Recognizing GPD can guide therapeutic decisions, appropriate genetic counseling, familial screening, and surveillance [6, 7]. However, patients with MNs with GPD may not have a family history of cancer, may lack a clinical phenotype, or have syndromic features that may be mild or unrecognized. In such instances, without testing for germline variants, the underlying GPD may not be identified. A study on myelodysplastic syndrome (MDS) showed that GPD occurred in patients of any age, even in those aged > 70 years [8]. This study showed that nearly 7% of the patients with MDS carried P/LP variants. GPD for bone marrow failure, DNA repair, and telomere biology disorders occurred at ages of < 40 years, whereas checkpoint disorders or germline variants in DDX41 occurred at a later age. A study of 1120 patients with pediatric cancer showed that 8.5% had germline mutations in cancer-predisposing genes and only 40% had a family history of cancer [9]. Thus, the 2022 European LeukemiaNet recommendation for the diagnosis and management of acute myeloid leukemia (AML) states that GPD should be considered in patients with any hematological malignancy, irrespective of age [10].

GPD with underlying bone marrow failure syndrome may require therapeutic modifications. In cases of GPD such as FA, hematopoietic stem cell transplantation (HSCT) with reduced-intensity conditioning is required because of the inherent hypersensitivity of these patients to genotoxic therapies. Significant treatment-related toxicities and prolonged cytopenia can occur with standard regimens [11, 12]. In a cohort of patients diagnosed with severe aplastic anemia who underwent HSCT, P/LP germline variants in 42 genes associated with inherited bone marrow failure syndromes were identified in 16.5% (121/732) of the patients [13]. In patients with GATA2 deficiency, various infectious complications during therapy should be considered [14]. Donor selection for HSCT is important as germline variants may be present in family members. Donor-derived malignancies have been reported in MNs with GPD for CEBPA [15], DDX41 [16], and GATA2 [17], and stem cell donors carrying pathogenic germline variants have demonstrated inferior outcomes characterized by challenges in stem cell mobilization or delayed engraftment failures [18].

Universal screening of individuals for GPD of MN is not currently the standard of care, but the American Society for Clinical Oncology recommends screening for hereditary cancer syndromes when (1) there is a personal or familial history suggestive of a hereditary cancer syndrome, (2) the screening test results can be accurately interpreted, and (3) the outcomes of the screening will contribute to the diagnosis or assist in managing the patient or family members at risk [19]. The Nordic guidelines recommend GPD testing when hematologic malignancies are diagnosed in a patient with a family history or signs/symptoms indicative of a hereditary condition, when gene variants are suspected to be germline based on somatic testing, or when MDS/AML is diagnosed in patients aged < 50 years in the presence of chromosome 7 aberrations [7].

The challenge in diagnosing germline variants in hematologic malignancies lies in the fact that peripheral blood is not an ideal source for genetic testing because the hematopoietic cells themselves are the source of the tumor. The recommended specimen for confirming germline variants is non-hematopoietic, such as skin fibroblasts, thus enabling less contamination by blood cells. However, this requires an additional procedure for a skin biopsy and additional time for culturing [20]. Other specimens, such as nail clippings, hair bulbs, buccal swabs, bone marrow, or peripheral blood at remission status, may be used but with caution. For saliva and buccal swabs, contamination by white blood cell may complicate interpretation, and a sufficient amount of DNA should be obtained from nail clippings and hair bulbs [21, 22]. Somatic mosaicism may occur, leading to situations in which a germline variant is not identified due to somatic reversion in blood cells; this has been described in cases involving RUNX1, SAMD9, and SAMD9L [23]. Inappropriate specimens can lead to false-negative or false-positive results, which may have a critical impact. Thus, the use of skin fibroblast samples are recommended for testing.

When an appropriate sample has been selected, a methodology to identify GPD variants should be considered. The selection depends on regulatory aspects, costs, and availability within the institution. Therefore, targeted gene panels using next-generation sequencing (NGS) should be considered. The genes included in the panels are most likely based on the current classifications of GPD, and the gene lists for hereditary disorders can be reviewed through the Gene Curation Coalition (GenCC) [24]. Importantly, differences in hotspots for somatic and germline variants may exist in certain genes. These differences should be considered during the panel selection, testing, and interpretation. In the United Kingdom, the National Health Service (NHS) in England sanctioned the implementation of whole-genome sequencing (WGS) as a standard care practice for all patients with acute leukemia. This approach involves performing paired tumor and germline WGS, which facilitates the identification of a greater number of germline variants than existing methods [25]. Germline testing strategies vary based on care plans and institutions in the United States and Korea. The availability of germline panels is increasing; however, WGS is not yet in clinical use in most scenarios.

For familial AML with CEBPA mutations, N-terminal germline variants with acquired somatic mutations in the C-terminus have mostly been reported. The presence of multiple CEBPA mutations or truncating alleles in tumor-based molecular profiling may warrant additional germline testing [26]. In myeloid neoplasms with DDX41 variants, 85% are germline and > 95% of the truncating alleles are germline variants. Among DDX41 variants, the c.3G > A, p.M1? start-loss allele is often detected as a germline variant. However, a low variant allele frequency (VAF) may be observed in tumor profiling because of technical issues, which can lead to misinterpretation as a somatic variant, necessitating caution in interpretation [27]. Different ethnic groups have different variant frequencies: Japanese and Korean individuals are enriched with c.1496dup, whereas c.3G > A and c.415_418dup are more common in individuals of Northern European descent [28–30]. The presence of multiple DDX41 variants, especially those with high VAF, suggests a germline mutation.

In the case of RUNX1, the same variants have been identified in both somatic and germline settings within hematologic malignancies, highlighting the challenge of determining when to use germline confirmation. Germline RUNX1 variants are distributed throughout the gene, necessitating sequencing the entire gene. These variants include missense, nonsense, frameshift, and whole-exon deletions or duplications [31]. A previous study involving 45 families revealed that partial or whole deletions of the RUNX1 locus comprised a significant portion, along with splice sites and intragenic duplications [32]. Germline ANKRD26 variants are located in the 5’ untranslated regions (UTR) of c.-116 to c.-134, leading to overexpression of ANKRD26 owing to the failure of regulation by transcription factors RUNX1 and FLI1; thus, this region should be included in analyses [33, 34]. Germline ETV6 mutations are typically found in the N-terminal central regulatory domain and C-terminal ETS motif. Notably, the distribution of somatic ETV6 variants showed a domain pattern similar to that of the germline variants.

For GATA2 deficiency, germline mutation involves truncating mutations, missense mutations within zinc finger 2, and noncoding variants in the + 9.5-kb regulatory region of GATA2. Most adolescent individuals with monosomy 7 MDS carry an underlying GATA2 deficiency [35]; thus, genetic testing for GATA2 may be necessary and is recommended in the Nordic guidelines [7]. Germline variants of SAMD9/SAMD9L can be accompanied by acquired loss-of-function mutations in the same gene (in cis) or by monosomy 7. This leads to loss of the mutated germline allele, complicating genetic testing because only the wild-type SAMD9/SAMD9L allele remains [36, 37].

As somatic tumor panels are widely used in hematologic malignancies, suspicious germline variants may be encountered in these panels. Additional testing is necessary to confirm that the variants are germline variants [38]. Drazer et al. identified germline variants in 21% of patients using tumor sequencing panels for hematologic malignancies and showed that a VAF of > 0.4 in the gene of interest may be predictive of a germline origin [39]. Experts have suggested that RUNX1 and ETV6 should be carefully assessed for germline origin if detected with a high VAF. It has been suggested that certain mutation patterns may indicate germline origin. These include the presence of mutations in both alleles of a gene, gene mutations accompanied by copy number variations (CNVs) in the same gene, large exon-spanning duplications or deletions, and the persistence of gene mutations in follow-up studies, even during remission [40].

There are numerous associated genes in MNs with GPD in bone marrow failure syndromes, such as FA, DBA, and Schwachman-Diamond syndrome (SDS). In FA, over 20 genes are implicated and for some genes like FANCA, 40% of the variants are detected through deletion/duplication analysis, necessitating careful consideration of the testing type [41]. Biallelic pathogenic variants of SBDS have been identified in most patients with SDS, with mutations commonly occurring within exon 2, c.258 + 2 T > C and c.183_184delinsCT [42]. Challenges in genetic testing include the presence of an SBDSP1 pseudogene, which shares 97% sequence identity with SBDS, complicating the identification or estimation of VAF because reads may be misaligned to SBDSP1 [43]. It is crucial to determine whether the variants are cis or trans, as a single heterozygous SBDS mutation with one wild-type allele is insufficient for the disease. Variants in other genes such as EFL1 and SRP54 may result in clinical features that overlap with those of SDS. The DBA genotype is heterogeneous and involves more than 20 ribosomal protein genes, with RPS19 being the most frequently mutated. Large deletions were found in RPS17, RPL35a, and RPS19, which should be considered during testing.

The process of interpreting germline variants adheres to the guidelines established by the American College of Medical Genetics and Genomics (ACMG) and the Association for Molecular Pathology (AMP) [44], which differs from the interpretation of somatic variants based on the AMP/College of American Pathologists/ACMG guideline [45]. Germline variants are classified based on a five-tier system: pathogenic, likely pathogenic, benign, likely benign, and uncertain significance. The rules for combining criteria to classify sequence variants are complex and require the assessment of evidence of pathogenicity (very strong, strong, moderate, and supporting) or benign (stand-alone, strong, and supporting). However, as the guideline is a general rule and gene-specific guidelines are lacking for the majority of genes included for testing the GPD of hematologic malignancies, practical recommendations have been proposed for interpreting germline variants for hematologic malignancies, bone marrow failure, and chronic cytopenia [46]. This guide provides thresholds for minor allele frequencies, in silico predictions, the use of constraint Z-scores, and functional evidence required for assessing pathogenicity. For the same variant, differences in variant interpretation may exist between the somatic and germline settings because different variant interpretation guidelines are used for different purposes and needs. Mutation type, minor allele frequency, published studies, in silico tools, and germline databases can be considered for both somatic and germline testing. However, for somatic variants, Food and Drug Administration (FDA)-approved therapies, somatic databases, variant frequencies in tumors, and professional guidelines should be considered when segregation information, in trans findings, and patient phenotype are also taken into account. Clinicians should be aware of the difference [47, 48]. Specific guidelines for variant interpretation have been proposed for only a few genes or conditions, such as RUNX1 [49, 50], TP53 [51], and RASopathy [52]; however, the list is expanding and ClinGen Expert panel reports should be prioritized. One report showed that only 21% of RUNX1 variants listed in ClinVar are clinically significant, thus suggesting caution in variant interpretation [49].

The ACMG has proposed points for consideration when reporting germline variations in patients undergoing tumor testing [38]. Individuals undergoing tumor testing must provide informed consent, acknowledging the potential discovery of germline pathogenic variants. It should be clearly communicated to patients that the identification of such variants will lead to a referral for genetic consultation and may necessitate confirmatory germline testing. Additionally, confirmatory germline testing should be conducted in a clinical laboratory that specializes in this area, with the results communicated by both qualified and experienced clinicians.

Most guidelines recommend that genetic testing be performed with pre- and post-test counseling [7, 19]. Currently, no standard surveillance guidelines are available for unaffected carriers of germline P/LP variants. However, peripheral blood cell counts, family history, and physical examinations are performed with an initial bone marrow workup and consultation for HSCT [7, 53]. Surveillance in asymptomatic carriers depends on the age and specific disorder, as some variants are prone to result in MNs at a young age, whereas others, such as DDX41, manifest later in life. Expert panels have recommended surveillance for children with leukemia-predisposing conditions because some cases may benefit from preemptive treatment with allogeneic HSCT [54]. Early referral to a transplant specialist and donor identification should be considered.