Detection of Dysplastic Features by Multiparameter Flow Cytometry in Myelodysplastic Syndromes in Relation to Cytomorphology and Cytogenetics.

Abstract

Dysplastic features can be detected in different cell lineages in myelodysplastic syndromes (MDS) by multiparameter flow cytometry (MFC). The aim of the present study has been the assessment of the flow cytometric detection of dysplastic features previously published to occur in MDS in relation to findings in cytomorphology (CM) and cytogenetics (CG). We analyzed 307 bone marrow samples from patients with suspected (n=130) or proven (n=177) MDS by MFC, CM, and CG in parallel. Blast counts as determined by CM and MFC, respectively, ranged from 0% to 21% (median, 3.5%) and from 0% to 23% (median, 3%; r=0.271, p<0.0001). The median number of aberrant features detected by MFC were 0 for blasts (range, 0 to 4), 2 for granulocytes (0 to 5), 1 for monocytes (0 to 5), and 0 for erythroid cells (0 to 2); median total number=3, range 0–11. The most frequent dysplastic features observed in the blast populations included aberrant coexpression of CD11b (13.7%), CD15 (10.4%) and CD64 (10.4%). The most frequent dysplastic features observed in the granulocytic cell populations included reduced side-scatter signal corresponding to hypogranulation (67.1%), aberrant coexpression of CD56 (32.9%), aberrant pattern of CD13/CD16 expression (31.6%), aberrant pattern of CD11b/CD16 expression (24.1%), and reduced expression of CD33 (13.4%). The most frequent dysplastic features observed in the monocytic cell populations included aberrant coexpression of CD56 (42.7%) and of CD16 (18.9%). The most frequent dysplastic features observed in the erythroid cell populations included a lack of CD71 expression (15.0%) and an aberrantly homogeneous expression of CD71 (9.1%). As compared to cases with no indication of MDS by CM (=non-MDS) cases with MDS according to CM were significantly associated with a reduced side-scatter signal in granulocytes (ratio granulocytes:lymphocytes 6.53±1.27 vs. 7.44±1.17, p<0.0001) as well as a higher number of dysplastic features in granulocytes (1.98±1.09 vs. 1.00±1.31, p<0.0001), monocytes (0.81±0.84 vs. 0.35±0.63, p<0.0001), and erythroid cells (0.33±0.47 vs. 0.20±0.40, p=0.061). Particularly, an aberrant expression of CD56 in monocytes occurred more frequently in 33 cases with CMML as compared to non-MDS cases (84.8% vs. 15.7%, p<0.0001). In cases with possible MDS according to CM the differences to non-MDS cases were less pronounced (reduced side-scatter signal in granulocytes 7.77±1.47 vs. 7.44±1.17, n.s., dysplastic features in granulocytes 1.90±1.21 vs. 1.00±1.31, p=0.003, monocytes 0.50±0.63 vs. 0.35±0.63, n.s., and erythroid cells 0.30±0.47 vs. 0.20±0.40, n.s.). In cases with aberrant cytogenetics (n=80, excluding those with –Y as sole aberration) dysplastic features by MFC occurred more frequently as compared to cases with normal karyotypes (reduced side-scatter signal in granulocytes 6.35±1.18 vs. 7.09±1.38, p<0.0001, dysplastic features in granulocytes 2.05±0.95 vs. 1.68±1.31, p=0.021, monocytes 0.80±0.89 vs. 0.71±0.81, n.s., and erythroid cells 0.38±0.49 vs. 0.31±0.48, n.s.). A total of more than two dysplastic features in blasts, granulocytes, and monocytes and/or a blast count >5% occurred in 68.4%, 63.3%, and 35.3% of cases with MDS, possible MDS, and non-MDS according to CM (p=0.0005). Interestingly, in some cases aberrant antigen expression has been observed in cell lineages not rated dysplastic by CM. This evaluation suggests that MFC may be used to identify dysplastic features in patients with suspected MDS. Sensitivity and specificity varies between MDS subtypes and should be clearly defined in future studies.