Solomon and colleagues assessed the relationship among the initial haemoglobin response to darbepoetin after two weight-based doses, the haemoglobin level achieved after 4 weeks, the subsequent darbepoetin dose and outcomes in 1872 patients from the TREAT trial who were randomized to darbepoetin.15 The initial dose of darbepoetin was 0.75 µg/kg Selleck Palbociclib of body weight and was repeated after 2 weeks if haemoglobin values did not exceed 140 g/L. Poor initial response to darbepoetin was defined as the lowest quartile of per cent change in haemoglobin level (<2%) after the first two standardized doses of the drug. Patients in the lowest quartile of haemoglobin responsiveness were more likely to have cardiovascular

disease, high CRP levels and low ferritin and transferrin saturation levels. The average haemoglobin level after 12 weeks remained marginally but statistically significantly lower among patients with a poor initial response (122 ± 9 g/L) than among those with a better initial response (124 ± 7 g/L, P < 0.001). The average monthly dose of darbepoetin after 12 weeks and throughout the remainder of the trial was substantially higher among patients a poor initial response (median dose, 232 µg; interquartile range, 126 to 390) than those with a better initial response (167 µg; interquartile

range, 95 to 310; P < 0.001). There was significant difference CB-839 in the use of intravenous iron or blood transfusion throughout the trial. Compared with patients with a better initial response, those with a poor initial response were at increased risk of a cardiovascular composite event (HR 1.31, 95% CI 1.09–1.59) and all-cause death (HR 1.41, 95% CI 1.12–1.78). The event rates for the cardiovascular composite outcome and all-cause death in the better initial response group

were comparable with oxyclozanide the placebo group. These findings indicate that requirement of high-dose ESA to achieve target haemoglobin rather than achieved haemoglobin may be responsible for the poor outcome. Interestingly, the event rates for stroke were comparable in the two response groups, but higher in both groups than in the placebo group. It still remains unclear whether the use of ESA or high haemoglobin target contributed to increased risk of stroke in patients treated with darbepoetin. A summary of the observational studies is provided in Table 2. In a US Medicare study of 75 283 prevalent patients receiving haemodialysis between July–December 1993, a haematocrit level of 33–36% was associated with a similar risk of mortality (adjusted RR 0.96, 95% CI 0.91–1.01) compared with a reference haematocrit level of 30–33%.16 In contrast, lower haematocrit levels were associated with increased risk of mortality (haematocrit <27% RR 1.33, 95% CI 1.26–1.40; haematocrit 27–30% RR 1.12, 95% CI 1.08–1.17). The pattern of higher mortality with lower haematocrit was similar in diabetic and non-diabetic patients.