Germline predisposition to hematopoietic malignancies

Abstract

Once thought to be exceedingly rare, the advent of next-generation sequencing has revealed a plethora of germline predisposition disorders that confer risk for hematopoietic malignancies (HMs). These syndromes are now recognized to be much more common than previously thought. The recognition of a germline susceptibility risk allele in an individual impacts the clinical management and health surveillance strategies in the index patient and relatives who share the causative DNA variant. Challenges to accurate clinical testing include a lack of familiarity in many health care providers, the requirement for DNA samples that reasonably approximate the germline state, and a lack of standardization among diagnostic platforms as to which genes are sequenced and their capabilities in detecting the full range of variant types that confer risk. Current knowledge gaps include a comprehensive understanding of all predisposition genes; whether scenarios exist in which an allogeneic stem cell transplant using donor hematopoietic stem cells with deleterious variants is permissive; and effective means of delivering genetic counseling and results disclosure for these conditions. We are hopeful that comprehensive germline genetic testing, universal germline testing for all patients with an HM, universal germline testing for allogeneic hematopoietic stem cell donors, and the development of preventive strategies to delay or even prevent malignancies will be available in the near future. These factors will likely contribute to improved health outcomes for at-risk individuals and their family members.

Introduction

A historical perspective on hereditary cancer syndromes and hereditary hematopoietic malignancies (HHMs)

The first publication describing the clustering of HMs likely dates back to an 1861 case report from Biermer (Fig. 1). This report describes a healthy woman with three children, each of whom had splenomegaly and died within the first 3 years of life (1). The woman subsequently re-married and had four children with her second husband, three of whom also developed splenomegaly, including one who died with acute leukemia (AL) documented at autopsy (1). In the first large scale study investigating familial leukemia in 1947, Videbaek demonstrated that 8.1% of 209 probands with leukemia possessed a family history of leukemia (2). However, this publication did not provide any pedigrees demonstrating a clear Mendelian inheritance pattern. Intriguingly, Videbaek also noted that a number of the families experienced a higher risk of solid malignancies (2). In 1975, Gunz et al. (3) studied 909 individuals with leukemia and showed their first-degree relatives experienced a 2.8–3-fold increased risk of leukemia. This risk was driven by excess cases of chronic lymphocytic leukemia (CLL) and AL, but no excess cases of chronic granulocytic leukemia (now referred to as chronic myeloid leukemia [CML]) were observed (3). The proportion of probands with a family history of leukemia in the Gunz study (7.9%) was similar to the original Videbaek’s study (8.1%). The group noted that multiple families experienced excess risk for both lymphomas and leukemias, but not an excess risk for solid tumors, and thus their work serves as the first published pedigree with Mendelian inheritance of ‘pure’ HMs (3). This family included a male proband with CML, a maternal aunt with acute leukemia and 8 maternal relatives with acute leukemia (3).

Figure 1

Timeline of discovery of genes that predispose to the development of hereditary hematopoietic malignancies (HHMs). The first description of an HHM in the medical literature appeared in 1861, and the first molecular description of an inherited gene mutation was published in 1999. The application of next-generation sequencing to patients and families has accelerated disorder discovery in the past several years.

Work during the 20th century from Drs. Lynch, Li, Fraumeni, King and their colleagues further established the molecular basis for hereditary cancer syndromes (4–13). The pioneering work of Dr Henry Lynch in the early 1960s established the hereditary basis for hereditary non-polyposis colorectal cancer, now termed Lynch syndrome (LS) (4). Li and Fraumeni’s seminal publication (5) similarly laid the groundwork for Li-Fraumeni syndrome (LFS). Both of these syndromes, LFS in particular, are notable for a predisposition to HMs. Li and Fraumeni’s initial report, for example, included four families. Two of these pedigrees were notable for cases diagnosed with acute lymphocytic leukemia (ALL) and acute myeloid leukemia (AML) (5). The risk for HMs in LFS has been further strengthened since Li and Fraumeni’s first study (6). Multiple case series have also been published linking LS-associated mutations with HMs, with lymphoma in particular being seen in multiple kindreds, despite HMs not being frequently considered to be a LS-associated tumor (7–9). Dr King, her colleagues and other groups subsequently demonstrated the genetic basis for hereditary breast and ovarian cancer driven by germline mutations in BRCA1 or BRCA2 (10–13). Deleterious germline genetic variants in BRCA1 and BRCA2 have since been associated with leukemogenic risk in both murine models and humans (14–16).

This work ultimately informed the discovery of the first HHM syndromes, starting with familial platelet disorder with associated myeloid malignancy (FPDMM, OMIM 601399; also known as hereditary thrombocytopenia and hematologic cancer predisposition syndrome (17), MONDO:0011071) driven by germline mutations in RUNX1, which encodes a master hematopoietic transcription factor (TF), in 1999 (18). The FPDMM phenotype has been expanded over time to include two phenocopies driven by germline mutations in ANKRD26 or ETV6, the latter of which encodes another TF critical for hematopoiesis (19–22). Germline mutations in CEBPA, which encodes for the CCAAT/enhancer-binding protein alpha TF, were subsequently shown to drive leukemogenesis in ‘pure’ HHMs in 2004 (23). Since these initial discoveries, subsequent work has implicated numerous other genes in HHM-associated syndromes, including but not limited to: GATA2, DDX41, SAMD9/SAMD9L and CSF3R (24–37).

Highlights from the Past 2 Years

With the use of next generation sequencing (NGS) applied to individuals with a personal history of multiple cancers or onset of disease at an earlier age than expected and a family history suggestive of germline predisposition, there has been an increase in recent years in the identification of germline predisposition disorders that confer risk for HMs (Fig. 2) (38–46). In the following section, we outline some of the most recent discoveries and key aspects of clinical testing that are imperative for accurate diagnosis for these conditions.

Figure 2

The paradox of rare disorders resulting in a common phenomenon. The riddle, ‘How much does a ton of feathers weigh?’ conveys the paradox that objects with little weight can combine to weigh a lot if there are enough of them. Likewise, each germline predisposition disorder that confers risk for HMs is seemingly rare in the population, but so many are now recognized that the phenomenon of germline risk can be considered to be relatively common overall. Standard gene names are used for each gene in which deleterious variants are known to confer risk for: myeloid malignancies (light blue feathers), including those causing Shwachman Diamond Syndrome (i.e. SBDS) and Shwachman Diamond Syndrome-like disorders (e.g. DNAJC21, EFL1 and SRP54); lymphoid malignancies (salmon feathers); immunodeficiencies (yellow feathers); hematopoietic [both myeloid and lymphoid] malignancies (green feathers); both hematopoietic and solid tumors (red feathers); and additional genes grouped as: BMF/DKC, bone marrow failure/dyskeratosis congenita [ACD, ADH5/ALDH2, ALAS2, CECR1, CTC1, CXCR4, DKC1, EFTUD1, ELANE, ERCC6L2, G6PC3, GATA1, GFI1, HAX1, LIG4, MDM4, NHP2, NOP10, PARN, POT1, RBM8A, RPL5, RPL11, RPL15, RPL18, RPL23, RPL26, RPL27, RPL31, RPL35, RPL35A, RPL36, RPS7, RPS10, RPS15A, RPS17, RPS19, RPS24, RPS26, RPS27, RPS28, RPS29, RTEL1, TERC, TERT, TINF2, USB1, VPS45, WRAP53]; FA, Fanconi anemia [BRCA1/FANCS, BRCA2/FANCD1, BRIP1/FANCJ, ERCC4/FANCQ, FANCA, FANCB, FANCC, FANCD2, FANCE, FANCF, FANCG, FANCI, FANCL, FANCM, MAD2L2/FANCV, PALB2/FANCN, RAD51/FANCR, RAD51C/FANCO, RFWD3/FANCW, SLX4/FANCP, UBE2T/FANCT, XRCC2/FANCU]; HBOC, hereditary breast and ovarian cancer [ATM, BARD1, BRCA1/2, BRIP1, CDH1, CHEK2, EPCAM, MLH1, MSH2, MSH6, NBN, NF1, PALB2, PMS2, PTEN, RAD51C, RAD51D, STK11, TP53]; and LS, Lynch syndrome [EPCAM, MLH1, MSH2, MSH6, PMS2, POLD1, POLE, RPS20].

Emerging new bone marrow failure syndrome: combined ADH5/ALDH2 deficiency

In 2020, two studies described a new bone marrow failure (BMF) syndrome in seven families of East Asian ethnicity, characterized by homozygous or compound-heterozygous variants in ADH5 in combination with the defective ALDH2 rs671 allele in the homozygous state (A/A) (47,48). This combined ADH5/ALDH2 deficiency leads to formaldehyde accumulation, overwhelming the DNA-repair capacity with increased levels of DNA damage. A subsequent study was able to show that patient-derived induced pluripotent stem cells displayed drastically defective cell expansion when stimulated into hematopoietic differentiation in vitro (49). All identified patients had a Fanconi anemia-like phenotype with childhood-onset BMF, changes in skin pigmentation, low birth weight, short stature and intellectual disability with no increased chromosome fragility/breakage or abnormalities of the thumb or radius (47,48,50). Therefore, we recommend that ADH5/ALDH2 should be part of screening for BMF syndromes, in particular in patients with East Asian ethnicity with a high population prevalence of ALDH2 rs671.

Complex spectrum of variants associated with germline predisposition

Truncating single-nucleotide variants, such as frameshift, nonsense and canonical splice variants leading to a truncated protein or absent protein via nonsense mediated decay have long been described and are the mechanism of disease in many germline predisposition syndromes. Likewise, missense variants causing loss-of-function, gain-of-function, or dominant-negative effect over the wild-type allele have been well described in genes such as ETV6 (51,52), GATA2 (53), RUNX1 (54), TP53 (55) and many others. The spectrum of variants and mechanism of disease has broadened over the past few years, which affects both the sequencing approach as well as the analysis and interpretation of the identified variants.

Copy number analysis has not been performed widely until increasing numbers of studies have shown that copy number variants (CNVs) are common causative alleles of germline predisposition syndromes. CNVs accounted for 10% of causative variants in young adults with myelodysplastic syndromes (MDS) or aplastic anemia diagnosed between the age of 18 to 40 (39), in 19% of patients with a telomere biology disorder (56) and in 17% of patients with hypocellular BMF syndromes (57). Whole-gene and intragenic deletions have been reported as segregating with disease in numerous families with RUNX1 and GATA2 germline variants (58–60), and can be easily overlooked by traditional NGS or whole-exome sequencing.

Recent functional studies have shown that a recurrent synonymous variant in GATA2, p.Thr117Thr and a missense variant, p.Ala286Val, cause disease by introducing a cryptic novel splice donor site, which in turn results in a truncated GATA2 protein (61–64). An in-frame nine amino acid insertion between the two zinc finger domains of GATA2 was reported to increase spacing between the two zinc fingers and proved to be defective in genetic rescue assays (65). Likewise, the RUNX1 c.968-10C>A variant was shown to generate a cryptic splice acceptor leading to a truncated RUNX1 protein (60), emphasizing the need to look beyond canonical splice variants or variants affecting known functional domains.

Two distinct RUNX1 germline translocations have been identified, with the predicted breakpoints of both separating the first promotor, which drives expression of RUNX1 isoform C, and the +23 enhancer from the rest of the gene (66,67). This emphasizes the importance of pursuing and publishing investigations beyond standard mutational testing if the family history and phenotype strongly suggest involvement of a particular gene. It will be very important for clinical laboratories to have nimble diagnostic testing platforms that are capable of detecting the variety of mutation types described in individuals and families with germline predisposition to HMs.

Although we recognize a variety of syndromes with strict Mendelian inheritance, additional risk factors may influence the phenotype seen in patients. For example, Karastaneva et al. (68) reported a pediatric patient with high-hyperdiploid ALL, persistent thrombocytopenia and germline ETV6 variant, who was also homozygous for the ARID5B risk allele (rs7090445-C) (69), each of which predispose to high-hyperdiploid ALL. Leukemogenic co-factors, such as the ARID5B risk allele or the GATA3 rs3824662 allele increasing susceptibility for Ph-like ALL (70), could explain incomplete penetrance and phenotypic heterogeneity seen in many germline syndromes.

Of note, a comprehensive list of genes is paramount for clinical testing and has to be adapted quickly based on the rapid addition of new genes/syndromes. Roloff et al. (71) reported that commercially available panels often miss a number of well-described germline syndromes, such as ANKRD26, DDX41, ETV6, SAMD9 and SAMD9L and may have decreased sensitivity for the detection of small CNVs. Unfortunately, busy clinicians who are not familiar with technical aspects of germline genetic testing may not realize the deficiencies of the platforms they are using and may mistakenly believe that they have ordered comprehensive testing. We look forward to academic and commercial laboratories providing clinical testing that clearly delineates its capabilities with reports that are easy to interpret for those without a sophisticated genetic background. Deficiencies of the technologies employed should also be outlined, with options for follow-up testing to ensure comprehensive testing.

Recognition of germline variants from ‘somatic’ tests on cancer cells

Somatic NGS panels are now routinely performed on malignant cells in blood and bone marrow to identify genetic variants important for risk stratification and treatment decisions, in genes such as NPM1, CEBPA, FLT3, RUNX1, ASXL1 and TP53 (72). These panels detect both somatic and germline variants, since germline alleles are present in all cells from an individual (Fig. 3). However, it is incumbent upon clinicians to recognize the potential germline nature of deleterious variants and pursue subsequent germline testing to test the germline status of such variants (72–75). DDX41 variants found on somatic AML/MDS NGS panels with a variant allele frequency (VAF) of >40% have been confirmed in germline in 94% of patients, particularly early truncating variants as well as germline variants common in particular populations, such as p.M1? and p.Asp140fs in Northern European populations and p.Ala500fs in Japanese/Korean populations (76). Importantly, the somatic hotspot variant p.Arg525His was identified in more than half of the patients with germline DDX41 variants (76). In general, patients with a ‘double-hit’ in one gene, in particular CEBPA, RUNX1 or DDX41 are likely to carry one variant in the germline at a VAF between 40 and 60%, with the second variant of somatic origin with variable VAF (77). In a pediatric AML study, 13% percent of variants identified in genes encoding TFs like CEBPA, ETV6, RUNX1, GATA2 and P53 were of germline origin (77–79). Simon et al. (80) reported that 27% of RUNX1 variants identified in RUNX1-mutant AMLs were of germline origin. However, the yield of deleterious germline RUNX1 variants is only 16% (81), when the variants are curated using the RUNX1-specific curation rules (17,82). This example emphasizes two important points: (1) not all germline variants are deleterious; and (2) accurate functional classification of germline variants is critical for variant interpretation and ultimately patient management. Moreover, germline frequencies may also depend on the populations being tested (83). Taken together, these recent studies suggest that germline testing should be performed when: (1) the variant is found at a VAF > 30%, suggesting its possible germline nature; (2) the variant is truncating and/or affects well-described germline hotspots; and/or (3) there is a ‘double-hit’ with one of the variants being within germline range. Special consideration is warranted when the variant occurs in genes such as CEBPA, DDX41, ETV6, GATA2, RUNX1, TP53 or variants in genes associated with telomere biology disorder (Table 1) (72–77,80,81,83–85).

Figure 3

Finding germline variants from tumor testing. Molecular profiling of tumor cells is now standard practice for many leukemias, and germline variants will be present in all non-germ cells from an individual. Obtaining informed consent that molecular testing of tumor cells can reveal germline variants is prudent, but adds complexity to testing. The identification of a deleterious variant in a gene known to confer inherited cancer risk should prompt consideration of its germline status. In our experience, such alleles are most often germline when the VAF is relatively high (75), and we prioritize patients who are found with such alleles at a VAF ≥ 30%. Once such alleles are identified, we recommend alerting the ordering provider and/or patient. Testing for the germline nature of the variant can be performed either by site-specific testing for the variant within DNA that reasonably approximates the germline state, which can be obtained from cultured skin fibroblasts, mesenchymal stromal cells, and/or hair bulbs, or via segregation within the family. We prefer the former approach, since patients with HMs return frequently to the clinic, and obtaining tissue from family members can be cumbersome, especially if they live out of state relative to the ordering provider. Red arrow identifies the affected individual in the pedigree, who is identified as a filled circle. Unfilled circles/squares, unaffected individuals; circles, females; squares, males.

Table 1

Somatic and germline testing to identify germline variants

	‘Somatic’ testing (DNA comes from malignant cells)	Proper germline testing (DNA comes from tissue reasonably approximating the germline*)
Covered genes	Largely defined already; platform is stable, but may NOT include all genes that confer germline risk.	Genes are still being discovered; platform needs to expand over time.
Detected variants	Prognostication is based on exonic variants; can use NGS panels easily; VAF should be within germline range (0.3–0.6). Caveat: Hematopoietic tissues undergo somatic reversion easily (especially SAMD9/SAMD9L and genes associated with bone marrow failure syndromes and immunodeficiency).	Variants in non-coding regions and CNVs are COMMON predisposition alleles; need NGS panels AND high-density microarrays or WES with spike-in probes covering non-coding regions of interest (e.g. TERC, 5′UTR of ANKRD26, and the intronic enhancer region of GATA2) with customized bioinformatic analysis to detect CNVs. Consider other techniques (e.g. FISH) if phenotype is highly specific for one gene/group of genes but mutational testing is negative.
Specific alleles	For some genes, (e.g. CEBPA, RUNX1 and TP53), the same allele can be a somatic or germline variant. ‘Double-hits’ with one allele with a VAF within the germline range are a good indicator for a germline predisposition syndrome (e.g. truncating DDX41 variant with VAF in germline range and p.Arg525His with a lower VAF).	For some genes (e.g. CHEK2 and DDX41), specific alleles have OVERWHELMINGLY been seen as germline alleles.

^*those approximating germline tissue: DNA derived from cultured skin fibroblasts (gold standard), cultured bone marrow-derived mesenchymal stromal cells, and/or hair bulbs.

Abbreviations: CNV, copy number variant; FISH, fluorescence in situ hybridization; NGS, next-generation sequencing; VAF, variant allele frequency; WES, whole-exome sequencing.

Although somatic NGS panels can identify germline variants, they should not be used to screen or test for HHMs for several reasons. First, somatic panels do not include all of the genes associated with germline predisposition syndromes, which restricts the data to genes covered by the panel and potentially leads to false-negative results. Second, non-coding regions that are well-described germline hotspots, such as the intronic enhancer region of GATA2 (40,86), the 5′UTR of ANKRD26 (87), or TERC, the telomerase RNA component (88,89), are typically not covered by somatic panels (Table 1). Third, somatic reversion in hematopoietic tissues, including blood and bone marrow used for somatic testing, is a phenomenon that occurs more often than previously thought and has been described in numerous BMF syndromes (e.g. FANCB (90), RPL4 (91), RPS19 (92), RPS26 (93), TINF2 (94) and UBE2T51 (95)) or immunodeficiency syndromes (e.g. CARD11 (96), DOCK8 (97) and IL2RG (98)). Somatic reversion has most frequently been reported in patients with SAMD9/SAMD9L gain-of-function variants either through loss of heterozygosity (−7 or del(7q)), uniparental disomy, or acquisition of loss of function variants (Table 1) (99–103). Although somatic reversion can lead to long-lasting remissions, overcorrection of SAMD9/SAMD9L variants by loss of multiple myeloid tumor suppressor genes on chromosome 7q and interferon-induced clonal expansion of hematopoietic stem cells with −7/del(7q) in the bone marrow of patients can cause MDS/AML (104).

Germline confirmation of any variant detected via somatic testing should be performed on DNA that reasonably approximates the germline state, which can be obtained from cultured skin fibroblasts, bone marrow-derived mesenchymal stromal cells, or hair roots (Table 1) (38,39,105). Cultured skin fibroblasts provide several advantages, including their ease of collection at the time of a bone marrow biopsy, laboratory familiarity with in vitro growth conditions from procedures like amniocentesis and chorionic villus sampling, ability to provide large quantities of DNA, and availability for downstream scientific studies, such as the generation of induced pluripotent stem cells by established protocols. Demonstrating segregation of a gene variant among family members is a second means of determining an allele’s germline status. Other tissue types, such as blood/bone marrow in remission, CD3-sorted T cells, saliva, buccal swabs, paraffin-embedded tissue or nails, all pose the risk of contamination with hematopoietic/malignant cells and should be avoided if possible.

Expectations and Hopes for the Future

We look forward to the near future when the shortcomings acknowledged in the sections above yield improvements in diagnosis and management for individuals and families with germline susceptibility alleles (Table 2).

Table 2

Current and potential future testing criteria for germline predisposition alleles

	In use now	In the future
Personal history of cancer	At a young age compared to the general population [e.g., MDS or AA diagnosed at 40 years old or younger]	Perform germline genetic testing on all patients diagnosed with a hematopoietic malignancy
	Two or more cancers (excluding non-melanoma skin cancer)	As outlined above
Personal history of syndromic features suggestive of germline risk	Includes platelet defects; pancytopenia; bleeding diathesis; skin or nail abnormalities; unexplained liver disease/cirrhosis; pulmonary fibrosis or alveolar proteinosis; limb anomalies; lymphedema; immune deficiencies/atypical infections, among others	no change
Deleterious variant identified in a gene known to confer risk for cancer and detected in malignant cells	Obtain germline DNA and test for deleterious gene variant to test germline versus somatic nature of the variant	Simultaneous collection of tissue for isolation of germline DNA and hematopoietic malignancy [e.g., skin punch at the time of diagnostic bone marrow biopsy]
Relevant family history in a proband diagnosed with a hematopoietic malignancy	Another diagnosis of a hematopoietic malignancy or a young-onset (<50 years old) within two generations of the proband	Standardization of universal germline genetic assessment for all patients diagnosed with a hematopoietic malignancy regardless of family history
	Other haematologic abnormality [e.g., excessive bleeding, cytopenias, macrocytosis, anemia in men] within two generations of the proband	As outlined above
Proband is being considered for allogeneic hematopoietic stem cell transplant	Review the personal and family history of all relatives being considered as allogeneic hematopoietic stem cell donors and perform appropriate germline testing prior to mobilization	Standardization of universal germline genetic assessment for all patients diagnosed with a hematopoietic malignancy would allow germline risk determination prior to consideration of related allogeneic stem cell donors
		Standardization of assessment of germline genetic risk of all hematopoietic stem cell donors regardless of relation to the transplant recipient

Widespread availability of comprehensive germline genetic testing

Busy clinicians assume that commercial laboratories and major academic medical centers provide testing that covers all relevant genes and is capable of detecting all types of deleterious germline genetic variants, including assessment of single nucleotide variants and CNVs (71). Currently however, most panels advertised as appropriate for testing ‘germline susceptibility to HMs’ fail to provide comprehensive testing (71). Many tests are incapable of detecting all mutation types and/or lack relevant genes (71). Although testing laboratories may be aware of how common somatic reversion occurs within hematopoietic tissues, many nonetheless accept hematopoietic tissues for germline testing without strong and obvious explanations that this testing is inherently flawed. Thus, clinicians may choose to send blood and/or bone marrow samples out of convenience, which even in remission, cannot be considered adequate for germline testing. This lack of awareness on the part of clinicians compounds the problem of ensuring comprehensive testing and results in clinicians being unaware that testing has not been comprehensive and giving patients and families the impression of having had adequate testing. Genetic counselors are familiar with the capabilities and limitations of various testing platforms and can be critical advisors to ensure appropriate testing of at-risk individuals. We hope that in the near future, testing laboratories will be stricter in their submission requirements and will restrict testing to samples that reasonably approximate the germline state (e.g., DNA from cultured skin fibroblasts, mesenchymal stromal cells, and/or hair bulbs) and/or will provide strong caveats to any testing performed on hematopoietic tissues. Moreover, we hope that standards for providing clinical testing for germline predisposition to HMs will ensure that testing must be capable of detecting all mutation types in all known susceptibility genes. The ever-expanding list of germline predisposition genes challenges these laboratories to remain flexible and current with published literature.

Universal germline testing for all patients diagnosed with a HM and potentially for all allogeneic hematopoietic stem cell donors

The American Society of Clinical Oncology urges use of germline genetic testing when the frequency of a positive finding is expected in at least 5% of the tested population (106). Therefore, with the HMs, we are already above that minimum for several diagnoses: MDS and aplastic anemia diagnosed in anyone 40 years old or younger (28,34,39–41,100,107,108). As recognition of HM predisposition syndromes increases and comprehensive germline testing becomes more available (109–111), universal germline testing for patients with particular diagnoses may become the standard of care at key points in treatment, such as at diagnosis and when allogeneic stem cell transplantation is considered (112). We recognize that use of a related hematopoietic stem cell donor for allogeneic transplantation who shares a deleterious germline genetic variant with the patient confers risk for graft failure, poor graft function and donor-derived HMs (113). However, we stress that any donor presents the risk of introducing a deleterious germline variant, and we are aware of several cases in which introduction of such variants occurred from the use of unrelated allogeneic stem cell donors. Thus, we advocate consideration for germline predisposition testing as a future standard practice at the diagnosis of a HM and potentially for all allogeneic stem cell donors as well, regardless of whether they are biologically related to the transplant recipient. Recognition of a germline cancer predisposition disorder brings several advantages: (i) opportunity to select related allogeneic stem cell donors who lack the familial allele; (ii) recommendation of appropriate cancer surveillance plans and/or management of other medical conditions associated with the inherited disorder; and (iii) opportunities for cascade testing of additional family members.

Development of preventive strategies to delay or even prevent malignancies in at-risk individuals

Currently, we lack sufficient knowledge about the natural history of many germline predisposition disorders, the factors that drive progression to overt malignancy, and biomarkers that predict which individuals are at highest risk for progression. Those with deleterious germline variants that confer risk for HMs are much more likely to develop clonal hematopoiesis (CH) than those within the general population (60,114), but it is unclear if the somatically-mutated clones contribute directly to the development of HMs, or even if those with CH are at increased risk of progression. Several lines of evidence suggest that systemic inflammation contributes to progression of CH to malignancy by driving the expansion of somatically-mutated clones (115,116). Thus, we hope that the near future brings a deeper understanding of the molecular mechanisms by which an inflammatory milieu may promote progression in those with germline susceptibility and the identification of which individuals are at the greatest risk of progression. Ultimately, the design and implementation of strategies that delay or prevent development of malignancies in at-risk individuals based on a molecular understanding of disease pathogenesis will improve the quality of life for these families.

Acknowledgements

We thank our patients and their families who inspire our work on germline predisposition to the hematopoietic malignancies.

Conflicts of interests. The authors have no conflicts of interest to claim.