RAEB-1 with Erythroid Hyperplasia and Erythroleukemia Share Clinico-Biological Features and Outcome: A Same Disease with Different Labels?

Introduction: Based on the 2008 World Health Organization classification (WHO 2008), erythroleukemia is defined by the presence of ≥50% erythroid precursors in bone marrow (BM) and ≥20% myeloblasts in the non-erythroid cell population. Multilineage dysplasia is almost always present with high rates of MDS-like cytogenetic abnormalities, specially complex karyotypes. Therefore an extensive comparison with myelodysplastic syndromes (MDS) with ≥50% erythropoesis seems crucial to elucidate whether erythroleukemia and MDS with erythroid hyperplasia should be considered as different biological entities.

Aim: To elucidate this issue, the outcome and cytogenetic alterations of erythroleukemia patients were studied and compared to MDS patients with ≥50% erythropoesis with <5% BM blasts (RA, RARS, CRDM, MDS-U) or those with ≥5%-<10% (RAEB-1). In this subset of patients, the diagnosis of RAEB-2 is not possible because those with ≥50% erythropoesis and ≥10% BM blasts were formally diagnosed with erythroleukemia when the blast percentage was assessed in the non-erythroid cell population.

Methods: We retrospectively analyzed 448 de novo MDS with ≥50% erythropoesis and 59 de novo erythroleukemias from the MDS Spanish registry (RESMD). Diagnosis was done according to WHO 2008 and patients with ≥80% erythropoiesis with less than 20% of myeloblasts in the non-erythroid cell compartment were excluded assuming a diagnosis of pure erythroid leukemia.

Results: Median age of presentation was 74 years (25-94 years), median follow-up was 29.4 months, 63% were males. Median overall survival (OS) of MDS patients with ≥50% erythropoiesis and <5% of BM blasts (n=389; group-1) was significantly longer than MDS with ≥50% erythropoiesis and ≥5%-<10% (n=59; group-2/RAEB-1) (69 months vs. 18 months, p<0.001). Although erythroleukemia patients (n=59) presented a shorter median OS than group-1 patients (69 months vs. 14.5 months, p<0.001), there was no significant differences compared to group-2 patients (RAEB-1) (18 months vs. 14.5 months, p=0.679). Figure 1. Percentage of abnormal karyotypes was significantly higher in the group-2 and EL vs. group-1 but there was no significant differences between group-2 and erythroleukemia (56.9% vs. 44.1%, p=0.165). Moreover no significant differences were observed in the percentage of high-risk karyotypes defined by the IPSS (complex karyotype, chromosome 7 abnormalities) between RAEB-1 and erythroleukemia (30.5% vs. 23.7%, p=0.408). Finally, the presence of a high-risk IPSS karyotype was capable to discriminate two risk groups in the subset of patients with ≥5% BM blasts (RAEB-1 and erythroleukemia). Figure 2.

Conclusion: Erythroleukemia and RAEB-1 with ≥50% erythropoiesis share clinico-biological features and outcome. Our findings suggest that erythroleukemia is a continuum of MDS with erythroid hyperplasia and karyotype rather than an arbitrary blast cut-off is the main prognostic marker in this subset of patients.

 Figure 1.

Figure 2.