02, 95% CI 1.01–1.03 (P < 0.001). Most CKD patients treated with ESA require concomitant iron supplementation, particularly when targeting higher haemoglobin levels. This raises the intriguing

possibility that iron therapy may be an important effect modifier contributing to the complex relationship between INCB024360 ESA dose, haemoglobin level and clinical outcomes. Previous epidemiologic data have linked augmented body iron stores and/or increasing IV iron doses with heightened risks of both cardiovascular disease28–30 and bacterial infections,31 although other studies have refuted these findings.32 High ferritin and low transferrin saturation values have similarly been associated with increased mortality,33,34 but these traditional iron markers may have been confounded

by non-iron-related conditions, such as infection, inflammation and protein-energy malnutrition. The effect of iron therapy on mortality has not been systematically CH5424802 studied in an ESA RCT and patients with iron deficiency or iron overload were specifically excluded from the four largest ESA trials. In the Normal Haematocrit Cardiac Trial, more patients received intravenous iron in the normal haematocrit group than in the low haematocrit group (85.1% vs 75.4%, P < 0.001), although serum ferritin levels at 12 months were lower in the former (391 ± 424 vs 503 ± 442 ng/mL, P = 0.005) and transferrin saturation values were comparable between the two groups.9 The odds ratio of mortality for patients in the normal haematocrit group who received intravenous iron dextran during the 6 months before death or censoring was 2.4 compared with those who did not receive intravenous iron (P < 0.001). During the 6 months period before death, the average doses of intravenous iron dextran

in the normal and low haematocrit groups were 214 ± 190 and 145 ± 179 mg/4 weeks period, respectively. On the other hand, more patients in the placebo group received intravenous iron than in the darbepoetin group in the TREAT trial (20.4% vs 14.8%, P < 0.001).10 In the CREATE trial, 52% and 42% of patients in high and low haemoglobin groups received at least one dose of intravenous iron.14 Similarly, overall use of iron was comparable Thalidomide in high (52%) and low (48.3%) haemoglobin groups in the CHOIR trial.12 None of these RCTs provided more data on iron therapy, iron studies and outcomes. Consequently, based on trial information to date, there is insufficient evidence to conclude whether iron loading contributed to the poorer outcomes associated with targeting higher haemoglobin levels with ESA. Currently, there is a reasonable body of evidence to indicate more harm than benefit from targeting higher haemoglobin levels with ESA therapy. Patients requiring higher doses of ESA experience increased mortality at any haemoglobin level and patients achieving target haemoglobin levels have better outcomes than those who fail to achieve.