Effect of high-flux dialysis on the anaemia of haemodialysis patients

Background. Anaemia is one of the major clinical characteristics of patients with chronic renal failure, and has a considerable effect on morbidity and mortality. Adequate dialysis is of paramount importance in correcting anaemia by removing small and medium-sized molecules, which may inhibit erythropoiesis. However, high-molecular-weight inhibitors cleared only by means of highly porous membranes have also been found ill uraemic serum and it has been claimed from uncontrolled studies that high-flux dialysis could improve anaemia in haemodialysis patients. Methods. We therefore planned this multicentre randomized controlled trial with the aim of testing whether the use of a large-pore biocompatible membrane for a fixed 12-week follow-up improves anaemia in haemodialysis patients in comparison with the use of a conventional cellulose membrane. Eighty-four (5.3%) of a total of 1576 adult haemodialysed patients attending 13 Dialysis Units fulfilled the entry criteria and were randomly assigned to the experimental treatment (42 patients) or conventional treatment (42 patients). Results. Haemoglobin levels increased nonsignificantly from 9.5+/-0.8 to 9.8+/-1.3 g/dl (P=0.069) in the population as a whole, with no significant difference between the two groups (P=0.485), Erythropoietin therapy was given to 32/39 patients (82%) in the conventional group, and 26/35 (74%) in the experimental group (P=0.783) with subcutaneous administration to 26/32 patients in conventional and to 23/26 patients in experimental group, P=0.495. Dialysis dose (Kt/V) remained constant in both groups (from 1.30+/-0.17 to 1.33+/-0.20 in the conventional group and from 1.28+/-0.26 to 1.26+/-0.21 in the experimental group, P=0.242). Median pre- and postdialysis beta(2)-microglobulin levels remained constant in the conventional group (31.9 and 34.1 mg/dl at baseline) and decreased in the experimental group (predialysis values from 31.1 to 24.7 mg/dl, P=0.004 and post-dialysis values from 24.8 to 20.8 mg/dl, P=0.002). Median erythropoietin doses were not different at baseline (70 IU/kg/week in conventional treatment and 90 IU/kg/week in experimental treatment, P=0.628) and remained constant during follow-up (from 70 to 69 IU/kg/week in the conventional group and from 90 to 91 IU/kg/week in the experimental group, P=0.410). Median erythropoietin plasma levels were in the normal range and remained constant (from 12.1 to 12.9 mU/ml in the conventional group and from 13.2 to 14.0 mU/ml in the experimental group, P=0.550). Conclusions. This study showed no difference in haemoglobin level increase between patients treated for 3 months with a high-flux biocompatible membrane in comparison with those treated with a standard membrane. When patients are highly selected, adequately dialysed, and have no iron or vitamin depletion, the effect of a high-flux membrane is much less than might be expected from the results of uncontrolled studies.