The biocompatibility of artificial organs

Thromboembolic events in very short-term ventricular assistance have been uncommon. However, patients who are bridged or have long-term artificial heart implantations have almost uniformly experienced serious thromboembolic sequelae. Anticoagulation regimens for this form of circulatory assistance have varied but results are inconclusive. Intravenous heparin is commonly used, while low molecular weight Dextran, aspirin, and other antithrombotic agents are occasionally used. Mechanical cardiac valve failures requiring reoperation are due primarily to thrombosis, tissue overgrowth, and structural failure. Reoperations of tissue valves indicate sterile degeneration, usually related to calcification and tears, as the principal cause of failure. Despite improvements in biomaterials and an appreciation of the influence of disordered flow on graft patency, antithrombotic therapy represents the most clinically relevant approach to prevention of occlusion of vascular prostheses. The location of neointimal hyperplasia in vascular prostheses is different from that in vein grafts, the effect concentrating at the anastomosis. No clear rationale exists for antithrombotic therapy for hyperplasia in this instance. However, a rationale does exist for early antithrombotic therapy to reduce surface thrombogenicity. The lessons learned from clinical trials of antithrombotic therapy in the case of vein grafts might be profitably studied with regard to synthetic grafts and other artificial organs. Early low-dose aspirin and, possibly, dextran-40 treatment are justified in this case. However, continuation of these therapies past the early healing phase does not appear warranted. Clinical and experimental studies with hemodialysis and other membranes suggest complement activation through the alternate pathway is commonly observed; the biologic responses are consistent with the known properties of C5a. Factors governing anaphylatoxin exposure can be ascribed to the nucleophilicity and number of surface reactive groups. Elimination of surface reactive groups, or binding of bystander plasma proteins such as albumin, may be expected to limit complement activation.