Recombinant Erythropoietin and Survival in Myelodysplastic Syndromes.

Several reports have recently raised the question of whether the use of recombinant erythropoietin (r-EPO) may influence clinical outcome and survival of neoplastic patients who receive this drug along with the therapy for their disease. r-EPO is frequently employed for the treatment of anemia in myelodysplastic syndromes (MDS). The efficacy of r-EPO in these patients have been extensively investigated in terms of improvement of Hb levels and quality of life parameters and its use is now recommended in selected settings of MDS. However, there is very small published evidence about the long term effects of r-EPO in patients with MDS and, above all, about its possible relationship with survival. In particular, low-risk MDS, which generally do not receive cytostatic treatments, represent an interesting model to evaluate the impact of r-EPO on the natural history of these diseases. We performed a retrospective analysis on our patients with MDS which was focused on determining overall survival and causes of death in relationship to the use or not of r-EPO. Between 1986 and 2001, 346 patients affected by MDS were identified in our database. Among them, we selected 169 anemic (Hb < 10 g/dl) patients with well defined diagnosis of MDS according to FAB criteria (excluding CMMoL and RAEB-T), low-intermediate IPSS score (retrospectively assessed) and regular follow-up. Ninety-nine patients belonged to the pre-r-EPO era or had never received r-EPO for other reasons. Seventy patients had instead received r-EPO for at least 8 weeks during the course of their disease. The two groups (r-EPO no/r-EPO yes) resulted comparable for age, type of diagnosis, additional treatments, Hb levels and transfusion requirement. Among patients who received r-EPO, 22 (31.4%) achieved an hematological erythroid improvement (HI-E), according to IWG criteria. In these patients the drug was given for 5–51 months (median 19). There was no difference between the two groups in terms of leukemic evolution (19% vs 21% at 4 years) and causes of death (infections, hemorrhages, second tumors, cardiovascular events). No death could be directly attributable to r-EPO. Median overall survival was also not statistically different between patients who had received (45 months) or not (41 months) r-EPO. However, when survival was analysed according to response to r-EPO, it was found significantly longer (p < 0.02) in responders (median 54 months) than in non-responders (median 37 months), as well as than in subjects who had never received the drug. Such a difference was not modified by the time (early, during the first year from diagnosis, or later) of r-EPO administration. Response to r-EPO was associated with baseline lower levels of endogenous EPO, lower percentage of marrow blasts and higher Hb values. At multivariate analysis, however, response to r-EPO maintained an independent favourable prognostic value on survival. This is the first large study focused on the long-term clinical outcome of MDS analysed according to the role of r-EPO. Our data indicate that the prolonged administration of this drug is safe and that response to r-EPO is significantly associated with better survival in MDS patients.