a critical regulator of mitosis, differentiation, and apoptosis by pleiotropically influencing DNA acetylation and gene expression. MYC is activated by mitogenic signals via the MAPK/ERK pathway, forms a heterodimer with the Max transcription factor, and is required for cellular entry into S-phase. Following transposition of the c-MYC gene, expressional dysregulation appears to exert a global effect of cell mitosis and apoptosis, leading to malignant transformation of the B-cells in which this translocation occurs.

Epstein-Barr Virus (EBV)  is associated with approximately 95% of cases of the endemic forms of Burkitt lymphoma and 20-30% of the sporadic forms; recent evidence has begun to elucidate the mechanism by which latent infection of B lymphocytes promotes tumorigenesis. ^{[29, 30]}  In memory B-cells with latent EBV infection, expression of Epstein-Barr nuclear antigen-1 (EBNA1) appears to exert a direct antiapoptotic effect within the cell. It also reduces expression of major histocompatibility complex (MHC) class I molecules and antigen-processing (TAPasin) proteins. Thus, latent EBV infection may promote B-cell tumorigenesis by altering the programmed cell death pathway, leading to cell survival even in the presence of chromosomal abnormalities (ie, translocations) and by facilitating tumor evasion of the immune system following transformation.

Newman et al examined the clinical significance of TP53 abnormalities in a large cohort of British children with B-cell non-Hodgkin lymphoma. TP53 abnormalities, which were found in nearly 55% of patients, were independently associated with significantly inferior survival, compared with patients who did not have a TP53 abnormality (progression-free survival, 70.0% vs 100% [P< 0.001]; overall survival, 78.0% vs 100% [P=0.002]). ^[31]

ALCL is commonly characterized by a T-cell phenotype, expression of the CD30 antigen, and an aggressive clinical profile with peripheral adenopathy and skin involvement. Approximately 90% of cases of ALCL harbor the t(2;5)(p23;q35), which fuses the region encoding the N-terminal portion of nucleophosmin (NPM) on chromosome 5 to the region encoding the tyrosine kinase domain of ALK on chromosome 2, creating an NPM-ALK chimera, a constitutively active tyrosine kinase. Overexpression of ALK in lymphoid cells appears to contributes to tumorigenesis by inappropriate phosphorylation of yet unidentified target proteins.

Genetics of neuroblastoma

Neuroblastoma is characterized by multiple genetic alterations including amplification of MYCN, mutations in ​ALK, and other genetic and segmental chromosomal alterations. ^[32]  MYCN, an oncogene that encodes the transcription factor N-MYC, is identified in 20% of neuroblastoma. ^[33]  The presence of ​MYCN amplification is a powerful predictive biomarker independent of age and stage, is associated with poor outcomes, and is used in risk stratification and treatment planning. ^[34]  ​Somatic mutations in ​ALK, encoding anaplastic lymphoma kinase, have also been demonstrated in 9-14% of high-risk neuroblastoma, providing a therapeutic target that can be leveraged for some patients. ^[35]  Since both ​MYCN and ALK are located on chromosome 2p, they can be co-amplified. Whole-genome sequencing has also identified loss-of-function alterations in ATRX, encoding RNA helicase ATRX, and TERT, encoding telomerase reverse transcriptase in 10-25% of neuroblastoma. ^[36]  Segmental chromosomal alterations associated with poor survival include loss of heterozygosity of 1p, typically observed in combination with MYCN amplification, and 11q, which inversely correlates with MYCN ampflication. ^[37]

Familial neuroblastoma is rare, comprising only 1-2% of all cases of neuroblastoma, and is associated with germline gain of function in ALK or germline loss of function in PHOX2B, encoding paired mesoderm homeobox protein 2B that is also observed in combination with central congenital hypoventilation syndrome and Hirschsprung disease.

Genetics of Wilms tumor

Knowledge of the genetic underpinnings of Wilms tumor was originally derived from analyses of patients with Wilms tumor predisposition syndromes. Wilms tumor occurs in the context of a cancer predisposition syndrome in 15% of cases. These syndromes include Beckwith-Wiedemann syndrome (BWS), Denys-Drash syndrome, and WAGR syndrome (Wilms tumor, aniridia, genitourinary anomalies, and intellectual disability). Cytogenetic studies of these patients identified the central roles in tumorigenesis for Wilms tumor 1 (WT1), the WNT pathway, and insulin-like growth factor (IGF) signaling. ^[38]  WT1, a tumor suppressor gene at 11p13, encodes a transcription factor that is critical for urogenital development. A second Wilms tumor locus has been identified at 11p15 with abnormal methylation at this site responsible for BWS. In another form of BWS, uniparental disomy or H19 hypermethylation results in biallelic expression of IGF2 and overactivation of the IGF signaling pathway that carries a 20% risk of developing Wilms tumor.

In the absence of an underlying cancer predisposition syndrome, somatic mutations in WT1, LOH at 11p15, the WNT pathway, TP53, and MYCN can drive malignant transformation.

Genetics of osteosarcoma

The genetic landscape of osteosarcoma is complex, comprised namely of large structural rearrangements likely secondary to chromothripsis, chromoplexy, and kataegis, in addition to a few frequently mutated genes. ^[39]  These rearrangements result in copy number alterations in ​PTEN,CDKN2A/2B, and MYC. ^[40]  The most frequenty altered gene is TP53, lost in >90% of cases of osteosarcoma. Deletion of RB1 is also frequently seen in up to 30% of cases of osteosarcoma. ^[41]

While most cases of osteosarcoma are sporadic, up to 25% of cases of osteosarcoma occur in the setting of an underlying germline mutation resulting in a cancer predisposition syndrome. These syndromes include Li-Fraumeni, hereditary retinoblastoma, Diamond-Blackfan anemia, and disorders involving mutations in the RECQ family resulting in DNA helicase abnormalities such as Rothmund-Thomson, RAPADILINO, Bloom, and Werner syndromes. The age of onset for tumors in these patients is younger than in sporadic cases.

Genetics of Ewing sarcoma

Ewing sarcoma is a small round blue cell tumor defined by recurrent balanced chromosomal translocations involving the TET or FET family of genes, including EWSR1 and FUS, and the ​ETS family gene FLI1. More than 85% of Ewing tumors are characterized by a translocation t(11;22)(q12;q24) that produces a chimeric transcription factor ​EWS fused to the DNA binding domain ​FLI1. ^[42]   Among the 15-20% of Ewing sarcoma cases that do not have an EWSR1-FLI1 fusion, variant fusions including fusions of EWSR1 to ERG, ETV2, ETV4, and FEV2 have been described. ^[43]  These FET-ETS gene fusions result in a chimeric peptide that is an oncoprotein that binds DNA, resulting in aberrant transcription and regulating genes involved in the cell cycle cell migration, telomerase activity, and other functions. Although retrospective studies have suggested that different transcripts have prognostic implications, prospective studies have been unable to validate this observation. ^[44]  

Additional recurrent mutations are rare. Copy number variations have been identiied, including chromosome 1q gains and chromosome 16q losses, both associated with poor clinical outcomes.

Genetics of rhabdomyosarcoma

The discovery of genetic fusions in pediatric rhabdomyosarcoma (RMS) and their prognostic signficance has resulted in a new classification system of RMS that takes into account fusion status. Chrosomal translocations, containing the common t(2;13)(q35;q14) or the rare t(1;13)(p36;q14), have been identified in most patients with alveolar RMS. ^[45]  Both translocations disrupt the FKHR gene (FOXO1) on chromosome 13, which encodes a widely expressed transcription factor, with juxtaposition of PAX3 on chromosome 2 or ​PAX7 on chromosome 1. This generates ​PAX3-FOXO1 and PAX7-FOXO1 fusion genes that encode transcription factors that have enhanced transcriptional activity and can contribute to malignant transformation with activation of various cancer-promoting pathways. ^[46]  The presence of PAX3/PAX7-FOXO1 is associated with a poor prognosis. In addition, tumors that are alveolar but fusion negative behave more like tumors with embryonal histology, suggesting that fusion status has more wide-reaching prognostic significance than morphology. As a result, clinical trials studying RMS are adopting risk stratification systems that recognize fusion status. ^[47]

In addition, several mutations have been shown to have prognostic significance in pediatric RMS, namely MYOD1 L122R. Whereas mutations in MYOD1 account for only 3% of fusion negative RMS, MYOD1 mutated tumors are very aggressive and have poor outcomes. ^[47]

Genetics of non-rhabdomyosarcoma soft tissue sarcoma

There are many genetic aberrations that characterize non-rhabdomyosarcoma soft tissue sarcoma (NRSTS). The following list highlights some of the more frequently observed genetic findings:

• Alveolar soft-part sarcoma: t(X;17)(p11.2;q25), which fuses the  ASPSCR1 gene to  TFE3
• Desmoid type fibromatosis: 85% of desmoids have gain-of-function mutations in B-catenin gene,  CTNNB1, and a minority of desmoids have germline or sporadic loss-of-function mutations in  APC, resulting in activation of B-catenin
• Desmoplastic small round cell tumor: t(11;22)(p13;q12), which fuses  EWSR1 with  WT1
• Infantile fibrosarcoma: t(12;15)(p13;q25), which fuses  ETV6 with variable partners including  NTRK1, NTRK2, or most commonly  NTRK3
• Myxoid liposarcoma: fusions of  DDIT3 on chromosome 12 with variable partners including  EWSR1 and  FUS
• Synovial sarcoma: t(X;18)(p11;q11), which fuses the  SYT gene at 18q11 to either the  SSX1 gene (xp11.23) or the  SSX2 gene (Xp11.21) with poorer prognosis associated with former fusion
• Undifferentiated small cell sarcoma: previously classified as  EWSR1-fusion negative "Ewing-like" or "Ewing family" of tumors, now known to have distinctive molecular and clinical features; includes fusions with  CIC on chromosome 19  such as CIC-DUX4 t(4;19)(q35;q13) and  BCOR-CCNB3 with X-chromosome inversion

Genetics of CNS tumors

The most common brain tumor in childhood is a low-grade glioma, comprising 30-40% of all pediatric CNS tumors. The genetic alteration most frequently observed in low-grade gliomas involves the MAPK (RAS-mitogen-activated protein kinase) pathway with aberrations in the BRAF oncogene secondary to mutations at BRAF V600E or fusions of BRAF with KIAA1549. Low-grade gliomas can also occur in the context of a cancer predisposition syndrome such as neurofibromatosis type 1 and tuberous sclerosis. Both of these syndromes are secondary to mutations in tumor suppressor genes that dysregulate signaling in MAPK and mTOR pathways. High-grade gliomas, namely pediatric glioblastoma, have recurrent somatic mutations in histone 3 variants, including H3.1 and H3.3 K27M seen in diffuse midline gliomas, in up to 50% of cases In addition, high-grade gliomas in children aged older than 3 years have somatic mutations in TP53 and overexpression of p53 that correlates with poorer progression-free survival.

Medulloblastoma is the most common CNS embryonal tumor in childhood. There are four subgroups of medulloblastoma: wingless (WNT), SHH (sonic Hedgehog), Group 3, and Group 4. WNT and SHH were named after the signaling pathwyas involved in these subgroups. The WNT subgroup, with genetic alterations including monosomy 6 and mutations in CTNNB1, is associated with the best prognosis. The SHH subgroup can be characterized by somatic or germline mutations in PTCH1, SUFU, and TP53, and MYCN amplification. Group 3 medulloblastoma is associated with the worst prognosis. The defining genomic alteration in this subgroup is MYC amplification seen in 15% of patients and associated with poor outcomes. Group 4 is the most common subgroup with segmental chromosomal alterations in chromosome 11 and 17 that portend a superior prognosis. Cancer predisposition syndromes associated with an increased risk of medulloblastoma include Gorlin syndrome with mutations in PTCH1 and SUFU, Turcot syndrome type 2 with mutations in APC, Li-Fraumeni with mutations in TP53, and subsets of Fanconi anemia with mutations in PALB2 and BRCA2.

Ependymoma can be molecularly classified based on DNA methylation profiling data. Ependymomas that occur in the posterior fossa are classified as PF-A tumors or PF-B tumors. Gain of chromosome 1q and loss of 13q are associated with poorer outcomes in posterior fossa ependymomas. Supratentorial ependymomas are characterized by the presence of RELA or YAP1 fusions. Deletions of CDKN2A/2B in RELA-fused tumors are associated with inferior outcomes and survival. Up to 50% of patients with neurofibromatosis type 2 (NF2) develop ependymomas. ^[48]

Cancer predisposition syndromes are defined by germline mutations in tumor suppressors or proto-oncogenes that predispose affected children to malignancy. Compared with adults, cancer predisposition syndromes play a greater role in malignancy formation in children. In general, it is thought that children with the following features or family history should be screened for a germline cancer predisposition: family history of the same or similar cancers; personal history of multifocal or multiple cancers; personal history of early age at diagnosis compared to sporadic tumors of the same type; phenotypic physical examination features consistent with a predisposition syndrome; and occurrence of specific tumor types associated with a genetic predisposition syndrome. ^[49]  Beyond this, screening for patients has expanded with next-generation sequencing technologies that allow for more comprehensive genetic screening of patients with current malignancies.

Germline mutations associated with cancer predisposition have been seen in 7-10% of children with malignancy using next-generation sequencing based screening. ^{[50, 51, 52, 15]}  A study by Zhang et al found germline mutations in cancer-predisposing genes were identified in 8.5% (95 patients) of the children and adolescents with cancer. The study sequenced DNA from blood samples of 1120 patients younger than 20 years of age. This included whole genomes (595 patients), whole exomes (456 patients), or both (69 patients). The study also added that family history did not predict the presence of an underlying predisposition syndrome in most patients. ^{[50, 51]}  These germline mutations affect multiple parts of pediatric oncology patient care, including choice of therapy and future malignancy screening. 

Enhanced screening for and early detection of malignancy is important in patients with germline cancer predisposition syndromes. In 2017, the American Association for Cancer Research (AACR) published consensus recommendation guidelines for malignancy screening in the 50 most common cancer predisposition syndromes that predispose to malignancy within the first 20 years of life. These included Li Fraumeni syndrome, neurofibromatoses (ie, NF1, NF2), overgrowth syndromes (ie, Beckwith Wiedemann), Wilms tumor–associated syndromes, neural tumor syndromes (ie, hereditary neuroblastoma, malignant rhabdoid tumor syndrome), gastrointestinal cancer syndromes (ie, familial adenomatous polyposis, Lynch syndrome), neuroendocrine syndromes (ie, multiple endocrine neoplasia 1, 2A, 2B), leukemia predisposition syndromes (ie, ataxia-pancytopenia syndrome), and DNA instability syndromes (ie, Fanconi anemia, ataxia telangiectasia) among other miscellaneous syndromes that did not fit into these categories (ie, PTEN hamartoma syndrome, pleuropulmonary blastoma syndrome). Based on published data for each of these syndromes, the group proposed recommendations for cancer screening and preventive services. Recommendations included the type of screening modality (ie, CT, MRI, laboratory studies, physical examination), ages at which screening should start, frequency at which screening should occur, and any preventive procedures (ie, surgery) that should be undertaken. ^[53]  Although the efficacy of the AACR screening recommendations will be observed in the future, there is evidence that screening patients with known cancer predisposition syndromes improves survival. A prospective 11-year study in patients with Li Fraumeni syndrome concluded that early tumor detection through surveillance was associated with improved long-term survival. ^[54]

Whereas certainly identification of children with cancer predisposition syndromes and subsequent enrollment in malignancy screening has benefits, there are also ethical questions surrounding this practice. The screening tools recommended for these children are often invasive. For example, many small children need anesthesia to be sedated for screening imaging. There is also not universal insurance coverage for all screening and preventive procedures. Additionally, there are potential psychological implications to identification of a cancer predisposition syndrome and screening for both the affected child and parent. Finally, there are likely cancer predisposition genes or syndromes that have not yet been defined; in some cases, patients may meet criteria to be assessed for a cancer predisposition syndrome (strong family history or personal history of multiple malignancies) but do not have a cancer predisposition found on screening. Uncertainty remains regarding how to screen these children. ^[55]

Whereas both pediatric and adult malignancies are driven by mutations that cause loss of tumor suppressor genes or activation of oncogenes, the genetics of pediatric and adult malignancies are quite distinct. Pediatric malignancies have fewer somatic mutations, different types of genetic alterations, and a higher prevalence of germline mutations compared with adult malignancies. Characterization of the genes situated at translocation breakpoints in childhood tumors has helped provide new insights into the mechanisms of malignant transformation, develop sequencing platforms specific to pediatric cancers, refine risk-adapted treatment algorithms based on molecular profiling, and identify potential therapeutic targets. Additional efforts are needed to further define the unique and complex genomic landscape of pediatric cancers.