Judging biocompatibility of materials with blood includes assessment of hemolysis resulting from flow in contact with those materials. Such hemolysis is influenced in part by the surface roughness and specific hydrodynamic features of the device used. Using a rotating-disk device, with polyethylene disks and human blood, it was shown that roughness under 4 microns has negligible effect and that hemolysis increases sharply for roughness above about 11 microns. Gross roughness (65 microns) causes qualitatively different hemolysis kinetics but not as severe hemolysis as extrapolated from low-roughness data. In the disk geometry, the corner is a key region which generates high hemolysis and thus minor alterations cause hemolytic variations which may tend to obscure materials influences. A series of corner-beveled polycarbonate disks were tested in comparison with the normal square corner and were found always to cause greater hemolysis. Tapering the bevel inward, so the taper angle was on the order of 2 degrees-4 degrees, reduced hemolysis relative to the 45 degree bevel but still was more hemolytic than the flat disk. Evidence suggests that glassy plastics can be machined with more reproducibility at the corners than plastics above their Tg such as polyethylene. General device design problems regarding hemolysis are discussed.

Three types of materials of special interest to the NIH Biomaterials Program were evaluated for their tendency to induce hemolysis when exposed to a laminar blood flow between rotating parallel disks. The three types were: (1) TDMAC-heparinized surfaces of polycarbonate (Lexan), silicone rubber, and polyvinylchloride; (2) polyacrylamide hydrogels (PAH) prepared by three different chemical processes; and (3) fluorinated ethylcellulose (FEC). All were compared to a polyethylene (PE) standard, to normalize data for variations in blood quality. Multiple tests, showing good reproducibility, demonstrated: FEC is a very low hemolyzer, about 60% of the PE; PAH surfaces are poorer than PE, giving 120-220% of PE hemolysis depending on fabrication and shipment history; and TDMAC-heparinized surfaces are highly hemolytic, in the range 160-440% of PE depending on substrate. Plastics used as substrates for the coating cited above were also evaluated: Delrin, Lexan, Nylon 6, propylene, and a polyether urethane. Tentative explanations are advanced for hemolytic variations, in terms of surface chemistry and material interactions with the blood.

Mechanical fragility of human red blood cells was evaluated by shearing the blood between rotating polyethylene disks and measuring the timedependent release of hemoglobin. Several blood additives were tested for their effect on this hemolysis: (a) two synthetic polynucleotides (polyadenylate and polycytidylate) as biopolymers with purine and pyrimidine moieties, respectively; (b) two low-molecular-weight purine derivatives (the drug theophylline and uric acid). It was found that polyadenylate always increased hemolysis, polycytidylate often reduced it, theophylline always reduced it, and uric acid was always ineffective. Drug localization data on theophylline showed a large uptake of the additive by cell membranes, the degree of hemolysis protection being proportional to the mass of the drug absorbed. Supplementary cell characterization by resistive pulse spectroscopy documented that the protective drugs caused cell volumes to increase, deformability to decrease, and osmotic fragility to decrease. Chemical and mechanical mechanisms for changes in mechanical fragility are proposed.