BACKGROUND: The potential risks of tumor growth promotion and thromboembolism associated with erythropoietin (Epo) therapy warrant cautious use of erythropoiesis stimulating agents (ESA). Recent data suggest a surprising and significant association between ESA use and increased incidence of leukemic transformation in primary myelofibrosis (PMF) (data from an analysis separately submitted for American Society of Hematology meeting, 2007). It is prudent, therefore, to avoid unnecessary ESA exposure in PMF, especially in patients with low likelihood of treatment response.

METHODS: Diagnosis of PMF was based on World Health Organization criteria. Study inclusion required availability of treatment history with ESA and adequate follow-up. We excluded patients who were receiving hemoglobin-improving agents concomitantly with an ESA, such as prednisone, androgens, danazol, or thalidomide. Transfusion dependency was defined by a history of minimum 2 units of red blood cell transfusions in a 1-month period for a hemoglobin < 8.5 g/dL, not associated with bleeding. An adequate treatment response was defined as a minimum 2.0 g/dL increase in hemoglobin level or becoming transfusion-independent over a minimum 1-month period.

RESULTS: A master database of PMF patients seen at the Mayo Clinic from 1976 through 2006 was queried to identify 43 patients (median age 67 years; 35% females) treated with ESAs and in whom adequate follow-up was available to determine response. Prior to ESA treatment, median hemoglobin was 9.2 g/dL (range 7.0–11.9), and 16 (37%) were red blood cell transfusion-dependent. In 25 patients, pre-treatment serum Epo level was available (median 34 mU/mL; range 2–323). Among the 43 study patients, 16 (37%) were previously treated with other PMF-specific therapies, and 18 (42%) were receiving cytoreductive treatment concurrent with an ESA. Overall, 23% of patients (10 of 43) responded to treatment with ESA. The response rate was 0% (0 of 16) in the transfusion-dependent group compared to 37% (10 of 27) in the transfusion-independent group (p=0.007). Among the transfusion-independent patients, the proportion of ESA responders was significantly higher in the presence of a baseline hemoglobin level < 10 g/dL prior to treatment (p=0.009). ESA response was not correlated with baseline serum Epo level in the context of either all 43 study patients (p=0.68) or the 27 transfusion-independent patients (p=0.39). ESA response was not affected by PMF-specific treatment history (p=0.46), use of concurrent cytoreductive therapy (p=0.28), cytogenetic findings (p=0.62), JAK2V617F presence (p=0.69) or its allele burden at time of initial PMF diagnosis (p=0.16).

CONCLUSIONS: Clear benefit of ESAs in PMF is limited to red blood cell transfusion-independent patients and those with baseline hemoglobin levels less than 10 g/dL.

Author notes

Disclosure: No relevant conflicts of interest to declare.