Synthetically modified cellulose: an alternative to synthetic membranes for use in haemodialysis?

Renal replacement therapy relies predominantly on the use of cellulose-based membranes. Such membranes have a biocompatibility profile which is inferior to membranes manufactured from synthetic polymers. Synthetically modified cellulose (SMC) is a new, low-flux haemodialysis membrane in which hydroxyl groups have been replaced with benzyl groups. The biocompatibility profile characterized by changes in white cell and platelet counts and the activation of complement components (C3a, C5a and C5b-9) have been studied in vivo and compared with those of cellulose acetate, unmodified cellulose (Cuprophan(C)) and low-flux polysulphone (Fresenius Polysulfone(TM)) in the same group of patients. For SMC, the white cell count at 15 min declined to 65.6% of pretreatment level, compared with 63.8% for the cellulose acetate, 79.6% for low-flux polysulphone and 28.1% for Cuprophan(C),thereafter returning to pretreatment levels. Both modified cellulose membranes were superior to unmodified cellulose (P = 0.001); the differences between the modified cellulose membranes were not significant statistically. The changes induced by all three cellulose-based membranes exceeded those for low-flux polysulphone (P = 0.001). Associated with the neutropenia was a reduction in platelet count, but this was independent of membrane type. The mean time-averaged concentrations of C3a(des Arg) over 150 min were 1168 ng ml(-1) (SMC), 1030 ng ml(-1) (cellulose acetate), 1297 ng ml(-1) (Cuprophan) and 790 ng ml(-1) (low-flux polysulphone). Equivalent values for C5a(des Arg) were 6.12 (SMC), 2.98 (cellulose acetate), 11.03 (Cuprophan) and 1.33 ng ml(-1) (low-flux polysulphone). C5b-9 values were 385 (SMC), 386 (cellulose acetate), 177 (Cuprophan) and 185 ng ml(-1) (low-flux polysulphone). For each of the complement components the differences between the membranes were significant [P = 0.0009 (C3a(des Arg)), P = 0.0001 (c5a(des Arg) and C5b-9)]. The levels of C5b-9 generated during dialysis also showed a significant positive correlation compared to C5a for all membranes considered as a single group (Pearson's correlation coefficient = 0.870, P = 0.0001). It is concluded that the modification of the cellobiosic unit is a promising approach to improve the biocompatibility profile of cellulose-based membranes. The two different methods of modification lead to similar improvements in biocompatibility compared with unmodified cellulose, but as yet do not match that of low-flux polysulphone. (C) 1997 Elsevier Science Limited. All rights reserved.