The Role of Molecular Profiling of Bone Marrow Samples in Confirming the Diagnosis of Myelodysplastic Syndrome in Patients Presenting with Cytopenia

Introduction: Diagnosis of myelodysplastic syndrome (MDS) can be very difficult when blast count in bone marrow is <5%.  The demonstration of a mutation in one or more of the MDS-related genes is usually considered an objective confirmation of MDS.  However, recent reports suggest that normal individuals may have circulating clonal hematopoietic cells carrying MDS-related mutations.  We studied the relevance the mutated allele frequency and number of mutated genes in confirming the diagnosis of MDS in patients with cytopenia as determined using bone marrow samples.

Methodology: We analyzed Next Generation Sequencing (NGS) data from of 294 consecutive bone marrow samples referred to rule out MDS and were reported to be positive for mutation in one or more MDS-related genes.   All samples were tested for mutations in the following genes:   TET2, SF3B1, ASXL1, DNMT3A, SRSF2, RUNX1, NRAS, ZRSR2, EZH2, ETV6, TP53, CBL, NPM1, JAK2, U2AF1, IDH1, KRAS, IDH2, FLT3, PTPN11, SETBP1, and BCOR. The average depth of NGS testing in this targeted sequencing was approximately 10,000X.

Results: Of the 294 MDS samples with mutations, 103 (35%) had blasts <5%. Of the 103 samples, 36 (35%) showed mutations in one gene; the remaining (65%) had mutations in more than one gene. The frequency of the mutant allele was <20% in only 11 of 103 cases (11%).  The remaining 92 patients had either mutations in two genes or in one gene, but the mutant allele frequency was >20%.  Four of the 11 patients (36%) with one gene mutation and <20% allele frequency had cytogenetic abnormalities confirming the diagnosis of MDS [der(1;7)(q10;p10), del(5q), trisomy 8. and del(11)(q23)].  Of the remaining 7 patients with allele frequency <20%, 3 had mutations in DNMT3A, 1 in U2AF1 gene, 1 in TET2 gene, 1 in TP53 and 1 in SF3B1 gene.   Of these 7 cases, only two cases had an allele frequency <10%, one in TP53 gene and one in SF3B1 gene. 

Of the 92 cases with mutations in two genes or in one gene with allele frequency >20%, 26 patients (28%) had cytogenetic abnormalities confirming the diagnosis of MDS.   In fact in this group of 26 patients with cytogenetic abnormalities, only one patient had mutations at <20% in all mutated genes (TET2, DNMT3A and TP53), but also had del(17p). Of the remaining patients 65 cases without cytogenetic abnormalities, with more than one gene mutation, at least one gene had mutant allele at >20%. 

There was no statistically significant difference in the degree of cytopenia between patients with <20% one mutation and no cytogenetic abnormalities (N=7) and the 96 cases with mutations in two genes or in one gene with allele frequency >20%. There was no significant difference in the degree of cytopenia between the 36 patients with one gene mutation and 67 patients with more than one gene mutation.

Conclusion: This data suggests that bone marrow samples from patients with peripheral cytopenia should be tested by cytogenetic and molecular profiling using NGS and the analysis of MDS-related genes.   Our data suggests that when proper criteria are used, molecular profiling of bone marrow in the proper clinical presentation can help in confirming the diagnosis of MDS.  Our data suggests that the presence of mutations in more than one gene and the detection of mutant allele frequency >20% may comprise reliable criteria for the diagnosis of MDS. The presence of mutation in 20% of DNA usually reflects mutation in 40% of the bone marrow cells.  Patients with mutant allele frequency between 10% and 20% in the bone marrow and cytopenia most likely have MDS, but further studies are needed.  Mutant allele frequency in bone marrow of <10% is extremely rare when testing is performed in patients presenting with cytopenia.