Introduction

For nearly 50 years, the functional replacement of the human heart has received extraordinary interest that stimulated the development of mechanical circulatory assist technology. Given the significant technical hurdles, early failures, and limitations in technology to achieve a practical long-term total replacement heart, research into alternative approaches to total replacement of the heart led to successful development of partial functional support with ventricular assist devices for both short- and long-term applications.

In the late 1980s, implantable pneumatic and electrically driven left ventricular assist devices (LVADs) were introduced into clinical investigation in the United States (1). These early devices were large volume displacement pumps implanted below the diaphragm and connected to the apex of the left ventricle with an inlet cannula and directed to the ascending aorta with an outflow graft.  Both pneumatic and electrically driven models utilized a percutaneous drive-line that served both as a method to deliver air or electricity to drive the pump and also function as a venting tube to eliminate the need for an internal displacement diaphragm. Early clinical successes with the first generation of implantable LVADs led to approval by the U.S. Food and Drug Administration (FDA) for use of these devices for bridge-to-transplant indications.  Following this successful application, one of these devices, the HeartMate VE was used in the REMATCH (Randomized Evaluation of Mechanical Assistance for Treatment of Congestive Heart Failure) trial that led to successful use of these devices for destination therapy (permanent pump use) (2).  Despite early successes, these devices were large and noisy and were associated with significant adverse events including stroke, bleeding, device infection, and device malfunction that limited application of these devices to patients in cardiogenic shock or who were inotrope-dependent with evidence of failing organ function (1,2).

In the early 1990s, an evolutionary step in the development of LVAD therapy occurred with the introduction of continuous-flow rotary pumps with axial centrifugal design (3). These pumps were significantly smaller in size, had fewer moving parts, and were quiet in operation. Early studies with both axial- and centrifugal-flow rotary pumps demonstrated improved mechanical reliability and fewer device-related infections and neurological injury leading to greatly improved survival (3,4). The introduction of rotary pump technology significantly increased adoption of LVAD therapy for both bridge-to-transplant and, particularly, destination-therapy indications (4).  Despite improvements in clinical outcomes with continuous-flow rotary pumps with axial design and mechanical pivot to support the internal rotor, these pumps demonstrated a propensity for  pump thrombosis that again slowed further adoption of this therapy (5).

Technological advancements in the field have continued with introduction of continuous-flow rotary pumps with bearing-less designs that incorporate various forms of hydrodynamic and magnetic levitation or total magnetic levitation of the internal impeller. Recent clinical studies have demonstrated superiority of these designs to previous continuous-flow rotary pumps using axial design with a mechanical pivot. The HeartMate 3 is a continuous-flow rotary pump with centrifugal design incorporating total magnetic levitation of the internal impeller (6,7). This device also incorporated a sintered surface throughout the pump to promote a pseudo-intimal biological lining of the pump. Further, the pump employs a pulse in the speed control algorithm that is designed to alter blood flow pathways in the pump to prevent areas of blood stasis. The superiority of this pump design was validated in the MOMENTUM 3 trial (Multicenter Study of MagLev Technology in Patients Undergoing Mechanical Circulatory Support Therapy with HeartMate 3) where the HeartMate 3 was demonstrated to be superior to the HeartMate II pump with a near zero percent risk of pump thrombosis and significantly lower risk of stroke (6,7).

One of the major concerns with continuous-flow rotary pumps is the significant reduction in pulse pressure although the totality of experience with continuous-flow pumps has demonstrated that long-term survival with a non-physiological reduced pulse pressure is entirely feasible with maintenance of normal end-organ function. However, a number of concerns have arisen with continuous-flow pumps that have caused unanticipated problems in the biology of the coagulation system causing bleeding, unanticipated dysfunction of aortic valves resulting in aortic insufficiency, and difficulties in assessing blood pressure and management of hypertension (8).

Furthermore, continuous-flow rotary pumps exert significant shear forces on blood elements that have been associated with development of acquired von Willebrand’s syndrome (9). This is due to the effects of the shear forces on destroying high molecular weight multimers of the von Willebrand protein and stimulation of the enzyme ADAMT13. The development of acquired von Willebrand disease in addition to the requirement for anticoagulation with aspirin and warfarin has led to significant bleeding events with continuous-flow pumps, particularly events related to mucosal bleeding including gastrointestinal bleeding and epistaxis. Additionally, whether due to the alteration in biology or von Willebrand protein or other biology or reduction in systemic pulse pressure, patients supported with a continuous-flow pump have an increased risk of developing gastrointestinal arterio-venous malformations that further increase the risk of gastrointestinal bleeding (10).

Despite the significant ongoing success achieved with continuous-flow LVAD technology, significant hurdles to greater adoption of this therapy remain.  Stroke and right heart failure are significant adverse events that still plague the therapy. Further insight into the issues of stroke are needed as well as development of VAD systems capable of supporting both right and left ventricles.