Thromboembolic and Infectious Complications of Total Artificial Heart Implantation

Thromboembolic and infectious events were found to be major complications of long-term total artificial heart implantation in two patients. Similar complications have been reported in other patients, as well as in animal studies. The thromboembolic events and the infectious complications appear to be interrelated. On the one hand, thrombi located on the valves and at the vascular anastomoses of the artificial heart were found to be infected at autopsy; such infections are known to exacerbate formation of thromboemboli. On the other hand, the generation of microthrombi may have contributed to the RES blockade seen in our patients. We hypothesize that this RES blockade led to a progressive decrease in lymphoid system function and impaired the patients' capacity to clear microorganisms from the circulation. These phenomena arose, in part, from the design of the artificial heart and were exacerbated by associated therapy, such as blood transfusions. Our data suggest several measures that might be taken in order to reduce the severity of both the thrombogenic and infectious complications. Improved anticoagulation regimens, which increase the ability of the physician to maintain the proper balance between thrombotic and hemorrhagic potential, are needed. This may require not only improved methods of monitoring anticoagulation and predicting changes in the effectiveness of various agents as other events supervene, but also new anticoagulant and antithrombotic drugs, for example, low molecular weight heparins and prostacyclin derivatives. It is also clear that the design of the artificial heart should be modified in order to improve fluid dynamics so that they will approximate as closely as possible those of the natural heart. This includes redesigning the mounting of the valves to eliminate crevices and discontinuities that allow stagnant flow and predispose to thrombus formation as well as imposing a dP/dt that minimizes shear-related hemolysis, thereby minimizing the need for blood transfusions. Prevention of infections presents a more difficult problem. Transcutaneous lines (regardless of their use) are an obvious route for infection, and attention should be given to minimizing the number and length of use of monitoring lines. However, until a totally implantable drive system is available, the drive lines will remain a potential avenue for the introduction of infections. The risk may be minimized by rigorous attention to care of the exit sites and by improved