Independent confirmation of the immunophenotypic ELN Scoring System in patients with MDS and CMML

We read with great interest the study by Veenstra et al.1 validating the 17 immunophenotypic core marker panel (ELN Scoring System), defined by the ELN-iMDS-Flow Working Group (ELN-iMDS).2 It includes aberrancies in myeloid progenitor cells (MPCs), neutrophils, monocytes, and nucleated erythroid cells. The presence of at least three aberrancies was indicative of either myelodysplastic neoplasm (MDS) or chronic myelomonocytic leukemia (CMML). Veenstra et al. could diagnose MDS with a sensitivity of 90% compared to an integrated diagnostic approach, which was also maintained within low-risk MDS (<5% bone marrow blasts) at 87%. This is comparable to the established comprehensive Integrated Flow Score (iFS)3 (all MDS: 91%) and represents a significant improvement compared to the 4-parameter Ogata-score4 (all MDS: 66%). In pathological controls (i.e., non-clonal cytopenias), concordance of ELN Scoring System was achieved in 76% (41/54) of patients. Some aberrancies were restricted to MDS patients but absent in pathological controls (aberrant expression of CD5, CD7, or CD56 on MPC and abnormal expression of CD33 on neutrophils). Exclusion of four markers predominantly associated with CMML (low side scatter and aberrant CD33 expression on neutrophils, aberrant percentage and CD13 expression of monocytes) (Kern et al.2) increased specificity to 96%. We aimed to confirm the applicability and therefore the significance of the ELN Scoring System in routine diagnostics. To do this, we analyzed a large, independent cohort of MDS (359 patients) and pathological controls (41 patients with non-clonal cytopenias, for details see legend of Figure 1). In addition to the study by Veenstra et al., samples of 38 CMML patients and 32 healthy bone marrow (HBM, hip surgery patients) were included. The antibody panel, staining, acquisition, and data analysis have been previously delineated.5 Applying the ELN Scoring System to our MDS cohort, the diagnosis was confirmed in 89% of all MDS cases (Figure 1A). High sensitivity was also maintained in the low-risk MDS group (<5% blasts: 84%; low-to-moderate low IPSS-M: 86%). These results were in line with those reported by Veenstra et al. Regarding false negative cases (38/359), 42% of these would have been assigned to MDS by alternative scoring systems (Ogata-score: 8%, iFS: 21%, both simultaneously: 13%). Almost all CMML samples (97%) had an ELN Score compatible with MDS/CMML, in line with the ELN-iMDS study by Kern et al.2 We also confirmed the four parameters described above as significantly associated with CMML. As determined in pathological controls, the specificity of the ELN Scoring System was notably higher in our cohort than in the Veenstra and Kern studies (98%, 75%, and 78%, respectively). However, this high specificity was confirmed in HBM (97%) (Figure 1B). A possible explanation for the higher specificity in our study might be the different composition of the pathological control cohort. For example, the proportion of patients with iron deficiency in whom the presence of dyspoietic changes has been described,6, 7 was significantly lower in our study compared to Veenstra et al. (15% vs. 37%). Furthermore, each group uses laboratory-specific reference values. For example, the cutoff for CD71 (coefficient of variation in %) in erythroid cells is very different in the two studies (84% vs. 62% in the Veenstra study), which might impact the specificity of the score. Considering the individual aberrancies (Figure 1C), we can verify the high prevalence of abnormal CD117 on MPC as well as the abnormal CD13/CD16 expression on neutrophils. Furthermore, we can confirm the high specificity of an abnormal percentage of MPC and neutrophils, cross-lineage expression of CD5, CD7, or CD56 on MPC, and an abnormal CD33 expression on neutrophils. In summary, the sensitivity of the ELN Scoring System validated in the Veenstra study1 is reproducible in our independent cohort with an even higher specificity in the present study. It also performs well in CMML patients (see Figure 1A). Using a combination of the ELN Scoring System and the Ogata-score could reduce false negative results. Notably, the Ogata-score can be assessed without additional measurements. Overall, the ELN Scoring System may represent a step toward the development of a robust and widely applicable flow cytometric approach as a part of integrated MDS diagnostics.8 The authors would like to thank all the technicians in the flow cytometry laboratory for processing the samples and acquiring the raw data, and Dirk Schweigler for data bank management. Uta Oelschlaegel: Conceptualization; investigation; writing—original draft. Jonas Schadt: Investigation; writing—review and editing. Katharina Epp: Investigation; writing—review and editing. Lisa Wagenführ: Investigation; writing—review and editing. Leo Ruhnke: Investigation; writing—review and editing. Frank Kroschinsky: Writing—review and editing; supervision. Martin Bornhäuser: Writing—review and editing; supervision. Katja Sockel: Investigation; writing—review and editing. Malte von Bonin: Investigation; writing—review and editing; supervision. Maximilian Alexander Röhnert: Conceptualization; investigation; writing—review and editing. The authors declare no conflicts of interest. This research received no funding. The data that support the findings of this study are available from the corresponding author upon reasonable request.