The effect of recombinant human erythropoietin on platelet counts is strongly modulated by the adequacy of iron supply

The effect of recombinant human erythropoietin (rHuEpo) on megakaryopoiesis remains controversial. Treatment with rHuEpo in renal failure patients has been associated with a slight elevation of platelet counts. In animal stud ies, high doses of rHuEpo produced an increase of platelet counts followed by a gradual return to normal after 7 to 15 days or even a substantial degr ee of thrombocytopenia. However, because iron deficiency is also known to b e associated with thrombocytosis, (functional) iron deficiency during rHuEp o could be contributing to these observations. We investigated the impact o f iron supply on changes in platelet counts induced by rHuEpo. Rats were ei ther fed normal food (normal rats) or received 1% carbonyl iron for 2 weeks or 3 months, as well as during the experiment, to achieve iron supplementa tion or overload, respectively. Rats of all three categories then received daily intravenous injections of rHuEpo (10, 50, or 150 U) or normal saline (0 U) for 20 days. With 0 to 10 U rHuEpo, platelets remained stable. In nor mal rats receiving 50 to 150 U rHuEpo, platelets increased to 120% to 140% of baseline at 4 to 12 days to level off at 120% at 16 to 20 days. This res ponse was less sustained in splenectomized animals. Iron-supplemented rats receiving 50 to 150 U rHuEpo also increased platelets initially, but the pe ak was at day 4, followed by a gradual return to baseline and even a modera te thrombocytopenia later on. Iron-overloaded rats receiving 50 to 150 U rH uEpo also had increased platelets at day 4, but the duration of platelet in crease was shorter, and they experienced a more pronounced degree of thromb ocytopenia in proportion to the dose of rHuEpo. Because the early elevation of platelets was of larger magnitude than hematocrit changes, it is unlike ly that it could be accounted for by shrinkage of plasma volume. Because it was observed in all three iron conditions, there appears to be some direct positive effect of rHuEpo on platelet production. However, after this tran sient effect, expanded erythropoiesis appears to exert a negative impact up on platelet production. Secondary thrombocytopenia was not related to splen ic pooling, and its very slow correction after cessation of rHuEpo therapy is not compatible with changes in platelet survival. Rather, it is consiste nt with stem cell competition between erythroid and megakaryocytic developm ent. However, this secondary thrombocytopenia is masked by (functional) iro n deficiency in rats not receiving an adequate iron supply from food or sto res. (C) 1999 by The American Society of Hematology.