Introduction: The clinical diagnosis of myelodysplastic syndrome (MDS) relies on the presence of persistent cytopenias, not otherwise explained, and evidence of morphologic dysplasia in the bone marrow. Low grade MDS (bone marrow blasts <5%) is defined by the presence of morphologic dysplasia in at least 10% of cells in one or more cell lineage. Low grade MDS is particularly challenging to diagnose, as morphologic dysplasia may be subtle and is subject to high inter-observer variability. The ability to diagnose low grade MDS can be improved by incorporating cytogenetic evaluation of the bone marrow, especially in the setting of equivocal morphologic dysplasia. However, many MDS cases (up to 60%) lack cytogenetic abnormalities, limiting the overall utility of cytogenetics. Multiple recent studies have demonstrated that the majority of MDS patients (over 80% in some studies) harbor recurrent somatic mutations in a core group of genes. We sought to determine if targeted DNA sequencing of genes recurrently mutated in MDS and AML could be useful in the evaluation of cytopenic patients with a normal karyotype being evaluated for the possible diagnosis of MDS.

Methods: We screened patients who presented for evaluation of MDS between 2002 and 2014 that had consented for sequencing studies and had banked samples. Patients were selected based on 1) World Health Organization defined cytopenia (WBC <1,800/µL, hemoglobin <10g/dL, platelets <100k/µL) in at least one lineage, 2) bone marrow blasts <5%, 3) WBC <14k/uL, 4) normal cytogenetics, and 5) absence of prior therapy for MDS. Bone marrow specimens were independently re-reviewed by two board certified hematopathologists. DNA was extracted from cryopreserved bone marrow and skin (to serve as a source of normal DNA) and enriched for a panel of 285 commonly mutated myeloid genes. Captured DNA libraries were sequenced on a HiSeq 2500 instrument with 2x101bp reads. The resulting data was analyzed for single nucleotide variants (SNVs) and insertions/deletions (indels) using VarScan2 in paired normal mode.

Results: Thirty-eight patients met the selection criteria, and 30 of these had bone marrow aspirates available for morphologic review and were included in the study. A mean unique coverage depth of 913x was achieved for targeted genes and all reported variants had >50x coverage, variant allele fractions (VAFs) >3%, and minor allele frequencies (MAFs) < 1% in any population. Of the 30 sequenced cases, 25 had a somatic mutation in at least one gene (mean 3.3 mutations/case, range 1-10 mutations/case). The most commonly mutated gene was TET2 (7 cases), followed by ASXL1 (5 cases), EZH2 (4 cases), SRSF2 (4 cases), and U2AF1 (4 cases). Of the 285 sequenced genes, 44 were mutated in at least one case, and 14 were mutated in 2 or more cases. The mean VAF (variant reads/total reads) of detected mutations was 27% (range 3-98%). Morphologic review demonstrated definitive dysplasia (≥10% of cells in least one lineage) made by two pathologists in 18 of 30 cases (supporting the clinical diagnosis of MDS), no dysplasia in 6 of 30 cases, and equivocal dysplasia (where hematopathologists did not agree that dysplasia was ≥10%) in 6 of 30 cases. Thirteen of 18 cases (72%) with definitive dysplasia had a mutation, 5/6 cases (83%) without dysplasia had mutations, and 6/6 (100%) cases with equivocal dysplasia harbored somatic mutations. The mean VAF of mutations was 17.5% in cases without dysplasia, 29% in cases with equivocal dysplasia, and 28% in cases with definitive dysplasia. All of these groups included mutations in canonical MDS genes such as TET2, DNMT3A, SRSF2, RUNX1, and EZH2.

Conclusions: In this cohort of 30 cytopenic patients with normal cytogenetics, 80% harbored a somatic mutation in at least one myeloid-associated gene. Somatic mutations were detected in 5 of 6 cases without definitive dysplasia (<10% dysplasia) and 6 of 6 cases with equivocal dysplasia. Notably, canonical MDS mutations were found even in the absence of dysplasia. These findings suggest that clonal hematopoiesis may be present in the majority of cytopenic patients independent of dysplasia, a finding that requires independent validation. Identification of somatic gene mutations in patients with morphologically equivocal MDS or cytopenic patients without definitive dysplasia provides a means for tracking clonal disease that could be used to monitor patients for subsequent development of definitive MDS.


Duncavage:DI&P Consulting: Consultancy; Cofactor Genomics: Consultancy. Jacoby:Sunesis: Research Funding; Novo Nordisk: Consultancy.

Author notes


Asterisk with author names denotes non-ASH members.