AbstractAbstract 4958Erythropoietin (EPO) have been widely employed in the treatment of patients with low-risk Myelodysplastic Syndromes (MDS) and anemia, with response rates ranging from 30 to 60%. These data, however, have been derived only from controlled clinical trials or unicentric single-arm studies; it is still lacking a wider survey evaluating the use of EPO in the real-life clinical practice. To address this issue, the Gruppo Romano Mielodisplasie (GROM) revised retrospectively 394 MDS patients (M/F 225/169, median age at diagnosis 73. 9 yrs, IR 67. 0 – 79. 3) treated with EPO from 1/2002 to 12/2010 by 11 Hematological Centers (5 university hospitals and 6 community-based hospitals) in the metropolitan area of Rome. According to WHO classification, there were 81 (20. 6%) patients with RA, 7 (1. 8 %) with SA, 160 (40. 7%) with RCMD, 17 (4. 3%) with RCMD-S, 75(19. 0%) with RAEB-1, 27 (6. 8%) with RAEB-2 and 27 (6. 8%) with isolated del5q. The IPSS score was calculated in the 307 patients with an available karyotype: 145 (47. 2%) patients were low-risk, 135 (44. 0%) int-1, 24 (7. 8%) int-2 and 3 (1. 0%) high-risk. Median interval from diagnosis to EPO start was 3. 7 months (IR 0. 9 – 12. 1). At EPO start, median age was 74. 5 yrs (IR 68. 3 – 79. 9) with a median haemoglobin level of 8. 9 g/dl (IR 8. 2 – 9. 6). Creatinine level was elevated in 64 (16. 2%) cases: 138 patients (35. 3%) had a previous transfusion requirement. Median serum EPO level was 50. 0 mU/L (IR 26. 2 – 110. 0). The initial doses of EPO were ≤ 40. 000 UI/week in 259 patients (65. 7%) (standard doses, α-EPO in 104 patients, β-EPO in 143 patients, darbepoietin in 12 patients) and 80000 UI/week in 135 patients (34. 3%) (high doses, α-EPO in 130 patients, β-EPO in 5 patients). An erythroid response was observed in 228 (57. 9%) patients, with Hb increase > 1. 5 g/dl in 210 patients (53. 3%) and disappearance of transfusion requirement in 18 (4. 6%): patients receiving initial high doses had a higher response rate compared to patients receiving standard doses [94/135 (69. 6%) vs 134/259 (51. 7%), p=0. 002]. Only 5 thrombotic events (1. 2%) were reported during the treatment. Predicting factors for erythroid response were no previous transfusion requirement (p<0. 001), serum EPO levels at baseline < 50 mU/l (p<0. 001), creatinine levels above the normal values (<0. 001), ferritin levels < 250 ng/ml (p=0. 009) and hemoglobin level at baseline > 8 g/dl (p=0. 017). Median overall survival from EPO start was 70. 7 months (CI 95% 52. 5 – 88. 8) in responders versus 41. 7 months (CI 95% 27, 6 – 55, 7) in resistant patients (p= 0. 018). Our real-life data from a single homogeneous geographic area outline that EPO treatment is safe and effective also in the current clinical practice, beyond controlled clinical trials. However, this latter type of studies is warranted to define the best initial dose of EPO in such patients. Disclosures:No relevant conflicts of interest to declare.