Assessment of Mutation Status in a Large Series of Patients with Cytopenias with Normal Karyotypes and without Increased Blasts

Background: Myelodysplastic syndromes (MDS) are clonal stem cell disorders that can variably affect the erythroid, myeloid, and megakaryocytic hematopoietic lineages producing persistent cytopenias. Distinguishing MDS without increased blasts from immune-mediated cytopenias is difficult as criteria for dysplasia lack specificity and karyotypic changes are frequently absent. Given the need for better diagnostic assays to identify the clonal hematopoietic populations characteristic of MDS, we examined mutation status of 8 genes that are commonly mutated in the early stages of MDS. We profiled a group of 505 samples from patients with cytopenia(s) that were clinically and/or morphologically suspicious for MDS but had no increase in blasts and normal karyotypes.

Methods: All 505 samples analyzed met these criteria: 1) clinical suspicion of MDS; 2) concurrent complete blood count with cytopenia(s); 3) bone marrow biopsy without increase in blasts; 4) normal karyotype (ie normal diploid, constitutional changes or –Y). In 170 cases, a FISH panel for 6 MDS changes was also performed and was negative. DNA was extracted from bone marrow aspirate (n=461) or blood (n =44). Sequencing of EZH2, TET2, exon 12 of ASXL1, JAK2 codon 617 and mutation hotspots in IDH1, IDH2, KRAS and NRAS was performed on the Ion Torrent PGM platform and analyzed using SeqNext software. To exclude artifacts and constitutional and low-level background alterations, a positive result required confirmation by Sanger sequencing (or pyrosequencing) and was defined as frameshift change, nonsense or previously characterized missense change, or novel missense change with high mutation probability based on level and structural features. All single nucleotide polymorphisms (SNPs) seen in normal controls or public databases were excluded. Ninety-two cases were subsequently analyzed for mutations in 21 additional MDS-associated genes using a TruSeq assay on the MiSeq platform.

Results: Among 505 cases, M:F ratio was 1.02; 90.9% of patients were ≥50 years old. The frequency of MDS-associated mutations in cytopenic samples that had normal karyotype was 19.6% (99/505); mutations were identified in TET2 (63 cases), ASXL1 (36), IDH1 or IDH2 (8), NRAS (5), EZH2 (4), JAK2 V617F (4) and KRAS (1). The lowest mutation rate was in samples with isolated neutropenia (8.8%) with any cases showing neutropenia having an unusually high frequency of IDH1 or IDH2 mutation (7/39, 17.9%); other CBC parameters did not correlate significantly with incidence or type of mutation (Table). Multiple mutations were seen in 40 (7.9%) samples, with dual TET2 and ASXL1 mutation in 13. Among cases with several mutations, unequivocal subclonal mutation changes were seen in 17. Among 69 cases with mutations detected by the 8-gene panel, a larger sequencing panel detected additional mutations in 49 (71%), most commonly SRSF2 (9) and RUNX1 (3). 6/23 cases (26.1%) that were negative in the 8-gene panel had MDS-associated mutation(s) identified in the larger panel.

Conclusions: Using a targeted 8-gene panel, we identified MDS-associated mutations in approximately 20% of samples suspicious for MDS based on cytopenia(s) but with normal karyotype and no increase in blasts. In this community-based, unselected series, mutation detection rate was largely independent of hematologic pattern. Mutation density and subclonal complexity were relatively low even using a larger panel of genes. The high detection rate illustrates the enhanced clinical utility of targeted sequencing panels for identifying clonal alterations in early/low-grade MDS. Further work is needed to delineate the prognostic significance of mutations identified in minimally atypical cytopenic bone marrow samples.