Originally published in Volume 35 Issue 8 of Artificial Organs, 15 August 2011

As described in the Design News Engineering Achievement Award, I was considered an engineer for the long haul as I began working on artificial heart technology in 1966 while employed at Thermo Electron Corporation (Waltham, MA, USA). This company was one of six companies that received funding contracts in 1966 from the National Heart, Lung, and Blood Institute (NHLBI) to develop artificial heart technology. Our concept at that time was to develop a nuclear-powered artificial heart capable of supporting the entire circulation in man. For animal trials, we chose a 50-watt plutonium 238 fuel capsule as the energy source. The heat generated by the fuel capsule was used to boil water to make steam, which ran a miniature steam engine. The rotary power that was produced by the engine was then used to drive a hydraulic pump, which activated the blood pump. Unfortunately, this system was not very efficient and required a significant amount of waste heat to be disposed of. To accomplish this, waste heat was transferred through a heat exchanger into the blood stream. The animal’s response to the increased thermal load was a slight increase in the animal’s core temperature as well as a slight increase in the respiration rate. Thermal and radiation experiments were conducted in primates, which established that up to 0.7 watts per kilogram could be used without ill effects. Only minor lymphocyte abnormalities were observed following the animals through three generations. Work on the development of a nuclear-powered system was eventually terminated to concentrate on electrically powered systems that were considered simpler and without the hazards of nuclear radiation. Unfortunately, battery technology was not yet at the level of sophistication in the late 1960s and early 1970s that we required for artificial heart systems. The rechargeable systems lacked the power densities and reliabilities that we needed.

In the 1960s, very little was known about implantable power sources, biomaterials, toxicity, the blood/biomaterial interface, and the mating of mechanical systems to the biological systems of man and animals. We pioneered the development of textured surfaces for both rigid and flexible surfaces and investigated a large variety of power sources. We designed and developed over 15 different blood pump configurations using a variety of different mechanical and tissue valves as well as evaluated various biomaterials and conduit designs.

Before we could undertake clinical trials in man, we had to develop and test a wide variety of materials, bonding processes, protective coatings, methods of attachment to the biological system as well as evaluate the response of the biological system to the foreign materials that we were using.

As there were very few acceptable biomaterials at that time, we were forced to undertake material development to meet our specific needs and to demonstrate that these materials were not toxic to the biological system and remained stable for an extended duration of time. The body is a hostile environment of hot saline replete with enzymes that eliminate or dissolve foreign objects. Materials that we developed were Tecoflex, Cardioflex, and our flexible and rigid textured surfaces.

A critical need at that time was to find a flexible material that could be used for the fabrication of a diaphragm to be used in our pulsatile pump. A typical pump diaphragm is required to flex at least 40 million cycles per year without hydrolytic degradation or fracture. In addition, these materials are required to be thromboresistant as they are in direct contact with blood. The development of a biomaterial interface for flexible and rigid materials was a major undertaking as blood would recognize the foreign material and begin the process of covering up that material with a coagulum, which was prone to embolization, which could then lead to a stroke.

Many approaches were investigated to eliminate this thromboembolic complication that limited animal experiments. A variety of materials were evaluated; surface coatings as well as surface charges were studied. Two approaches that showed promise were the use of anticoagulant drugs coupled with careful hydraulic design.

An opposite approach followed by us consisted of fabricating a textured surface to encourage cells and protein to penetrate the porous structure and essentially anchor the coagulum that was being formed. Through the natural healing process, the formed and anchored coagulum would convert to a pseudo neointima. Once established, the patient’s blood would only see its own blood elements and not the underlying biomaterial. Clinical results utilizing these surfaces were very satisfying as they did not require warfarin as an anticoagulant; only aspirin was required. These surfaces reduced thromboembolic complications to acceptable levels.

Pneumatically driven axisymmetric blood pumps, total artificial hearts, and electrically driven devices were being tested in animals in the late 1960s and beyond. After completing several hundred animal experiments, which were primarily conducted in calves, we gained valuable information about the feasibility of pump designs and power sources that led us to initiate the first clinical trial in 1975. This study was undertaken before the Food and Drug Administration (FDA) began their involvement in this technology and only institutional review board approval was required. The first device that was used was an all-titanium pump with disc valves that I designed to have a 100-mL stroke volume. This device was the first introduction of titanium in this area of technology. This device was designated the Model 7 axisymmetric pump or sometimes called the abdominal left ventricular assist device; it was transposed between the left ventricle and the descending abdominal aorta just superior to the bifurcation. The device was implanted in the abdominal cavity and was intended to support patients who could not be weaned from bypass. In total, 20 devices were implanted at the Texas Heart Institute, Houston, TX, USA under the direction of Dr. John Norman and Dr. Bud Frazier.

The second clinical trial was also to investigate the use of a left ventricular assist device (LVAD) to support patients who could not be weaned from bypass after the surgical procedure was completed. It was anticipated that the natural heart would recover if it was allowed to rest. With recovery of the stunned myocardium, the LVAD could be removed. To accomplish this, I designed the Model 10 LVAD as a device that could be positioned externally for easy removal without re-entering the chest cavity. The LVAD was designed with long and flexible conduits that contained porcine xenograft valves. With the pump mounted on the external chest wall, the conduits could then enter the thoracic cavity and be attached to the left ventricle and the ascending aorta. With this implant position, the pump could easily be removed without going on bypass when cardiac recovery was confirmed.

Pump removal consisted of shutting the pump off, cross-clamping the conduits at the skin line on both the inlet and outlet conduit positions. The inner Dacron graft was then dissected and oversewn. Each stub was then positioned below the skin surface to allow final skin closure. The sternum was not opened nor was bypass needed. This clinical trial was conducted in Boston, MA, USA under the direction of Dr. William Bernhard of the Children’s Hospital Medical Center. The results of the first two trials for postcardiotomy applications were as follows: Of the 42 patients who were implanted, 25 died within 24 h, 13 died between 1 and 7 days, and 4 patients were successfully resuscitated with pump removal.

Pneumatic operation of a blood pump is satisfactory for short-term applications; however, it is not ideal for long-term use. The HeartMate I (HMI) (Thermo Electron Corp.) was conceived in 1975, which consisted of a blood pump with a stroke volume of 83 mL that could produce 10 L/min of blood flow. To satisfy both short- and long-term use, the pump was designed to be powered by either pneumatic actuation or by direct electric actuation. Transfer of electrical energy into the body was accomplished by either utilizing a transcutaneous energy transfer system, developed by Thermo Electron, or by direct percutaneous wire transfer. For direct pneumatic actuation, a percutaneous lead carries the pneumatic pulse from the external console to the blood pump.

With the advent of HMI, a third clinical trial was initiated. In this trial, the pump was intended to support patients who were awaiting cardiac transplant. The pneumatic version of the pump was used in this trial, as were porcine xenograft valves and textured surfaces. One hundred and sixteen patients were entered into the trial between 1985 and 1994. The HMI received FDA approval in 1994 for use as a bridge to transplant. Of importance in this trial was the determination that the HMI could be safely used in patients without the use of warfarin as an anticoagulant. This was accomplished by utilizing our textured surfaces, xenograft valves, appropriate hydraulic design, and appropriate biomaterials. This combination greatly reduced the need for anticoagulant drugs. To demonstrate this, we conservatively recommended at the early phases of the trial the following drug regimen: aspirin, dipyridamole, and warfarin. By the end of the trial it was clear that warfarin was not needed as a result of what we experienced when we were required to stop using it when patients developed bleeding disorders. Warfarin was eliminated without any adverse effects. The ability to eliminate warfarin was considered to be a major step forward in the treatment of these desperately ill patients. In addition, this trial also demonstrated that patients could be safely supported for extended durations in excess of 1 year. This third trial was conducted under the direction of Dr. Bud Frazier of the Texas Heart Institute. With this success, the fourth clinical trial was initiated in 1992 and utilized the HMI electrically actuated device. This consisted of the initial version designated as the HeartMate Vented Electric (HM VE), developed and manufactured at Thermo Electron Corp., followed by an improved version designated as the HM XVE, developed and manufactured by Thermedics, Inc. (Woburn, MA, USA), a newly formed subsidiary of Thermo Electron Corp.

Eighty-six patients received this device between 1992 and 1998 as a bridge to transplant. Of significance in this trial was the approval from the FDA to release patients being supported by the device to their own home while they waited for their transplant; this was another significant step forward. The release of patients from the hospital was made possible by the design of the HM VE system. The blood pump was identical to the HMI pneumatic device, which was previously approved. We did not need to prove that the pump was safe as this was already done; we only needed to prove that the electrical actuation did not alter the previous data that were collected in the HMI trial.

The drive mechanism that was used in this trial was direct electrical actuation. Two nested helical face cams were mounted on the back of the blood pump pusher plate or piston. Cam follower bearings that rode on the cams were attached to a rotating direct current torque motor that was mounted on the pump housing. With this arrangement, the motor rotation was converted to axial translation, which provided the force necessary to translate the pusher plate in a back-and-forth motion. The motion and force to expel the blood from the blood pump was produced by the electric motor while the force to retract the pusher plate to its start position was produced by the incoming blood flow into the empty pump chamber. The motor made one revolution and stopped for each pump stroke. The rate at which the pump operated was determined by a computer mounted in a small controller positioned on the outside of the patient. Power was provided to the motor and controller from two batteries carried by the patient. With this system, the patient was truly independent as he or she was not attached to any external console that would limit their activities. This trial was conducted under the direction of Dr. Bud Frazier of the Texas Heart Institute.

The fifth clinical trial was again with the HMI, electrically actuated. This trial was designated as the Randomized Evaluation of Mechanical Assistance for the Treatment of Congestive Heart failure (REMATCH) trial sponsored by Thermo Cardiosystems, Inc. (TCI, Woburn, MA, USA), a newly formed subsidiary of Thermedics, Inc., and ran between 1996 and 2003. One hundred and forty patients were randomized against medical therapy. This trial was successful and demonstrated that patients being supported on the HMI device did significantly better than those being treated with medical therapy. FDA approval was obtained in 2003 for use as a destination device. This trial was designed to be the gold standard. It was a trial jointly developed by the FDA, the NHLBI, Columbia University, and TCI. This was a randomized trial intended for nontransplant candidates for use as destination therapy. TCI was blinded to the control patient data until the end of the trial and only received the device patient data. The trial once again proved that patients supported on the HMI device did significantly better than those being treated with medical therapy. This trial was conducted under the direction of Dr. Eric Rose of Columbia Presbyterian Hospital in New York, NY, USA.

The sixth clinical trial evaluated the second-generation devices that were conceived in 1984 with the advent of the Hemopump. This device was a miniature, rotary, high-speed catheter-mounted pump that was developed by the Nimbus Company (Sacramento, CA, USA). This short-term temporary device was subsequently redesigned in 1990 to become a chronic high-speed axial flow device that utilized blood-immersed bearings. In 1996, I purchased the Nimbus Company and proceeded to convert their chronic high-speed axial pump to what is now known as the HeartMate II (HMII). After considerable animal testing and evaluation, the sixth clinical trial was undertaken with the HMII in 2003. Four hundred and eighty-six patients were entered into the trial using the device as a bridge to transplant and 460 additional patients were entered into the trial using the device as a destination device. In 2008, the device was approved by the FDA for use as a bridge to transplant and in 2010 it was approved for destination therapy. In summary, progress has been made with the HeartMate technology as can be seen in Table 1.

Table 1. HeartMate technology versus optimal medical management
Survival rate Optimal medical management (%) HM XVE (%) HMII (%)
1 year 25 52 68
2 years 8 23 58

TCI merged with Thoratec Laboratories in 2001 to form what is now the Thoratec Corporation.


Victor Poirier, BSME, MBA has over 45 years of experience developing artificial heart technology and biomaterials. He received national recognition as Engineer of the Year in 1992, and in 2003 he was inducted into the National Academy of Engineering. He was the president, CEO, and founder of Thermedics Inc. and Thermo Cardiosystems, Inc. He successfully obtained FDA approval and subsequently commercialized the HeartMate technology with two devices approved as a bridge to transplant and one device approved for permanent use. He was also instrumental in bringing second-generation technology (HMII) to commercial clinical utilization.