Genetic Predisposition to Hematologic Malignancies

Abstract

Hematologic malignancies (HMs) have been increasingly recognized in association with an underlying genetic predisposition syndrome (GPS) in individuals of all ages. It is critical for hematology and oncology providers to be aware of the diagnostic findings, physical examination findings, and aspects of family history that raise suspicion for an underlying GPS. Moreover, recognition of how somatic gene panel testing, frequently done at the time of HM diagnosis, may raise suspicion for an underlying germline condition based on the mutation profile reported, is prudent. With knowledge of an underlying germline condition, the chemotherapy used for a given HM may be impacted and the role of hematopoietic stem cell transplant more critically considered. Off-therapy monitoring after HM treatment is completed will also likely be impacted. In this work, we review key features of several GPSs associated with increased risks for HM while also outlining the diagnostic workup to identify GPSs and treatment considerations for affected patients. Armed with this knowledge, treating providers may evaluate the possibility of a GPS in patients with leukemia/lymphoma and modify their treatment plan accordingly.

Until recently, hematologic malignancies (HMs), which represent ∼25% of all pediatric cancers (Linabery and Ross 2008), were largely thought to be sporadic in nature and little was known about germline predisposition syndromes that increase the risk for leukemia or lymphoma. We now know that genetic predisposition syndromes (GPSs) play a larger role in both pediatric and adult HMs than previously understood. Although studies are limited, ∼4% of children with acute lymphoblastic leukemia (ALL) have been reported to have a GPS (Klco and Mullighan 2021), and 14%–39% of adults with a myeloid disorder had a co-occurring GPS diagnosis (DiNardo et al. 2016; Keel et al. 2016). This range reported in adults is similar to the rate of genetic predisposition in breast, colorectal, and other solid tumors.

We also know, based on natural history data, that some germline mutations, such as those in SAMD9 and SAMD9L, confer HM risk early in life, whereas others, like germline DDX41 abnormalities and short telomere disorders, may take six or seven decades to contribute to leukemic transformation (Gurnari et al. 2023). This variable timeline suggests that some hematopoietic developmental stages may be more susceptible than others to secondary somatic events that result in malignancy or that the presence of a specific germline mutation may, in itself, directly impact the risk for leukemic transformation at particular developmental stages. Moreover, the penetrance for myeloid or lymphoid disease varies based on the germline condition present (Klco and Mullighan 2021). In this work, we will review clinical symptoms and signs that are associated with GPSs that predispose to HM and highlight some of the clinical features that may be seen in affected patients. We will also discuss how the interplay between clinical history and presentation, family history, and results of somatic testing may help inform GPS risk and we will review the optimal strategies to confirm germline versus somatic origin of identified mutations. Moreover, we will describe clinical scenarios in which certain HM diagnoses warrant germline testing regardless of clinical presentation or family history. For those identified as having a GPS and associated HM, we will discuss how germline results can impact decisions regarding treatment, the role of hematopoietic stem cell transplant (HSCT) as well as potential HSCT donor considerations. Finally, we know that surveillance for HMs in a patient with a known GPS is a topic of much uncertainty and provider variability, and we will highlight some of the considerations at play when considering frequency and type of monitoring. In summary, increased awareness of the interplay between GPSs and HMs will ensure optimal medical management of patients and may provide opportunities for prevention or early intervention at the time of malignancy diagnosis.

DIAGNOSTIC CONSIDERATIONS

Clinical History and Symptoms/Signs Suggestive of GPS in Patients with HM

In keeping with increased awareness for an underlying GPS manifesting with an HM, the latest World Health Organization (WHO) leukemia classification system includes “germline predisposition to myeloid malignancies” as a provisional category (Obrochta and Godley 2018; Khoury et al. 2022). The International Consensus Classification of Myeloid Neoplasms and Acute Leukemia, a group of individuals with expertise in both clinical care and pathologic/genetic aspects of these disorders, further enhanced awareness by describing the clinical characteristics and natural history of a number of HMs that present in childhood or adulthood and are associated with underlying GPSs (Arber et al. 2022; Rudelius et al. 2023). Moreover, the National Comprehensive Care Network (NCCN) guidelines now recommend germline testing for all patients with myelodysplastic syndrome (MDS) under the age of 40. Aligned with these changes, it is imperative that treating physicians consider GPS in any child or young adult newly diagnosed with an HM and pay close attention to blood abnormalities and/or constitutional symptoms/signs that might be indicators of an underlying GPS.

When evaluating patients with HMs for an underlying GPS, many factors play a role in the risk assessment and differential diagnosis, including physical examination findings, antecedent laboratory findings, features identified on bone marrow (BM) aspirate/biopsy, and family history. In this section, we highlight several of the key features that should raise suspicion for an underlying GPS and warrant referral for genetic counseling/testing. However, it is important to note that not all GPSs associated with HM have secondary clinical features. In addition, incomplete penetrance and variable expressivity may mask awareness of a GPS in a family, even if the index case has many features of the germline disease. Conversely, an individual found to have a GPS may themselves lack the constitutional findings of that syndrome due to variable expression, making their presentation less obvious. Therefore, it is important to remember that the absence of the features described below does not rule out an underlying GPS and germline genetic testing may still be warranted.

Due to variability in phenotype and presentation for many of the GPS, we recommend performing a comprehensive medical history and physical examination paying particular attention to the features described in Table 1. Additional intake questions may be necessary depending on the differential diagnosis for a given patient.

Table 1.

Personal/family history and/or physical examination findings that raise suspicion of an underlying genetic predisposition syndrome (GPS)

Clinical history or examination findings	Genes and/or syndrome associated with clinical finding
Café-au-lait spots	Neurofibromatosis type 1 (NF1), Fanconi anemia (FA), constitutional mismatch repair deficiency (CMMRD), RASopathies
Thrombocytopenia (isolated) or bleeding disproportionate to platelet count	RUNX1-familial platelet disorder with associated myeloid malignancies, ANKRD26-related thrombocytopenia, ETV6 thrombocytopenia and predisposition to leukemia
Pancreatic insufficiency	Shwachman–Diamond syndrome (SDS)
Early graying	Telomere biology disorders (TBDs)
Dysplastic nails	TBDs
Ataxia	Ataxia–telangiectasia (ATM), ataxia–pancytopenia (SAMD9L)
Hearing loss	GATA2 deficiency, SRP72
Skeletal anomalies	FA, MIRAGE syndrome (SAMD9), Diamond–Blackfan anemia (DBA), SDS
Neutropenia	SDS, ELANE-related neutropenia
Short stature	SDS, FA, MIRAGE, ataxia–pancytopenia, DBA, RASopathies, TBDs
Pulmonary fibrosis	TBDs
Warts (many, challenging to treat)	GATA2 deficiency
Recurrent infections, especially viral, opportunistic infection, or atypical bacteria	GATA2 deficiency
Lymphatic abnormalities/hydrocele	GATA2 deficiency
Squamous cell carcinomas of the head and neck, gastrointestinal (GI) tract, and other solid organ cancers	FA, TBDs
Early-onset solid tumors	Li–Fraumeni syndrome (TP53), CMMRD

Nonmolecular Tests Used in the Diagnostic Workup of Patients with Suspected GPS

Clinical testing performed at the time of HM diagnosis may provide additional clues as to the risk for a GPS and raise suspicion of an underlying predisposition. A preceding history of unexplained cytopenias or elevated mean corpuscular volume (MCV) may suggest marrow dysfunction associated with a GPS. In conjunction with an abnormal complete blood count (CBC), liver function tests may aid diagnosis in some GPSs like a telomere biology disorder (TBD) or Fanconi anemia (FA). Elevation of hemoglobin F may also be suggestive, although can be seen in the context of myeloid malignancy, and thus should be interpreted with caution (Reinhardt et al. 1998; Kudo et al. 2000; Choi et al. 2002; Alter et al. 2013; Avagyan and Shimamura 2022). Qualitative and quantitative platelet defects resulting in menorrhagia and/or easy bruising/bleeding may be a result of germline abnormalities in RUNX1, ANKRD26, and ETV6 (Marquez et al. 2014; Hock and Shimamura 2017; Sood et al. 2017; Avagyan and Shimamura 2022). Screening for immunologic abnormalities, by testing lymphocyte subsets for T, B, and natural killer (NK) populations as well as levels of IgA, IgG, and IgM, may increase suspicion for germline GATA2 deficiency (Spinner et al. 2014; Rossini et al. 2022). This transcription factor, which is crucial for the maintenance function of CD4⁺ T cells, B cells, monocytes, and NK cells may impair the function of all subgroups, although, interestingly, patients with GATA2 deficiency tend to have HM restricted to the myeloid lineage (Spinner et al. 2014; Collin et al. 2015; Bruzzese et al. 2020; Rossini et al. 2022).

More specific diagnostic studies may be prudent and include those described below. As both chromosomal breakage and telomere length can be impacted by chemotherapy, it is important to obtain these tests before treatment. If done following treatment start, a negative test may still be reassuring, although a positive result may be falsely concerning and merely reflect the impact of antecedent treatment, or, in the context of short telomeres, malignant cells that inherently have shorter length.

Chromosome Breakage

Chromosomal instability syndromes are a group of inherited disorders associated with genomic instability and breakage either spontaneously or in response to DNA-damaging agents. Two of these disorders, FA and Nijmegen breakage syndrome, may have, when exposed to the DNA cross-linking agents mitomycin C (MMC) or diepoxybutane (DEB), increased rates of breakage and form chromosomal abnormalities like radials. It is important, however, to know the limitations of this testing. Peripheral blood (PB) testing can be falsely negative as somatic reversion in hematopoietic stem cells (HSCs) can occur, mitigating the impact of DNA-damaging agents on cells derived from the BM. Skin fibroblast testing is, therefore, a gold standard for this evaluation and should be performed if the results on blood are normal but high suspicion of an underlying syndrome remains (Fargo et al. 2014). Moreover, as exposure to recent chemotherapy may also increase breakage, the risk for false-positive testing in that context must be recognized (Gutierrez-Rodrigues et al. 2023). The presence of increased chromosomal breakage with MMC or DEB, without definitive identification of a genetic abnormality, is sufficient to make the diagnosis of FA or Nijmegen breakage.

Telomere Length

Flow-FISH telomere length measurements are used to evaluate for dyskeratosis congenita and other TBDs. Fluorescent signals from telomeres of PB cell subsets are assessed on a single-cell basis by flow cytometry. Depending on the laboratory used and test selection, results can be reported for lymphocytes, granulocytes, B cells, naive and memory T cells, and NK cells; alternatively, only lymphocytes and granulocytes may be tested although the former 6-cell subset testing is the gold standard. Telomere lengths must be analyzed in relation to age-matched controls. Short telomeres in granulocytes are a nonspecific finding and are not diagnostic of a TBD. For patients with TBDs, the results often show low or very low lengths in multiple lymphocyte subsets. Length less than the first percentile in at least three different lymphocyte subsets is very sensitive and specific for TBD; results between the first and tenth percentile may also be suggestive in certain clinical contexts. Lengths >10% are unlikely to reflect a TBD; however, telomere length may be less sensitive for diagnosing a TBD in adults (Gutierrez-Rodrigues et al. 2023). Telomeres may also be shorter in patients with immunologic conditions. Telomere length testing for TBD has not been validated in patients with peripheral blasts and should be interpreted with caution. Moreover, as referenced above, chemotherapy also shortens telomere length, and/or malignant marrow cells often inherently have shorter telomeres. Thus, the interpretation of telomere length must be considered in the context of the patient’s clinical status and prior treatment.

Pancreatic Trypsinogen and Pancreatic Isoamylase

Serum pancreatic trypsinogen and isoamylase are often low in patients with Shwachman–Diamond syndrome (SDS); however, the age of the patient is important in the interpretation of the laboratory results. Trypsinogen is generally low in patients with SDS before the age of 3 and then increases to the normal range in older patients. Serum isoamylase levels are low in all patients with SDS; however, the utility of this test in patients younger than 3 years old is limited since isoamylase levels are also low in healthy children early in life. A very low fecal elastase may indicate exocrine pancreatic dysfunction; however, normal fecal elastase levels do not rule out SDS (Burroughs et al. 2009).

Erythrocyte Adenosine Deaminase

Elevated levels of erythrocyte adenosine deaminase (eADA) raise suspicion for Diamond–Blackfan anemia (DBA). However, it is important to be aware that up to 16% of patients with DBA may have normal eADA levels (Fargo et al. 2013). Elevation of eADA may also be seen with reticulocytosis or during the recovery phase from anemia.

Bone Marrow Analysis

Analysis of the BM may provide additional clues. While dysplasia in the myeloid lineage on morphologic review may be seen with idiopathic de novo myeloid disease, some GPSs demonstrate characteristic dysmorphology at baseline that does not indicate MDS but may be a clue to the underlying germline condition. For example, characteristic dysmyelopoiesis is present at baseline in SDS, dyserythropoiesis is seen in FA, and dysmegakaryopoiesis in germline RUNX1, ANKRD26, ETV6, and GATA2 (Li and Bledsoe 2023). In the context of MDS, pediatric disease may often present as a hypocellular marrow, which is in stark contrast to the hypercellularity typically seen in adult MDS. Additional morphologic features, including but not limited to the identification of micromegakaryocytes, increased reticulin fibrosis, and evidence of ring sideroblasts by iron stain, may also reflect marrow dysfunction or MDS that preceded an HM (Li and Bledsoe 2023).

Fluorescent in situ hybridization (FISH) as well as conventional cytogenetic analysis (CCA) of the BM also provide additional insights and should be performed in all new diagnoses of HM. CCA, which requires considerable manpower for analysis, is limited by the number of evaluable cells analyzed and what can be seen under a microscope. FISH analysis further facilitates the identification of cytogenetic fusions given the use of specific DNA probes. It can also identify copy number variants (i.e., submicroscopic imbalances like deletions and duplications) of a specific chromosome (Levy and Burnside 2019). In contrast to adult MDS, which is typically hypercellular, pediatric MDS is often hypocellular, so FISH is recommended even when cytogenetics are normal. Interpretation of FISH, however, must also be considered in clinical context. For example, in HM, the population of abnormal clonal cells may have a growth disadvantage compared to healthy cells. Interphase FISH testing, which does not require cell culture, is, therefore, prudent in HMs even if karyotypic testing is abnormal as it allows for the study of nondividing cells thus facilitating the identification of minor abnormal clones and/or small chromosomal deletions not detected by CCA at the DNA level (Flactif et al. 1994; Andreasson et al. 1997; Westbrook et al. 2000; Rigolin et al. 2001). Metaphase FISH, which does require culture to ensure cells are synchronized inmetaphase, aids inpositional information by allowing direct visualization of the chromosomes and the exact position of the signal probes on a given chromosome (Rigolin et al. 2001).

Complete or partial loss of chromosome 7 is the most common abnormality associated with germline conditions in pediatric and young adult HM and should prompt consideration of additional germline work up. Monosomy 7/deletion 7q is seen in approximately one-third of MDS or acute myeloid leukemia (AML) cases associated with FA and SDS and is particularly common in MDS/AML occurring in patients with germline mutations in GATA2, SAMD9, or SAMD9L. It has also been reported in HMs developing in patients with severe congenital neutropenia (SCN), RUNX1-familial platelet disorder (FPD), TBDs, and germline ERCC6L2, among others (Wlodarski et al. 2018). Trisomy 8 can also be seen with GATA2 deficiency and gains of chromosomes 1q and 3q have been reported in FA (Cioc et al. 2010; Wlodarski et al. 2016; Avagyan and Shimamura 2022; Li and Bledsoe 2023). A recent review by Li and Bledsoe (2023) summarizes additional cytogenetic abnormalities seen in pediatric MDS/AML suggestive of an underlying germline disorder.

Importantly, FISH is restricted in its ability to detect microduplications as well as uniparental disomy and atypical rearrangements. While CCA may aid in the identification of some structural issues not identified by FISH, the use of next-generation sequencing (NGS) provides an opportunity for more comprehensive screening. This testing refers to a high-throughput technology in which billions of sequencing reactions occur, each of which is analyzed simultaneously in a given sample. In the context of DNA NGS, identification of clinically relevant point mutations, insertions, or deletion of a given gene is facilitated. NGS RNA panels (i.e., RNA-seq) are useful to identify chromosomal translocations/fusions and gene expression profiles.

When mutations are reported by NGS, a variant allele frequency (VAF), which reflects the ratio of mutant to wild-type gene, is reported. In the context of the identification of a germline genetic variant, a high VAF (i.e., >30%) is typically seen, although exceptions occur. Awareness regarding the sensitivity of a given assay to detect a rare germline mutation, and the risk of that gene to undergo genetic reversion to a normal state is imperative as we enhance our understanding of GPSs and their association with HM. Specifically, spontaneous genetic reversion or somatic rescue for a given germline abnormality may be seen (Revy et al. 2019). This results in cells that either lack the germline mutation due to somatic reversion or develop a compensatory somatic change elsewhere in the germline mutated gene or in a related gene, resulting in functional correction of the germline defect that may then possess a growth advantage that can aid normal hematopoiesis and thereby mask detection of the inciting genetic variant (Waisfisz et al. 1999; Gross et al. 2002; Peffault de Latour and Soulier 2016). Of note, such somatic rescue of germline variants may either be adaptive (i.e., not associated with a risk for development of neoplasia) or maladaptive (i.e., increasing risk for HM). Given this potential for somatic reversion or rescue in BM-derived cells, germline testing of hematopoietic samples may not identify the germline variant and skin fibroblast testing is essential.

Similar to characteristic cytogenetic abnormalities like monosomy 7, specific NGS abnormalities in myeloid neoplasia, particularly in pediatric patients, may provide a clue to an underlying germline condition. SDS is associated with frequent somatic mutations in EIF6, and in MDS TP53 mutations. MDS in FA frequently harbors RUNX1 or RAS pathway mutations. MDS occurring in TBDs may have TP53, TET2, or ASXL1 mutations. MDS occurring in SCN frequently harbors CSF3R with or without RUNX1 mutation. GATA2-related MDS often harbors STAG2 or ASXL1 mutations (Li and Bledsoe 2023).

Germline Genetic Testing Modalities, Interpretation, and Limitations

Options for clinical germline genetic testing have evolved significantly in the last several years. Previously, testing was largely single gene or small panels, which limited the identification of genetic abnormalities associated with rare genetic syndromes. More recently larger panels that screen for several genetic conditions concurrently provide greater potential for identification of a GPS, many of which are rare and may lack associated clinical features. With the expanding options for clinical genetic testing, it is important for providers to be aware of the benefits and drawbacks of each testing approach to ensure that the optimal test is selected for a given patient in the context of their personal medical history, family history, and familial preferences regarding germline testing. Awareness of testing limitations is also prudent. For example, germline GATA2 mutations include several recurrent pathogenic variants that are deep in an intron and would be missed on any test that does not specifically target the intronic regions. ANKRD26 mutations are primarily within the promoter and can be missed with some testing modalities, and SBDS has a pseudogene that can make sequencing challenging and can lead to inaccurate results. When mutations in one of these genes are high on the differential diagnosis, it is important to evaluate the test being selected and verify with the laboratory that they have adequate coverage of all critical regions of the gene.

Awareness of atypical cell division events, which can result in aberrant gene expression is also important. For example, uniparental disomy occurs when an individual receives two copies of a chromosome, or part of a chromosome, from one parent and no copy from the other parent. This can manifest as heterodisomy, because of an error in meiosis 1, in which a pair of nonidentical chromosomes are inherited from one parent. Alternatively, isodisomy, in which a single chromosome from one parent is duplicated, due to an error in later-stage meiosis II, may also occur. Uniparental disomy is significant as it may result in large blocks of homozygosity, which can result in the expression of recessive genes with significant clinical implications. For example, isodisomy can lead to the expression of a pathogenic recessive phenotype given this chromosomal duplication. This aberration should be considered when a child has evidence of a rare recessive disorder and yet only one parent is identified as being a carrier for this genetic abnormality.

As referenced earlier, somatic genetic rescue (SGR) refers to compensatory mechanisms that may explain varied penetrance of a GPS in a kindred and/or impact the frequency of leukemic progression in these syndromes (Li and Bledsoe 2023). SGR manifests in two ways, either via genetic reversion or by the development of compensatory mechanisms that offset the deleterious phenotype (Schratz 2023). Copy neutral loss of heterozygosity (CN-LOH) is the most common cause of direct SGR in autosomal dominant disorders, leading to restoration of two wild-type gene copies by replacement of the chromosomal region harboring the germline mutation with a copy of the wild-type chromosome from the other allele. When reversion occurs, affected cells are functionally corrected and often possess a growth advantage that can aid normal hematopoiesis and potentially mask the detection of the inciting gene abnormality (Revy et al. 2019). This adaptive approach is seen in germline GATA2, FA, and TBDs (Schratz 2023).

Additional secondary adaptive mechanisms can also be protective. For example, in the context of SAMD9/SAMD9L GPS, more than 50% of individuals are found to have an SGR due to one of several mechanisms. The first, termed a “second site mutation” is the acquisition of a second SAMD9 or SAMD9L mutation in cis that offsets the effect of the germline mutation. Removal of the mutant allele through mitotic recombination that results in uniparental isodisomy of chromosome 7 (UPD7q) may also be adaptive as can focal gene deletions of chromosome 7 that impact the mutant allele. Conversely, aneuploidy resulting in monosomy 7 due to haploinsufficiency of additional genes on chromosome 7, can facilitate driver mutations that increase the risk of leukemogenesis, in individuals with this condition (Tesi et al. 2017; Wong et al. 2018; Sahoo et al. 2021; Gutierrez-Rodrigues et al. 2023; Schratz 2023). When suspicion of syndromes with a known risk for SGR is high (e.g., germline GATA2, SAMD9, SAMD9L, RUNX1, etc.) nonhematopoietic fibroblasts are the optimal sample type for genetic testing, for both the proband and their relatives. Lastly, patients with an underlying GPS may have both a germline and somatic mutation of the clinically relevant gene, and it is imperative to discern which one of the two mutations is germline (Avagyan and Shimamura 2022).

In the context of PB or BM DNA sequencing, germline variants typically have a VAF of 30%–50% or up to 100% depending on heterozygosity versus homozygosity status. When the VAF is <30%, it is more likely that the variant is somatic, although mosaicism, SGR, and technical limitations of the testing, such as allelic dropout, can lead to germline variants being reported at a lower-than-expected VAF. When a somatic gene panel identifies a variant that can be seen in germline disease at a mutation frequency of 30% or greater, consideration of follow-up germline genetic testing is prudent (Kraft and Godley 2020; Feurstein et al. 2022a,b; Gutierrez-Rodrigues et al. 2023). Secondary somatic mutations may provide additional clues to germline risk. For example, somatic mutations in PPM1D, POT1, and the TERT promoter can be associated with TBD; EIF6 mutations are seen in SDS, transient monosomy 7 is observed in germline mutations of SAMD9/SAMD9L, and secondary somatic DDX41 mutations may be seen in conjunction with the germline abnormality.

When germline testing is indicated, providers must carefully consider not only what genetic test to order for a patient with an HM but also the most appropriate sample type. For patients with acute leukemia who have negative MRD and no morphologic evidence of disease, germline genetic testing on blood or saliva/buccal samples can be considered. However, the blood and saliva of patients with a clonal HM that involves the BM may contain many lymphocytes, increasing the risk for classification of a somatic mutation as germline. In addition, in some instances, buccal swabs are associated with low DNA recovery, which precludes definitive results (Padron et al. 2018; Stubbins et al. 2022). Moreover, particularly in patients with a myeloid malignancy, there is the possibility of a reversion event in the blood and/or BM that can lead to a false-negative result in the blood/BM/saliva. Taken together, the ideal sample type for patients with any active acute or chronic leukemia or lymphoma with a history of BM involvement is DNA from cultured fibroblasts. Reliance on cultured fibroblasts for germline testing obviates the risk for contamination with somatic mutations derived from clonal hematopoietic cells. However, in patients with lymphoma that has not impacted the BM, blood or saliva can be considered. For patients with preceding allogeneic HSCT, cultured fibroblasts from an organ source outside of the BM is mandated. Of note, several centers are actively exploring the use of other sources of germline DNA, including DNA from fingernails and hair follicles.

With expanding options for genetic testing and interpretation, informing one’s patient regarding options for genetic testing has become more nuanced. Table 2 describes the most common molecular genetic tests available, and highlights scenarios when ordering of a given test is most appropriate. The pros and cons of each modality are reviewed and additional considerations that are warranted in their context are provided.

Table 2.

Molecular germline genetic testing panels

Type of test	When to order	Benefits	Drawbacks	Additional considerations
Targeted variant testing	Variant previously identified in family member(s), confirming germline status of variant identified on somatic testing	Low cost, easy-to-interpret results, no possibility of incidental findings	When confirming germline status of variant identified on somatic testing, the true germline mutation in the gene could be missed (e.g., the mutation identified could be the somatic second hit in a given gene that results in malignant transformation)	For recessive conditions, consider full gene analysis to rule out a second mutation that would impact risk assessment
Single gene	Phenotype highly consistent with a specific gene (e.g., neutropenia and abnormal pancreatic enzymes consistent with Shwachman–Diamond syndrome [SDS])	Potentially lower cost compared to larger panels or whole-exome sequencing (WES), good option for families who are concerned about “incidental” findings, typically covers relevant noncoding variants	If negative, may need to order a second more comprehensive germline screening test that can lead to delay in diagnosis and additional cost to family	Verify that copy number variation (CNV) analysis is done and, when relevant, known promoter/intronic variants included
Single syndrome	Phenotype highly consistent with a specific syndrome (e.g., patient with classic features of Fanconi anemia [FA] and abnormal chromosome breakage)	Potentially lower cost than larger panels or WES, good option for families who are concerned about “incidental” findings, typically covers relevant noncoding variants	If negative, may need to order a second more comprehensive germline screening test that can lead to delay in diagnosis and additional cost to family	Verify that CNV analysis is done and, when relevant, known promoter/intronic variants included
Multigene panel testing	Broad differential diagnosis	Evaluates multiple syndromes at once; has deeper sequencing of specific genes of interest compared to WES and typically covers relevant noncoding variants	Increased chance of incidental findings, more variants of uncertain significance (VUSs) identified	Verify that CNV analysis is done and, when relevant, known promoter/intronic variants included
Whole exome sequencing	Broad differential diagnosis, prior negative panel test(s) and suspicion for underlying syndrome remains	Submitting parental samples helps with analysis and provides insight into de novo versus inherited status; rapid WES is an option for patients who need results urgently for treatment (can be as fast as 7-day turnaround)	Does not cover noncoding regions, can miss deletions/duplications, analysis is phenotype driven/filtered (include on test requisition form if you have specific genes you want to make sure are analyzed and provide clear phenotype/clinical data), more expensive than panels	Once ordered, it can be harder to obtain insurance approval for a targeted panel test
Whole genome sequencing	Broad differential diagnosis, used when prior negative panel test(s) and suspicion for underlying syndrome remains	Covers noncoding regions, better deletion/duplication coverage than WES, can identify some structural variants, submitting parental samples helps with analysis and provides insight into de novo versus inherited status	Generally, has lower read depth/coverage than exome (might miss mosaic results), analysis is phenotype-driven/filtered (include on test requisition form if you have specific genes you want to make sure are analyzed and provide clear phenotype/clinical data), more expensive than panels	Many insurance companies do not cover WGS
Microarray (array comparative genomic hybridization [aCGH] or single-nucleotide polymorphism [SNP] array)	Patient with developmental delays, dysmorphic features, and multiple congenital anomalies; patient with whole gene deletion identified on other testing (microarray can help determine the extent of the deletion and if other genes are involved)	Good first line test for patients with multiple congenital anomalies with broad differential diagnosis	Can miss small deletions (e.g., single gene deletions or intragenic deletions); does not identify sequence variants	Can detect consanguinity and will report expected degree of relatedness of parents—recommend discussing this during consent conversation; unlike CGH arrays, SNP arrays can detect long contiguous stretches of DNA and uniparental disomies and copy-neutral loss of heterozygosity (LOH); SNP arrays can have higher resolution depending on the number and location of SNPs utilized

It is important for all providers ordering genetic testing to be familiar with interpreting the genetic results and understanding the difference between pathogenic/likely pathogenic variants and variants of uncertain significance (VUSs). In the HM space, many of the GPSs described are relatively newly discovered with limited cases reported in the literature, thus functionally significant mutations may be reported as VUSs on clinical reports. However, it is important to note that the majority of VUSs are later determined to be benign findings, having no impact on the health of the patient or their family members. Typically, VUSs are not used in clinical or treatment decision making, and cascade testing is not offered to family members unless it is likely to aid in the reclassification of the variant. Because of this, it is critical for providers to review the literature and available genetic databases that comprehensively annotate previously identified VUS to discern whether a given VUS has functional data or clinical attributes in affected individuals to support its pathogenicity. Where feasible, additional functional tests (e.g., chromosome breakage for patients with VUSs in FA genes, telomere length for patients with VUSs in TBD genes) or other evaluations that might aid the diagnosis of a GPS with a syndromic phenotype (e.g., echocardiogram for a patient with suspicious VUS in a RASopathy gene) can provide further clarity regarding the clinical relevance of a VUS.

OVERVIEW OF GPSs ASSOCIATED WITH PEDIATRIC HEMATOLOGIC MALIGNANCY

With the increased use of NGS testing at the time of HM diagnosis or in studies of families with high rates of HMs, the identification of genetic abnormalities that predispose to malignancy is increasing. It is important to remember that many of these GPSs may lack a preceding abnormal phenotype. Several studies have been conducted to discern the frequency of causative germline variants in HM (Trottier and Godley 2021). The Fred Hutch published a retrospective series of all patients who underwent HSCT at their center between 1990 and 2012 and found that 13.6% of patients <45 years old who were treated for MDS were found to have suspicious variants in known germline predisposition genes (Keel et al. 2016). Targeted panel testing identified GPS in 18% of adult patients referred to MD Anderson’s Hematologic Malignancy Program (DiNardo et al. 2016). A St. Jude cohort of 1120 pediatric cancer patients screened with whole genome, whole exome, and targeted sequencing found that germline mutations were seen in 8.5% of patients, 4.4% of whom were treated for leukemia. Importantly, only 23% of patients with an identified mutation had a family history of cancer predisposition (Klco and Mullighan 2021). We know that clinical penetrance of some GPS can be variable in kindreds and that some germline conditions result from de novo germline variants, particularly mutations seen in Diamond–Blackfan-associated genes, GATA2, SAMD9, the TINF2 subset of TBDs and TP53, given these mutational events arise early in embryogenesis (Gonzalez et al. 2009; Davidsson et al. 2018; McReynolds et al. 2018; Trottier and Godley 2021). Because of this, the absence of a suspicious family history does not obviate the need for germline testing. Knowledge of an underlying germline condition in an HM patient may impact one’s treatment approach to their malignancy and also provide awareness that allows families to make informed and personal reproductive and family planning decisions (Trottier and Godley 2021).

In Table 3, we have highlighted several GPSs that confer risk of both myeloid and lymphoid malignancy in pediatric and adult patients and have described the most common features associated with each condition. The table is separated into three distinct categories based on phenotypic presentation: (1) germline predisposition syndromes primarily associated with HMs not affecting multiple organ systems, (2) cancer predisposition syndromes with increased risk for HMs, and (3) BM failure syndromes. Category 1 encompasses syndromes where the primary phenotype is leukemia and other cancer risk or multiorgan involvement is rare. Syndromes in this category should be carefully considered for any patient presenting with a personal and family history of leukemia without a strong history of other cancers or additional medical history. Category 2 encompasses syndromes that are associated with a wide variety of cancers, including solid, central nervous system (CNS), and HMs. Syndromes in this category should be considered for patients with a personal history of a solid tumor and an HM as well as patients with a strong family history of cancer, particularly cancers diagnosed at younger than average age. Category 3 encompasses classically defined BM failure syndromes and syndromes that present with additional organ dysfunction.

Table 3.

Genes associated with genetic predisposition syndrome (GPS): clinical characteristics and hematologic malignancy (HM) presentation

Gene(s)	Syndrome	Inheritance pattern	Most common hematologic malignancies	Lifetime risk for hematologic malignancy	Other malignancy risks	Additional clinical features	Additional findings that support diagnosis	Additional comments	References
Germline predispositions primarily associated with hematologic malignancies not affecting multiple organ systems
DDX41	DDX41-associated familial myelodysplastic syndrome and acute myeloid leukemia	AD	Myelodysplastic syndrome (MDS)/AML, chronic myeloid leukemia (CML), lymphoid neoplasms	Reduced penetrance	N/A	Cytopenia	Somatic DDX41 mutations affecting the other allele are common, some have deletion on 5q35.3 leading to haploinsufficient DDX41 expression	Favorable prognosis; male predominance of HM (3:1); HM typically occurs in older adults	Polprasert et al. (2015), Li et al. (2022), and Tiong et al. (2023)
CEBPA	CEBPA-associated familial acute myeloid leukemia	AD	AML	Near complete penetrance with germline amino-terminal variant, reduced penetrance with germline carboxy-terminal variant	N/A	N/A	Germline amino terminus often accompanied by somatic carboxy-terminal mutation	Favorable prognosis; pros and cons of transplant should be discussed with the family given favorable prognosis but increased risk of new leukemia	Tawana et al. (2015, 2017)
ETV6	ETV6 thrombocytopenia and predisposition to leukemia	AD	B-cell ALL, MDS/AML	30%	N/A	Thrombocytopenia	Often hyperdiploid leukemia with loss of ETV6 wild-type (WT) allele; small hyperchromatic megakaryocytes, disseminated toxic granulations, and dysplastic eosinophils in the absence of frank myelodysplasia		Feurstein and Godley et al. (2017), Di Paola and Porter (2019), and Homan et al. (2023)
RUNX1	RUNX1-familial platelet disorder with associated myeloid malignancies	AD	MDS/AML, ALL, CML, lymphoma	25%–50%	N/A	Thrombocytopenia, allergic conditions, gastrointestinal issues	Somatic mutations often affecting the second RUNX1 allele and GATA2		Brown et al. (2020), Cunningham et al. (2023), and Homan et al. (2023)
ANKRD26	ANKRD26-related thrombocytopenia	AD	MDS/AML, CML, chronic lymphocytic leukemia (CLL)	<10%	N/A	Thrombocytopenia, qualitative platelet dysfunction	Deficiency of platelet α-granules, elevated thrombopoietin levels, dysmegakaryopoiesis		Kewan et al. (2020) and Homan et al. (2023)
PAX5	N/A	AD	B-cell ALL	Unknown	N/A	Neurodevelopmental disorder with developmental delay and/or autism spectrum disorder	Somatic deletion of WT allele		Shah et al. (2013) and Gofin et al. (2022)
KDM1A	N/A	AD	Multiple myeloma (MM)	Unknown	N/A	N/A			Wei et al. (2018)
Cancer predisposition syndromes with risk for hematologic malignancies
TP53	Li–Fraumeni syndrome	AD	ALL, AML, MDS, CML, lymphomas	4%–6%	Breast cancer, osteosarcoma, soft tissue sarcomas, adrenocortical carcinoma, choroid plexus carcinoma, brain tumors, and other solid tumors	N/A	Hypodiploid ALL	Consider whole body/brain magnetic resonance imaging (MRI) to evaluate for other malignancies before initiating hematopoietic stem cell transplant (HSCT); therapy-related myeloid malignancies including MDS and AML are common in patients with Li–Fraumeni syndrome (LFS) and often associated with poor prognosis with standard therapies and even allogenic HSCT; up to 20% de novo rate	Swaminathan et al. (2019) and de Andrade et al. (2021)
MBD4	MBD4 deficiency	AR	AML	Unknown, suspected high penetrance	Polyposis, uveal melanoma, and nervous system tumors	N/A	Hypermutator genomic signature, somatic DNMT3A mutations common		Sanders et al. (2018) and Palles et al. (2022)
MLH1, MSH2, MSH6, PMS2, EPCAM	Constitutional mismatch repair deficiency (CMMRD)	AR, majority inherited from parent with Lynch syndrome	Lymphoma (typically non-Hodgkin), ALL, AML	33%–42%	Colorectal polyposis, colorectal cancer, brain tumors, gastric, small bowel, pancreatic, endometrial, urinary tract cancers, and other solid tumors	Café-au-lait spots	Hypermutator genomic signature, MMR deficient	Consider whole body/brain MRI and colonoscopy to evaluate for other malignancies before initiating HSCT; consanguinity of the parents and/or homozygosity for a founder mutation is observed in more than 50% of individuals with CMMRD	Bakry et al. (2014), Abedalthagafi (2018), and Durno et al. (2021)
POT1	POT1 tumor predisposition	AD	CLL	3.6-fold increased risk for CLL	Melanoma, angiosarcoma (primarily cardiac), glioma, meningioma, and other solid tumors	N/A	Unknown		Speedy et al. (2016) and Herrera-Mullar et al. (2023)
Bone marrow failure syndromes
AR: FANCA, FANCC, FANCD1 (BRCA2), FANCD2, FANCE, FANCF, FANCG (XRCC9), FANCI, FANCJ (BRIP1), FANCL, FANCM, FANCN (PALB2), FANCO (RAD51C), FANCP (SLX4), FANCQ (ERCC4), FANCS (BRCA1), FANCT (UBE2T), FANCU (XRCC2), FANCV (REV7/ MAD2L2), FANCW (RFWD3), FANCY (FAAP100) AD: FANCR (RAD51) XLR: FANCB	Fanconi anemia	AR, AD, XLR	MDS/AML, T-ALL, CMML	>50% by age 50	Head and neck squamous cell carcinomas, anogenital squamous cell carcinoma, and liver tumors; for FANCD1/ BRCA2 subtype: additional cancer risks, including CNS tumors, Wilms tumors, other solid tumors	Bone marrow failure, short stature, skin hyper-/ hypopigmentation, skeletal, GU, neurologic, and heart abnormalities	Abnormal chromosome breakage; somatic RUNX1 mutations	Can experience chemosensitivty and risks from alkylating and DNA cross-linking agents	Alter (2014)
AD: TERC, TINF2, ZCCHC8, NAF1, RPA1, MDM4, NPM1 AR: NOP10, NHP2, WRAP53, CTC1, STN1, DCLRE1B, USB1 AD/AR: TERT, RTEL1, PARN, ACD, POT1 XLR: DKC1	Telomere biology disorders (dyskeratosis congenita)	AD, AR, XLR	MDS/AML, lymphoma	Unknown, genotype/phenotype correlations exist	Head and neck squamous cell carcinomas, gastrointestinal cancers	Bone marrow failure, leukoplakia, dysplastic nails, abnormal skin pigmentation, pulmonary fibrosis, premature graying, liver cirrhosis, arteriovenous malformation, skeletal anomalies, short stature, cytopenia, and immunodeficiency	Telomere lengths typically below tenth percentile for age in multiple subsets		Savage (2022)
AR: SBDS, DNAJC21, EFL1 AD: SRP54 (most are de novo)	Schwachman–Diamond syndrome	AR, ADAR, AD	MDS/AML	∼17%	N/A	Bone marrow failure, exocrine pancreatic dysfunction, neutropenia, short stature, skeletal anomalies (including chondrodysplasia and congenital thoracic dystrophy), hepatomegaly, cognitive and behavioral impairment	Somatic del(20)(q11) and i(7) (q10), somatic TP53 mutations, abnormal trypsinogen, and/or isoamylase levels		Furutani et al. (2022)
AD: RPL5, RPL9, RPL11, RPL15, RPL18, RPL26, RPL27, RPL31, RPL35, RPL35A, RPS7, RPS10, RPS15A, RPS17, RPS19, RPS24, RPS26, RPS27, RPS28, RPS29 XLR: GATA1, TSR2	Diamond–Blackfan anemia	AD, XLR	MDS/AML	Reduced penetrance, <15% by age 45	Osteosarcoma, colorectal cancer, other solid tumors	Bone marrow failure, short stature, cleft/lip palate, thumb, cardiac, craniofacial, and kidney anomalies	Elevated erythrocyte adenosine deaminase (eADA), macrocytic anemia with red cell aplasia		Vlachos et al. (2012, 2018)
ELANE	ELANE-related neutropenia	AD	MDS/AML	∼15%–36% within 15 years of initiating treatment with G-CSF	N/A	Congenital neutropenia, cyclic neutropenia, severe and recurrent infections	Monosomy 7, somatic mutations in CSF3R, and RUNX1		Rosenberg et al. (2008) and Dale et al. (2022)
AR: HAX1, G6PC3, CSF3R, VPS45, JAGN1 AD: GFI1, CXCR4 XLR: WAS	Severe congenital neutropenia (see above for ELANE-related severe congenital neutropenia)	AR, AD, XLR	MDS/AML	Unknown	N/A	Congenital neutropenia, cyclic neutropenia, severe and recurrent infections, osteopenia, seizures, developmental delays, genital, and heart abnormalities	Specific features are gene dependent		Fadeel et al. (2021)
GATA2	GATA2 deficiency (also referred to as monocytopenia and mycobacterial infection (MonoMAC)/dendritic cell, monocyte, B and NK lymphoid (DCML) deficiency	AD	MDS/AML, ALL, lymphoma, myeloproliferative neoplasm (MPN)	80% by age 40	N/A	Immunodeficiency, disseminated human papillomavirus (HPV) and mycobacterial infections, lymphedema, hearing loss, pulmonary alveolar proteinosis, and warts	Somatic monosomy 7 and trisomy 8 and somatic mutations in SETBP1 and ASXL1; B-, dendritic cell (DC)-, natural killer (NK)-, and mono-cytopenia, hypocellular myelodysplasia with loss of monocytes and hematogones, megakaryocytes with separated nuclear lobes, micromegakaryocytes, and megakaryocytes with hypolobated nuclei	High-risk disease with risk for life-threatening infections, warrants consideration of timely HSCT	Wlodarski et al. (2017), McReynolds et al. (2018), and Sahoo et al. (2020)
MECOM	N/A	AD	MDS/AML	Unknown	N/A	Congenital hypomegakaryocytic thrombocytopenia, radioulnar cynostosis, recurrent bacterial and fungal infections, deafness, renal defects, and congenital heart defects	B-cell deficiency, hypogammaglobulinemia		Germeshausen et al. (2018) and Lozano Chinga et al. (2023)
SAMD9	MIRAGE	AD	MDS/AML	Reduced penetrance	N/A	MIRAGE: myelodysplasia, infection, restriction of growth, adrenal hypoplasia, genital phenotypes, enteropathy, and cytopenia	Monosomy 7	Somatic hematopoietic rescue can occur leading to spontaneous hematological remission	Sahoo et al. (2020)
SAMD9L	Ataxia–pancytopenia	AD	MDS/AML	Reduced penetrance	N/A	Ataxia, pancytopenia	Monosomy 7	Somatic hematopoietic rescue can occur leading to spontaneous hematological remission; germline mutations are gain of function	Tesi et al. (2017) and Sahoo et al. (2020)
ERCC6L2	N/A	AR	MDS/AML, T-ALL	Unknown, suspected high penetrance	N/A	Neurologic abnormalities reported	AML-M6, Monosomy 7, complex karyotype with somatic mutations in TP53	Poor prognosis Warrants consideration of preemptive HSCT before progression to HM	Armes et al. (2022) and Baccelli et al. (2023)
MPL	Congenital amegakaryocytic thrombocytopenia (CAMT)	AR, AD	MDS/AML	Unknown	N/A	Thrombocytopenia, intracranial bleeds, bone marrow failure, and nervous system anomalies	Few or absent megakaryocytes	HSCT is only cure and is often recommended at a young age due to bleeding risks	Germeshausen and Ballmaier (2021)
RBM8A	Thrombocytopenia absent radius	AR	MDS/AML, ALL	Unknown, reduced penetrance	N/A	Bilateral absence of radii with the presence of both thumbs, thrombocytopenia, cow milk allergy, skeletal, heart, and genitourinary anomalies	Unknown	Unknown	Adam et al. (1993) and Jameson-Lee et al. (2018)
Other
N/A	Trisomy 21 (Down syndrome)	N/A	Transient myeloproliferative disorder (TMD), acute megakaryoblastic leukemia (AMKL), ALL	5%–30% risk of TMD, which develops into AMKL in 20% of individuals, 7–20-fold increased risk of childhood ALL	Increased risk for germ cell tumors, and testicular tumors	Hypotonia, short stature, intellectual disability, heart defects, hearing loss, and obstructive sleep apnea	Somatic GATA1 short and cohesin gene mutations		Xavier et al. (2009), Hasaart et al. (2021), and Osuna-Marco et al. (2021)
BLM	Bloom syndrome (sometimes referred to as Bloom–Torre–Machacek syndrome or congenital telangiectatic erythema)	AR	AML, ALL, non-Hodgkin lymphoma	Unknown	GI cancers, GU cancers, osteosarcoma, squamous cell carcinoma of skin, and other solid tumors	Extreme sun sensitivity, frequent infections, growth delay and short stature, endocrine and immune system abnormalities	Unknown	AML often occurs after chemotherapy for previous malignancy	Osuna-Marco et al. (2021)
PTPN11, NF1, CBL	RASopathies	AD	Transient myeloproliferative neoplasm in infancy, JMML	For germline PTPN11: ∼6% develop transient myeloproliferative neoplasm in infancy, ∼3% JMML	Risks are gene-dependent but can include brain tumors, neuroblastoma, rhabdomyosarcoma, and other solid tumors	Short stature, heart abnormalities, skeletal differences, distinct facial features, developmental delay		Children with JMML and germline predisposition have a high rate of spontaneous resolution of JMML	Strullu et al. (2014)

(AD) Autosomal dominant, (AR) autosomal recessive, (XLR) X-linked recessive, (AML) acute myeloid leukemia, (ALL) acute lymphoblastic leukemia, (JMML) juvenile myelomonocytic leukemia, (GI) gastrointestinal, (GU) genitourinary, (CNS) central nervous system, (G-CSF) granulocyte colony-stimulating factor, (WT) wild-type.

While Table 3 primarily focuses on genes associated with MDS or leukemia, it is important to recognize that there are several conditions that should be considered in individuals with a strong personal and family history of lymphoma. DNA repair disorders and primary immunodeficiencies are the most common syndromes associated with lymphoma risks. In addition, EBV⁺ lymphomas during childhood should raise the possibility of genetic predisposition either due to an immunologic condition or a GPS with immunologic dysfunction. Additionally, it is important to be aware that although only a small percentage of lymphomas occur in the setting of an underlying genetic predisposition, familial aggregation of lymphoma is well established, and risks are increased for individuals with a close relative with either Hodgkin lymphoma or non-Hodgkin lymphoma. First-degree relatives of NHL, HL, and CLL patients are estimated to have a ∼1.7-fold, 3.1-fold, and 8.5-fold elevated risk of developing NHL, HL, and CLL, respectively, and families should be counseled accordingly (Cerhan and Slager 2015).

Lastly, there are several genes, including TET2, CSF3R SRP73, and MECOM, that have recently been proposed as playing a potential role in germline genetic leukemia predisposition that is not included in the table. As the scientific discovery of GPSs associated with HM seems to be growing exponentially, we anticipate this list will continue to expand significantly.

TREATMENT CONSIDERATIONS IN CHILDREN WITH AN HM DIAGNOSIS AND UNDERLYING GPS

With the 5th edition of the WHO Classification of Hematolymphoid Tumors and with development of the 2022 International Consensus Classification of Myeloid Neoplasms and Acute Leukemias, HMs derived from germline predisposition were, for the first time, incorporated into HM classification systems (Arber et al. 2022; Khoury et al. 2022). With this heightened awareness, it is imperative that treating physicians apply their knowledge of the underlying GPS, when present, to identify the optimal treatment approach (Gurnari et al. 2023). In the case of chromosomal breakage disorders like FA, ataxia–telangiectasia, and Nijmegen breakage disorder, in which genes integral to DNA repair are mutated, one may be particularly sensitive to alkylators like cyclophosphamide as well as radiation. Patients with dyskeratosis congenita and other TBDs also fare poorly with alkylator treatment (Deeg et al. 1983; Gluckman et al. 1995; Rossini et al. 2022; Hudda and Myers 2023). With knowledge of an underlying GPS, definitive treatment may entail the use of an HSCT to eradicate the aberrant population of BM cells at risk for leukemic transformation. This is largely the case when a myeloid hematopoietic malignancy is found in a child with a GPS, even if the cytomolecular risk features of their malignancy would not support HSCT otherwise. In addition, individuals with GATA2 deficiency are at risk for serious infectious complications given their baseline immune dysfunction, prompting swift transition to HSCT for treatment consolidation when feasible, to minimize risk for infectious complications (Wlodarski et al. 2017; Gurnari et al. 2023). While HSCT is not recommended routinely in most children with a GPS and acute lymphoblastic lymphoma, such individuals may be prone to emergence of additional leukemic clones during or following completion of their initial leukemia therapy, thereby warranting close observation (McReynolds et al. 2022; Gutierrez-Rodrigues et al. 2023). Taken together, special considerations regarding treatment exist in individuals with an HM and underlying GPS and awareness of these nuances provides an opportunity for the advancement of disease-free survival with minimization of morbidity.

Toxicities following treatment may also occur at a higher frequency than the general population. Patients with TBDs are at risk for liver disease, pulmonary fibrosis, and arteriovenous malformations following treatment (Savage 2022). In addition, as a subset of GPSs are associated with both hematologic and solid tumors, the risk for secondary HMs following chemotherapy for a solid tumor may be heightened, warranting surveillance for this risk while also monitoring for solid tumor disease recurrence.

HSCT Considerations

For patients undergoing allogeneic HSCT with a related donor, knowledge of a GPS in the family can play a major role in donor selection. For a patient with a GPS, all potential family member donors should be tested for that GPS regardless of whether the family member has clinical features suggestive of the condition. When there is a strong suspicion for a GPS but no genetic diagnosis is identified, family donor evaluation may include a BM examination to assess for abnormalities. A donor with a known germline leukemia predisposition abnormality should not be used; the rationale for this stance extends beyond the obvious concern for risk for donor-derived malignancies following HSCT (Trottier and Godley 2021). For example, previous studies have demonstrated that, when related donors with deleterious germline variants such as CEPBA or RUNX1 or GATA2 are utilized, inferior HSCT outcomes are seen due to poor HSC mobilization of the donor, delayed or failed engraftment, poor immune function post-HSCT (Fogarty et al. 2003; Owen et al. 2008; Trottier and Godley 2021), as well as risk for leukemia development in the donor after stem cell mobilization/collection (Xiao et al. 2011; Kraft and Godley 2020).

In the absence of a suspicious family history in the patient undergoing a transplant, a comprehensive health history and screening CBC may be sufficient for the potential related donor. Universal NGS panel testing of donors has not become the standard of care given the ethics surrounding this approach. Moreover, widespread germline testing of donors may increase the potential for identification of VUSs that may, needlessly, delay HSCT putting the recipient at higher risk of recurrence and/or infection and create undue stress in a potential donor who has, to date, had no clinical stigmata of the GPS in consideration.

While universal donor testing is not recommended at present, treating physicians must be aware of the potential risk for the development of a donor-derived clonal population in an HSCT recipient. Two to five percent of leukemia relapses post-HSCT are donor derived (Williams et al. 2021). In this context, it is hypothesized that such conferred risk might reflect an unrecognized GPS in the donor versus donor clonal hematopoiesis that evolved a second hit in the recipient and resulted in disease. Cases of donor-derived malignancy after related and unrelated transplants suggest that pathogenic variants in genes such as CEPBA, DDX41, GATA2, and RUNX1 may contribute to poor HSCT outcomes (Fogarty et al. 2003; Owen et al. 2008; Williams et al. 2021). Although universal donor testing for GPS is not felt to be cost effective or ethical, the National Marrow Donor Program 2023 policy allows donors to opt in for notification, and ideally genetic counseling, if a genetic variant in a recipient is confirmed or suspected to be donor derived (Williams et al. 2021).

In the context of patients with GPS who need HSCT, DNA damage disorders like FA or a TBD have significantly inferior outcomes to those with ribosome biology disorders or disorders of hematopoiesis (Agarwal 2023; Hudda and Myers 2023). Moreover, TRM is reported to be 10% or higher in patients with a GPS (Agarwal 2023). Subsequent advances in HSCT approaches have strived to improve these outcomes by trialing alternative reduced-intensity conditioning regimens that include fludarabine or treosulfan and/or eliminate radiation (MacMillan and Wagner 2010; Peffault de Latour et al. 2015; Ebens et al. 2017; Mehta et al. 2017; Hudda and Myers 2023). A recent prospective multicenter trial of conditioning with treosulfan, an alkylating agent with a favorable toxicity profile, fludarabine, and, in some cases, rabbit antithymocyte globulin (ATG), in 14 patients with inherited BM failure without progression to HM reported 100% engraftment and 92% survival (Burroughs et al. 2017). This study is now expanded to see whether similar results are seen in a larger cohort (NCT04965597). Efforts are also ongoing to develop a similar trial of treosulfan conditioning for patients with known BM failure syndrome and subsequent HM.

With changes in HSCT conditioning, we are optimistic that outcomes for patients with GPS and secondary HM will improve, although the need to monitor for both disease recurrence and secondary malignancy is imperative. In addition, there are a number of post-HSCT complications that are of higher risk in patients with GPS that warrant consideration. Late effects in post-HSCT survivors, including but not limited to delayed immune reconstitution, iron overload, pulmonary complications, infertility, renal impairment, and growth failure, are more prevalent and more severe in a number of GPSs (Alter 2017). Detailed discussion of these risks and their clinical features, specific to individual GPSs seen in pediatrics, are well described in the 2017 blood article by Blanche Alter (Avagyan and Shimamura 2022). A recent American Society of Hematology (ASH) education review by Hudda and Myers (2023) provides a thorough review of post-HSCT monitoring recommended for some of the most common GPSs seen in children.

ADDITIONAL CONSIDERATIONS

Disease Surveillance in Individuals with a Known GPS

The frequency of disease monitoring in patients with a GPS, both those with and without a history of HM, is not clear. Recommendations are informed by the magnitude of the malignancy risk, outcomes of malignancy, and disease-specific patterns of progression to malignancy. For conditions with poor survival, once malignancy develops, surveillance offers an opportunity to intervene with HSCT before progression to frank malignancy. For example, disease progression has been studied longitudinally in SDS, a GPS with a high risk of TP53-mutated myeloid malignancy and poor survival once malignancy develops (Myers et al. 2020). Although small heterozygous TP53-mutated clones develop early in life and may persist stably for years, the acquisition of biallelic TP53 mutations in the marrow is a harbinger of malignancy in SDS (Kennedy et al. 2021; Reilly and Shimamura 2023).

As most GPSs lack longitudinal data to guide practice approaches, the present standard of care is largely informed by expert opinion and/or extrapolated from other hereditary hematopoietic disorders (Godley and Shimamura 2017; Porter et al. 2017; Roloff et al. 2021). A baseline BM evaluation at the time of diagnosis of a GPS is generally recommended, before HM development. Marrow examinations allow assessment for early signs of clonal evolution including dysplasia, cytogenetics, FISH, and somatic mutation analysis. Patients are then generally followed every 3–6 months with clinical evaluation and CBC, with consideration of annual BM examinations for conditions with high leukemia risk. Should cytopenias develop in one or more lineages, repeat CBC in 2–4 weeks is prudent with a low threshold to repeat BM testing if there is ongoing clinical concern. Data are lacking on the utility of molecular surveillance of the PB, although this is increasingly used as a screen for clonal progression in the absence of serial BM evaluations. If a somatic panel shows evidence of clonal evolution in the PB, additional marrow assessment may be warranted (Porter et al. 2017). If concerning clonal abnormalities or dysplasias are noted in the BM without frank malignancy, more frequent marrow testing using multiple modalities to assess for disease progression (i.e., FISH, karyotype, NGS somatic panel) is recommended (Roloff et al. 2021).

Given there are inherent differences in the natural history of acute lymphoid versus myeloid malignancies, decisions regarding the type and frequency of disease surveillance may vary. For example, due to the rapid onset of NHL, ALL, and some subsets of AML, there is limited data that PB and/or BM surveillance facilitates early detection of HM that translates into improved outcomes. However, as some heritable forms of AML and MDS can be more indolent, disease progression has been detected with serial monitoring for PB cytopenias and potentially surveillance BM evaluations (Porter et al. 2017; Förster et al. 2023). Of note, blood counts alone have been demonstrated to be an insensitive marker of impending myeloid malignancy in patients with SDS, so regular marrow examinations are recommended in that context (Myers et al. 2020; Reilly and Shimamura 2023).

Regarding risk for HM recurrence, the initial frequency of evaluations should be done as per the standard monitoring proposed for patients with de novo HM including, at a minimum, visits, and laboratory work done at regular intervals until 3 years from the end of treatment. At that time, many centers transition to yearly evaluations. However, given the underlying GPS and potential risk for a new HM, more frequent monitoring may be prudent and should be tailored to align with general surveillance guidelines for a given GPS as referenced above (Porter et al. 2017). Monitoring for disease post-HSCT can also introduce confounders. While surveillance with serial CBC is reasonable, some providers feel that molecular analysis in the setting of full donor chimerism should be avoided as such testing may result in the identification of gene variants derived from the donor, which may raise undue concern in transplanted individuals (Alter 2017; Trottier and Godley 2021). However, given the risk for recurrence of a primary HM as well as the potential for secondary hematopoietic malignancy in a patient with a GPS, awareness of the significance of donor mutations may be beneficial, and practice variations regarding disease monitoring exist as a result.

Cascade Testing and Reproductive Considerations in Patients with a GPS

Patients with GPS should be educated about the risks for their family members, including a discussion of the inheritance pattern, penetrance of HM, variability in phenotype, and surveillance that will be recommended for any family members who have the mutation. Physicians should collaborate with genetic services at their institution to coordinate cascade testing for family members or find a local certified genetic counselor specializing in hereditary cancer risks through the National Society of Genetic Counselors (NSGC) via the online genetic counselor tool at findageneticcounselor.nsgc.org.

Patients should also be made aware of various reproductive options, including natural conception with prenatal genetic testing through amniocentesis or chorionic villus sampling, natural conception with postnatal genetic testing, egg, or sperm donation, in vitro fertilization with preimplantation genetic testing (IVF with PGT), and adoption. Patients interested in learning more about PGT as well as other reproductive options should be referred to a prenatal genetic counselor or maternal–fetal medicine provider. For patients with autosomal recessive conditions, it is important to discuss the value of testing one’s partner to better understand risks for offspring.

SUMMARY

In summary, as awareness of genetic predisposition to HM expands, close collaboration between families, physicians, genetic counselors, scientists, and bioinformatics experts is critical. Diagnostic evaluation of GPS integrating clinical history, family history, physical examination, laboratory testing, and consideration of germline genetic testing for individuals with an HM has been outlined. Diagnosis of a GPS impacts decisions regarding chemotherapy treatment, the role of HSCT, as well as HSCT donor considerations. Moreover, such testing may also raise awareness as to the benefit of disease surveillance in other family members with the GPS. Lastly, as more children and adults with GPS are identified, there is a growing need for evidence-based guidelines regarding disease surveillance. As health care providers integrate the diagnostic workup for a GPS into the evaluation, monitoring, and treatment of HM patients, we anticipate greater survival of HM patients with GPS and less treatment-related morbidity/mortality.