Evaluation of membranes for plasmapheresis.

Commercial flat-sheet microporous membranes were evaluated for potential use in plasmapheresis with a specially designed filtration module. Significant differences in filtration rates were observed with different membranes. Saline filtration data were not useful in predicting the capacity of the membranes to filter plasma from whole blood. For all membranes studied, no rejection of plasma proteins was detected. No activation or deactivation of clotting factors was detected as a result of filtration. In addition, little or no hemolysis was caused by filtration with the various membranes. Saline, cell-free plasma, platelet-poor plasma, and whole blood were perfused over a track-etched membrane and the resulting filtration rates were compared. The cell-free plasma filtration rate decreased significantly with time, probably owing primarily to protein adsorption in the membrane pores. Cell-free plasma and saline filtration data were used to calculate an apparent adsorbed layer thickness in the membrane pores. Perfusion of platelet-poor plasma and whole blood resulted in time-dependent filtration rates that were much lower than those obtained when cell-free plasma was perfused. Results of the study support recent theoretical models that postulate that the rate-limiting process for blood filtration is the formation of a layer of blood cells (particle polarization) on the membrane surface.