Germline cancer predisposition is increasingly recognized in patients with myeloid malignancies, including MDS and AML. At the Best of Hematology & Breast Cancer 2026 conference, Dr. Sioban B. Keel of Fred Hutchinson Cancer Center outlined how inherited bone marrow failure syndromes, DDX41 mutations, and other germline variants influence disease risk, transplant decisions, donor selection, and long-term surveillance. Her presentation emphasized that up to 7–10% of patients with hematologic cancers may carry an inherited predisposition—reshaping how clinicians approach diagnosis, clonal hematopoiesis, and precision care.

Dr. Sioban B. Keel is a physician at Fred Hutchinson Cancer Center, Robert and Phyllis Henigson Endowed Chair in Hematology at UW Medicine, and Associate Professor in the Division of Hematology and Oncology at the University of Washington School of Medicine. At the Best of Hematology & Breast Cancer 2026 conference in Seattle, Dr. Keel presented a state-of-the-art update on inherited bone marrow failure and hematologic malignancy predisposition syndromes, with practical guidance for practicing hematologists and oncologists on recognizing germline risk, understanding how these disorders inform patient care, and appreciating their role in reshaping our understanding of clonal hematopoiesis.

The transcript report below has not been reviewed by the speaker and may contain errors.

Overview of Inherited Predisposition to Hematologic Malignancies

When discussing inherited predisposition to hematologic malignancies, the focus is on a broad, heterogeneous group of conditions. These include the classical bone marrow failure syndromes, disorders where myeloid malignancy is the most prominent feature, mature cancer predisposition syndromes often thought of primarily for solid tumor cancer risk that also pose risk of hematologic malignancy, and a small group of disorders classified as inborn errors of immunity that are increasingly being recognized. Importantly, many of these patients present for the first time in adult clinics.

Incidence and Risk of Hematologic Malignancy

The first question addressed was the incidence of these conditions and their associated risk of hematologic malignancy. A large study of over 400 patients with MDS undergoing related hematopoietic stem cell transplantation, spanning the full age spectrum, provides important insights. Patients were not selected based on age, physical features, or family history. They underwent enhanced whole exome sequencing for germline variants, and 7% were found to have a pathogenic or likely pathogenic germline variant. This is almost certainly an underestimate, as the germline nature of the variant was determined based on having both a sample from the patient and the donor.

The important finding is that the identified germline pathogenic mutations reflected the underlying biology. In young individuals, the incidence of germline mutation was enhanced, with 33% of those aged 0-20 years having mutations. The genes mutated in young patients tended to be the classical inherited bone marrow failure disorders. In older individuals, mutations were more commonly in the classical cancer predisposition genes. Short telomere syndromes spanned a broader spectrum of ages, while DDX41, which will be discussed in more detail, was restricted to the older population.

The prevalence of germline cancer predisposition in myeloid malignancy patients is now recognized to be comparable to cancer predisposition incidence in many other solid tumors, including ovarian, colorectal, and prostate cancer, where estimates are somewhere around 10% of individuals having an underlying inherited cancer predisposition syndrome. This is highlighted because germline cancer predisposition testing has long been incorporated into routine clinical care in the solid tumor space, but less so in hematology.

Population-Based Prevalence Data

The question has also been approached from a genome-first, population-scale perspective using two large exome sequencing cohorts with linked health records. In this study, investigators quantified the prevalence of pathogenic or likely pathogenic germline variants across eight well-established myeloid malignancy predisposition genes. One in 500 individuals harbored a pathogenic germline variant, with DDX41 being the most prominent at 1 in 500 individuals. If pooling together those with germline mutations in the transcription factors CEBPA, RUNX1, or ETV6, the prevalence was about 1 in 7,000.

However, prevalence is only part of the story. What truly matters is how many of these patients actually develop a hematologic malignancy. In this population-based study, the penetrance of developing a hematologic malignancy among germline mutation carriers was substantial, on the order of 1 in 19 to 25 individuals. The penetrance really varied by age. When the analysis was limited to CEBPA, ETV6, and RUNX1—genes that have stronger and earlier myeloid risk—the penetrance was 1 in 4 individuals. For DDX41, which is the most commonly mutated gene in individuals with myeloid malignancy and is more prevalent, the penetrance of actually developing myeloid disease was relatively low.

Germline Variation and Somatic Evolution

Another emerging concept is that germline variation doesn't just increase cancer risk but also shapes the trajectory of somatic evolution. In a large UK Biobank analysis of whole exome sequencing, investigators queried 236 cancer predisposition genes and identified autosomal dominant cancer predisposition in roughly 10% of patients, 8% in this study. The most commonly mutated genes were CHEK2, ATM, and BRCA2. Individuals with an autosomal dominant germline predisposition syndrome were more likely to develop clonal hematopoiesis, and the particular somatic mutations in those clones were dependent on the underlying germline allele. The curves demonstrate that having a germline variant plus having clonal hematopoiesis on top of it increases the risk of developing a frank hematologic malignancy much more substantially.

Key takeaways from this section are that germline predisposition to myeloid malignancy is common, the age of presentation is a surrogate for the biological pathway involved, and there is variable penetrance for hematologic malignancy development depending on the gene. CEBPA, ETV6, and RUNX1 are highly penetrant cancer predispositions, whereas DDX41 is less so.

Clinical Settings for Suspecting Germline Predisposition

The clinical settings that should prompt consideration of germline predisposition include AML or MDS patients who are under 50, particularly if they have a longstanding history of macrocytosis with or without cytopenias. In adolescents and young adults with MDS and chromosome 7 abnormalities, GATA2 deficiency and mutations in genes called SAMD9 or SAMD9L should be considered, as these are particularly enriched in adolescent and young adult patients. TP53-mutated hypodiploid ALL is another key setting that frequently reflects an underlying syndrome in the pediatric population. Individuals with myeloid malignancy and a secondary cancer (non-squamous cell skin carcinoma excluded) should also raise suspicion, as should anyone with clinical features suggestive of a germline predisposition syndrome at any age.

How Germline Diagnoses Change Patient Care

The next question addressed was how these diagnoses change care. A germline diagnosis has direct actionable implications for patient care. It informs how patients are monitored, not only for hematologic complications but also for non-hematopoietic manifestations that may emerge over time. It is essential for accurate family counseling and family planning. Germline information is also critical when considering a hematopoietic stem cell transplantation pathway. It guides if and when to transplant, as well as the choice of conditioning regimen, since many of these inherited syndromes are associated with increased sensitivity to chemotherapy. Perhaps most importantly, it informs donor selection.

Donor-Derived Malignancies

An abstract from ASH this year addressed donor-derived malignancies. Donor-derived malignancies after transplant have been reported with latencies ranging from months to decades. There are many donor and recipient-related factors that contribute to the development of donor-derived malignancies, and they have been reported in both related and unrelated donor settings in the context of donors with mutations in specific genes.

The ASH abstract represented the largest retrospective study of donor-derived malignancy to date, across multiple transplant centers in Spain. Investigators identified 34 donor-derived malignancies occurring between 2007 and 2024. These were predominantly myeloid malignancies. Notably, a large number of these patients had germline predisposition—40% of the donor-derived malignancies. The genes involved are shown, with the most common being DDX41.

DDX41-Associated Myeloid Malignancies

DDX41-associated myeloid malignancies are the most common inherited myeloid malignancy predisposition syndrome. DDX41 is a multifunction RNA-binding protein that accounts for about 2-6% of so-called sporadic MDS. The presentation looks just like regular MDS, with age of presentation typically in the 70s.

A couple of keys to identifying germline DDX41: somatic mutations in DDX41 are pretty rare, only about 4% of cases, so suspicion should be raised if a DDX41 mutation is seen. There is, for reasons not fully understood, male predominance in the development of myeloid malignancies. The disease is oftentimes characterized by normal karyotype, and over 50% of cases have a second somatic DDX41 mutation. If two mutations in DDX41 are seen, one should really look for one of those being a germline mutation. These mutations are common in the population on a population basis, about 1 in 403 individuals.

The takeaway is that these disorders inform care. When feasible, avoid choosing a donor that has the same germline predisposition as the recipient. Since January 2023, the National Marrow Donor Program has created a policy allowing donors to opt in at workup to receive notification of a germline variant that might be identified through post-transplant recipient testing. This is changing the field. An open question is whether routine germline testing should be considered for all donors, though the field hasn't reached that point yet.

What These Disorders Teach About Clonal Hematopoiesis

The final section addresses what these disorders are teaching about clonal hematopoiesis. The concept of mosaicism is introduced as a framework. In genetics, mosaicism refers to a situation in which a genetic change occurs after fertilization, meaning it's post-zygotic, so only a subset of cells carry that change. In the hematopoietic system, post-zygotic mutations can occur at any point from very early development, including in utero, to later in life.

For example, mutations in JAK2, the V617F mutation in polycythemia vera, have been shown to occur in some individuals before birth and can persist dormant for a very long time before clinical disease manifests. In hematopoietic post-zygotic mosaicism, the acquisition of a mutation in a single hematopoietic stem cell persists in the progeny of that cell. Clinical examples seen in the clinic include CHIP, PNH, and somatic activating mutations in lymphoid malignancies. These mutations can be of different kinds: single nucleotide variants, copy number alterations, insertions, deletions, or structural variants.

Genetic Reversion

Before discussing specific diseases, it is helpful to review the concept of genetic reversion. Reversion occurs to reverse a proposed genetic change that restores or functionally compensates for a pathogenic variant, giving the affected cell a selective growth advantage. There are several ways a cell can have a reversion event: true back reversion in an autosomal dominant disorder where the mutation just goes away, a second-site mutation that compensates for the original germline mutation, or most importantly, copy neutral loss of heterozygosity, which is essentially where one allele—one parental allele that lacks the mutation—is duplicated, thus losing the germline mutation.

Somatic Mosaicism in Inherited Bone Marrow Failure Syndromes

Within this framework, hematopoietic somatic mosaicism in inherited bone marrow failure and malignancy syndromes can be understood. Starting with a hematopoietic stem cell that carries a germline variant restricting stem cell fitness, under selective pressure, events arise to improve fitness. These can push hematopoiesis in two very different ways. One could acquire oncogenic hits that move toward malignancy, such as monosomy 7 or TP53 mutations. Optimistically, one could also acquire a reversion event that restores functional hematopoiesis. Importantly, in both settings, there is selection for these events because it improves hematopoietic stem cell fitness, but in one, it leads to cancer, and in another, it leads to functional improvement.

These are now well described in many inherited bone marrow failure syndromes. Really elegant work in an abstract presented at this year's ASH from the St. Jude's group looked at this concept in Diamond-Blackfan anemia, an inherited pure red cell aplasia. It was long seen in the clinic that some of these patients were born anemic and just got better—the anemia improved through unknown mechanisms. This work showed that upwards of 10% of patients with Diamond-Blackfan anemia acquired uniparental disomy, potentially accounting for those individuals whose anemia improved.

Shwachman-Diamond Syndrome as a Clinical Example

One other example of how these genetic principles manifest in clinical practice is Shwachman-Diamond syndrome, an autosomal recessive disorder marked by bone marrow failure and pancreatic insufficiency. It's due to biallelic mutations in SBDS that lead to a defect in the ribosome and chronic activation of p53. These patients have a very high lifetime incidence of AML and MDS, and unfortunately, once they develop AML or MDS, even with transplant, outcomes are very poor.

As a reminder of how mutations in SBDS lead to Shwachman-Diamond syndrome: SBDS is a protein that assists in the removal of another protein called EIF6, allowing the two ribosomal subunits to join. In Shwachman-Diamond syndrome, when SBDS is dysfunctional, EIF6 can't be removed and the two ribosomal subunits are not allowed to join. This creates p53 activity and leads to problems with hematopoietic stem cell fitness.

In Shwachman-Diamond syndrome, about 50% of patients have TP53 mutations and about 60% have mutations in EIF6. Within a single clone, these two mutations are mutually exclusive. In foundational work by Kannengiesser and colleagues and others, germline SBDS deficiency causes persistent ribosome joining defect and reduces stem cell fitness. Under this selective pressure, somatic evolution follows two paths.

One path is malignant and defined by progression to biallelic TP53 mutations within the same clone. A single TP53 alteration can exist for many years, but it's the acquisition of a second mutation in a single clone that leads to cancer. Importantly, biallelic TP53 mutations can be detected years before malignancy develops.

Alternatively, in an adaptive mechanism of somatic reversion events, patients acquire EIF6 mutations. This still improves hematopoietic stem cell fitness, but that mutation allows ribosome subunit joining and improves that process.

A clinical pearl highlighted was an example of an individual followed serially over time with three clones: a TP53 clone, an EIF6 clone, and another TP53 clone. When bulk clinical NGS was performed on whole marrow to track these clones over time, it appeared things were steady. But closer examination by single-cell analyses at year three revealed the presence of a biallelic TP53 clone. The reason for showing this is that while single-cell analyses aren't being done in the clinic yet, it offers the promise that if biallelic clones could be identified, there's a real opportunity for clinical intervention before patients develop MDS.

Key Takeaways and Clinical Recommendations

The key takeaway points are that somatic mosaicism can go both ways—rescue or malignancy—and that actionable clinical monitoring strategies are evolving in this context. Very nice abstracts at ASH this year addressed this. An open question is whether these patients could be monitored with peripheral blood rather than requiring bone marrow examination, and there was encouraging data at this year's ASH on that.

Another important point about somatic mosaicism in the clinic is that it can complicate the initial diagnosis of a germline disorder. If a person has a somatic reversion event, an inherited mutation being detected in bone marrow might appear at a VAF lower than the 50% heterozygous state, and a skin biopsy might be needed in that context. In Diamond-Blackfan anemia, this concept may explain some of the clinical heterogeneity in these disorders and why some patients have less severe phenotypes.

The two broad recommendations to take away are to remain vigilant for germline cancer predisposition across all cancers—solid tumors and hematologic—as it occurs in somewhere around 7-10% of cases. There is now an opportunity in the clinic to leverage evolving care and surveillance strategies and referrals to dedicated programs for these patients.