Background: The efficiency of dialysis membranes is generally evaluated by assessing their capacity to remove small, water-soluble and non-protein-bound reference markers such as urea or creatinine. However, recent data suggest that protein-bound and/or lipophilic substances might be responsible for biochemical alterations characterizing the uraemic syndrome.
Methods: In the present study, the total concentrations of four uraemic retention compounds (indoxyl sulphate, hippuric acid, 3-carboxy-4-methyl-5-propyl-2-furanpropionic acid (CMPF) and p-cresol) and of tryptophan, the only protein-bound amino acid and a precursor of indoxyl sulphate, were compared with those of urea and creatinine in pre- and post-dialysis serum and in dialysate of 10 patients; two high-flux (HF) membranes (cellulose triacetate (CTA) and polysulphone (PS)) and a low-flux polysulphone (LFPS) membrane were compared in a crossover design, using HPLC.
Results: Except for hippuric acid (67.3+/-17.5% decrease), major differences were found in the percentage removal of the classical uraemic markers on one hand (creatinine 66.6+/-7.0% and urea 75.5+/-5.8% decrease) and the studied protein-bound and/or lipophilic substances on the other (indoxyl sulphate, 35.4+/-15.3% and p-cresol 29.0+/-14.2% decrease; tryptophan, 27.5+/-40.3%, and CMPF, 22.4+/-17.5% increase; P<0.01 vs urea and creatinine in all cases). Hippuric acid removal was more pronounced than that of the remaining protein-bound compounds (P<0. 01). After correction for haemoconcentration, per cent increase of tryptophan and CMPF was less substantial, while per cent negative changes for the remaining compounds became more important. There was a correlation between creatinine and urea per cent removal at min 240 (r=0.51, P<0.01), but all the other compounds showed no significant correlation with either of these two. The three membranes were similar regarding the changes of total solute concentrations from the start to the end of dialysis.
Conclusions: Urea and creatinine are far more efficiently removed than the other compounds under study, except for hippuric acid. There are no striking differences between the HF membranes. Moreover, compared with the LF membrane these HF membranes do not appear to be superior in removing the studied compounds.