Myelodysplastic syndromes (MDS) constitute a heterogeneous group of hematopoietic stem cell disorders, characterized by ineffective hematopoiesis resulting in cytopenias and a variable risk to develop acute myeloid leukemia. Discrimination between MDS and non-clonal causes of cytopenia is regularly based on cytomorphology (CM) and conventional cytogenetics (CCG) and can be challenging in case of minor morphologic dysplastic features or if karyotypic abnormalities are lacking. In these cases an integrated multi-diagnostic approach is warranted.

The aim of this prospective study was to investigate the role of flow cytometry (FC) compared to other diagnostic tools in MDS. We evaluated bone marrow samples of 316 patients with unexplained cytopenias, who were routinely analyzed by CM and FC combined with CCG, i.e. karyotyping or interphase fluorescence in situ hybridization (iFISH) when karyotyping failed. CM as gold standard classified 82 patients as MDS and 114 patients as non-clonal diseases while in 120 patients the diagnosis was indeterminate. Four patients were classified as normal. CCG analysis was normal in 192 patients, abnormal in 58 patients (MDS-associated abnormalities in 28), and not performed in 66 patients because there was no indication based on the CM result. Additional iFISH analysis in 4 patients where CCG failed, showed no abnormalities. FC analysis, performed according to European LeukemiaNet guidelines (Van de Loosdrecht et al., Haematologica 2009 and Leukemia 2012), disclosed MDS in 90 patients, showed signs of dysmyelopoiesis in 74 patients, and no signs of MDS in 141 patients (Van de Loosdrecht et al., J. Natl Compr Canc Net 2013). FC as diagnostic tool in MDS reached a sensitivity of 83% and a specificity of 77%; the positive predictive value of FC was 78% and the negative predictive value was 83%, when compared to outcome of the multi-diagnostic approach in which CM was leading.

Focusing on the patient population with indeterminable CM (N=120), CCG analysis identified 1 patient with a chronic lymphatic leukemia associated abnormality and 5 patients with typical MDS abnormalities (Schanz et al., JCO 2012). The latter were also recognized by FC. In the remaining 114 CM-inconclusive cases, where CCG did not add to the diagnosis (note: ∼60% of MDS patients have normal karyotypes) FC identified 11 patients with a typical MDS profile, 28 patients with dysmyelopoiesis and 69 patients without any signs of dysmyelopoiesis. Based on the negative predictive value of FC (83%, as described above), we postulate that most of these 69 inconclusive cytopenic cases are no MDS. These cases might be classified as idiopathic cytopenia of unknown significance (ICUS), in the absence of other causes.

In summary, this prospective single center study analyzed the integrated diagnostic work-up of patients with unexplained cytopenias. We demonstrated that FC is instrumental in discriminating clonal and non-clonal causes of cytopenia by excluding underlying MDS when validated standard diagnostic tools such as CM and CCG are not informative.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.