Erythropoietin for the anemia of low-risk myelodysplastic syndromes

Comment on Jädersten et al, page 803

Recombinant EPO + G-CSF is an effective treatment of anemia in 40% of low-risk MDS, often with prolonged response and without increasing the risk of progression to AML.

There is now strong evidence that erythropoietin (EPO) + granulocyte colony-stimulating factor (G-CSF) can correct anemia in about 40 % of myelodysplastic syndromes (MDSs), with a synergistic effect between the 2 drugs.^1,2  The Stanford and Nordic MDS groups showed that low pretreatment serum EPO levels (< 500 U/L) and relatively low transfusion requirement (≤ 2 RBC units/month) were favorable prognostic factors of response to EPO + G-CSF, whereas patients classified in the low or intermediate-1 groups according to the MDS International prognostic scoring system (IPSS; based on percentage of marrow blasts, number of cytopenias, and karyotype) also had better response.^1-3  However, most published studies with EPO in MDS had short follow-up and the long-term effect of EPO + G-CSF on anemia, natural disease history, and survival remained uncertain. In particular, the risk that those growth factors (especially G-CSF) could trigger progression to AML remained of concern.FIG1 

In this issue, Jädersten and colleagues report the long term outcome of 129 MDS patients treated between 1990 and 1999 with EPO + G-CSF. The response rate was 39%. In the 41 responders (out of 48) who received maintenance treatment, median response duration was 23 months (range, 3-116⁺ months). Responses were more durable in patients with no excess of marrow blasts (RA and RARS), with an IPSS of low or intermediate 1 and in complete responders. Many patients required only low-maintenance doses of EPO and G-CSF (the latter could be stopped in several cases). Relapses that occurred on treatment were associated with disease progression in 30% of the cases, but remained unexplained in the other patients. Responders to EPO + G-CSF had better survival rates than nonresponders, and only 1 of the 20 patients with a response duration greater than 2 years progressed to AML.

The authors also compared their 129 patients with 334 well-matched untreated MDS patients selected from the cohort that served to establish the IPSS.³  They found that the incidence of progression to more aggressive disease, especially AML, and survival were similar in the 2 patient populations (Figure 3 of Jädersten et al). Although the comparison did not result from a randomized study, these findings are reassuring. Indeed, EPO, with or without G-CSF, is largely administered worldwide to improve anemia in MDS.

The authors finally confirm that, in high- and intermediate-2–risk MDS (according to IPSS), EPO, with or without G-CSF, is of limited use. In low- and intermediate-1–risk MDS patients, on the contrary, EPO remains a strong contender for new investigational treatments of anemia (with the probable exception of MDS with the 5q–syndrome, where responses to lenalidomide appear clearly superior).⁴  Whether novel erythropoiesis-stimulating proteins with prolonged half-life, like darbepoietin alpha, will further improve the results obtained with human recombinant EPO, as recently suggested,⁵  will have to be confirmed. ▪