EHA Library - The official digital education library of European Hematology Association (EHA)

Myelodysplastic syndromes (MDS) are a heterogeneous entity characterized by dysplasia, hypercellular bone marrow, cytopenias and a risk of transformation to acute leukaemia (AL). Treatment is mainly based in the International Prognostic Scoring System (IPSS). In the last few years a lot has been written about MDS prognosis, with new scores for survival increasing in complexity making their use in everyday practice somewhat difficult. Although other prognostic factors, for instance, bone marrow fibrosis, LDH and β2-microglobulin elevation are readily available to clinical practice, others like gene mutations TP53, EZH2, ETV6, RUNX1, NRAS, ASXL1, SRSF2, U2AF1 and SF3B1 are not so easy to obtain. Two retrospective studies show a benefit in survival in patients treated with erythropoietin (Epo)±G-CSF, however without an impact on progression to AL.

We analyzed the records of 102 patients with de novo MDS between October 2009 and March 2014. Clinical and laboratorial data were collected and overall survival (OS) was estimated stratified according their serum Epo level: low vs normal vs <100 mUI/mL vs 100-500 mUI/mL vs ≥ 500 mUI/mL.

The patients had median age of 74 years, with a Male:Female ratio of 0.8. According to the WHO classification, MDS subtypes where distributed as follows: Refractory Cytopenia with Multilineage Dysplasia (RCMD) (n=52), Refractory Cytopenia with Unilineage Dysplasia (RCUD) (n=12), Refractory Anemia with Excess Blasts (RAEB)-1 (n=8), (RAEB)-2 (n=8), Refractory Anemia with Ring Sideroblasts (RARS) (n=6), 5q- Syndrome (n=4) and Chronic Myelomonocytic Leukaemia (n=12). Thirty seven patients presented with low IPSS score, 39 with intermediate-1, 10 with intermediate-2 and 1 with high score. The mean erythropoietin was significantly lower in RCMD patients when compared to other MDS subtypes (p<0.05). Eleven patients (7 RAEB-2, 2 RCMD, 1 RAEB-1 and 1 CMML) progressed to AL. Their mean erythropoietin value was higher than in patients who did not evolve (p<0.05). We found that patients with elevated serum Epo had significantly lower survival rates than those with normal Epo levels (p=0.0336). The predictive value of serum Epo was maintained after Cox regression adjustment for LDH, age, IPSS, R-IPSS, hemoglobin, neutrophils, platelets, ferritin, β2-microglobulin and bone marrow blasts. In multivariate analysis, the independent survival predictors were serum Epo (p<0.001), ferritin (p<0.001), β2M (p=0.008) and marrow blast percentage (p<0.001).

In the present study, serum baseline erythropoietin was identified as in independent poor prognostic factor, with higher values in patients with progression to AL and decreased OS.