Experimental studies, supporting a protective role of erythropoietin on vascular disease, has not been sustained by three clinical trials “Sing AK, NEJM 2006” [1] (full text), “Drüeke T.B, NEJM 2006” [2] (full text), “Pfeffer, M.A, NEJM 2009” [3] (full text),  performed in advanced chronic kidney disease patients showing that the target of normal haemoglobin levels does not improve chronic renal disease and overall morbidity. It is probable that low, earlier doses of erythropoiesis stimulating agents may be much more effective to improve outcome of chronic kidney patients than high, later doses. Large retrospective studies may be useful not only to analyse any sort of differences in anemia correction, but also to underline the most important factors which may change the response to erythropoietin treatment, giving new ideas for future prospective trials.

Aims and method

A large, multicenter, retrospective study was performed in 4 Lombardy Renal Units to evaluate anemia status of dialysis patients with particular attention to the hemoglobin target, erythropoiesis stimulating agents doses and methods of administration, iron parameters, amount of iron supplementation and dialysis efficiency. Moreover, we carried out a subgroup analysis of vitamin B replete patients to evaluate a supposed relationship between high homocysteine levels, low methylation potential and low hemoglobin levels.


446 patients, 254 males and 152 females, participated to the study. Mean age was 68 years, while mean dialysis age was 74 months. All centers reached the hemoglobin target, with a mean value equal to 11,4 g/dl. Patients with high urea kt/V (mean value of 1,48) and with subcutaneous administration of erythropoietin had a lower consumption of erythropoietin than patients with normal urea kt/V (mean value of 1,13) and with intravenous administration of erythropoietin, respectively 6742 vs. 11394 IU/week (p< 0,01) and 6397 vs. 10613 IU/week (p< 0,01). Patients with severe hyperhomocysteinemia (group C) not only had higher hemoglobin levels than patients with moderate hyperhomocysteinemia (group B) and normal homocysteine values (group A), but also lower doses of erythropoietin than patients of group A and B, and thus a lower erythropoietin resistivity index (resistivity index: erythropoietin/body weight*hemoglobin) than patients of group A and B. No differences were found for folic acid and vitamin B12 between the 3 groups. Hyperhomocysteinemic patients had higher platelet count than patients of group A (figure 1)


Erythropoietin subcutaneous administration and high dialysis efficiency were associated with lowered erythropoiesis stimulating agents doses. Dialysis patients with severe hyperhomocysteinemia had lower erythropoiesis stimulating agent resistivity index than patients with normal or moderately high homocysteine levels. These last results seem contrary to some experimental studies which show an inverse association between homocysteine and erythropoietin synthesis.