What's the heart, after all? Just two pumps in series (right and left ventricles) with a couple of filling chambers (atria). Yet can you think of any project more challenging than designing an implantable artificial heart or heart-assist device? For decades, engineers pursued that goal with limited success. But in recent years, several projects have made the breakthrough to clinical test, and entry to the commercial market. Each engineering team involved followed a different path. And each group continues to develop and improve its equipment.

Help or replace. In 1992, Victor Poirier, president of Thermo Cardiosystems Inc., Woburn, MA, was selected by readers of Design News as the Engineer of the Year. He won for his work on the company's Left Ventricle Assist Device (LVAD). That product was approved by the FDA in 1994, and has been implanted in hundreds of patients. It represents both a great step forward and a recognition of limits by the Thermo Cardiosystems design team.

In 1966, when Poirier and his colleagues began working on heart pumps, they intended to develop a total implantable artificial heart. As the project progressed, the team discovered a host of problems, and then reconsidered their basic objective. "We looked at the need," says Poirier, "and about 80% of the patients only needed one-sided support." The team decided to develop a left-sided pump operated with air, leaving the power source outside the body. Such a left-sided pump is called an LVAD.

At the time, engineering teams around the U.S. were working with both constant-flow and pulsatile heart pumps. Poirier and his associates decided to build a pulsatile pump to mimic the behavior of the natural heart. The pneumatically powered unit they developed (described in Design News, 2/10/92) pumps blood using a blood-contacting diaphragm bonded to a piston. By using a cam-operated piston to drive the diaphragm, the Thermo Cardiosystems engineers produced a very controlled motion. A subsequent version of the pump is electrically driven, with an external power source. Feed-throughs pass through the skin and permit air and electrical power to reach the pumps.

Poirier and his colleagues solved two basic problems that all designers of heart pumps must deal with. The first involves avoidance of eddys or standing vortices within the pump. If eddies or a standing vortex develop, blood will clot, and that's dangerous for the patient. To prevent eddies, the designers designed their pump to wash out completely on each stroke. They dealt with the vortex problem by using a cam that causes the piston to wobble, and makes the vortex wander within the pump. If it doesn't stand still, no clots can form.

Secondly, Thermo Cardiosystems had to deal with the risk of clotting on blood-contacting surfaces within the pump. They did so by developing a textured surface for all such surfaces. The human body quickly forms a fibrous sheath on these textured surfaces. After that, the body never sees the underlying biomaterials, which eliminates clotting problems. This proprietary texturing technology remains a fundamental advantage for Thermo Cardiosystems. The alternative to texturing surfaces is to develop a surface that is smooth to the size of blood cells--sub-micron size. That approach also works, but presents a new set of manufacturing problems.

What will the future bring at Thermo Cardiosystems? Well, the company has developed an external battery-driven power system. It uses direct coupling to drive the implanted electrically powered version of the pulsatile pump. This system, with a pulsatile pump and an implanted power source, is in clinical trials. "We hope to get FDA approval for this version this year," says Poirier.

Further, Poirier and his colleagues are now developing a constant-flow pump to replace the piston-driven pulsatile units. The pulsatile pumps work well, but they're big. Using a high-speed, axial-flow rotary pump produces a smaller package. "Because they're small, you can put in two," explains Poirier. Using two creates a Bilateral Ventricular Assist Device (BIVAD), which effectively does the work of both of the heart's pumping chambers.

In addition to offering a size advantage, the constant-flow pump can produce a pulse. "The amount of flow you can produce across a pump depends upon the pressure on each side," Poirier points out. "If inlet pressure is low and outlet pressure high, flow will drop. With this (the axial flow pump) attached to a natural heart, the inlet sees pressure varying from 120 to 5 mm of mercury. The outlet pressure is the arterial pressure, which varies between 120 and 80 mm of mercury." This difference causes flow changes; a pulse. Thermo Cardiosystems is also looking at a centrifugal pump. Poirier expects the axial rotary to be first to market; a magnetically suspended centrifugal pump is in R&D. At Thermo Cardiosystems, the pulsatile pumps represent generation one, an axial flow vaned pump with mechanical bearings generation two, and then another continuous-flow pump with magnetic bearings as generation three.

The approach taken at Thermo Cardiosystems is to assist the heart, not replace it. Pumps developed by Poirier and his associates are now used as a bridge device until a heart transplant can take place. If they can be shown to function in the body for, say, two to five years, they can become a permanently implanted assist device.

Another approach to the VAD. Engineers at Thoratec Laboratories Inc., Berkeley, CA, chose another approach to the design of a ventricular assist device. "It's a sack-type pump," explains David J. Farrar, vice president of R&D at Thoratec, "a pulsatile, air-driven pump with a blood-pumping sack." In this design, air compresses a diaphragm exterior to the sack. The diaphragm acts upon the blood sack and causes ejection of blood. Thoratec's device, called the Thoratec VAD(R) System, uses pneumatic power that's generated external to the patient's body.

Initial pump design took place at Pennsylvania State University. Thoratec engineers liked the pump because it's fairly small when compared to some other pulsatile implantables, and produces good washing of surfaces as well as good cardiac flow outputs. The Thoratec design represents a good example of a VAD that relies upon internal smoothness to prevent clot formation.

Because of supporting conduits or cannula designed by Thoratec, this VAD can function on either side of the heart. Unlike some other devices, it can function as either an LVAD or an RVAD.

Farrar believes that the biomaterial used to make the blood-contacting surfaces of the VAD represents a significant advantage for Thoratec. "It's a material called Thoralon^TM. It's a blend of a highly refined base polymer--a polyether urethane urea--with surface-modifying additives. The surface-modifying additive is a silicon-containing high-molecular-weight polymer blended with the base polymer." This material produces a very smooth surface that does not provide locations where clots can form.

Today, Thoratec is working to reduce the size of the external driver used with its VAD. Presently the driver is on a cart. The new unit, called the Thoratec TLC-II Portable Driver, will allow a patient to move about readily. In another research area, Farrar and his colleagues are seeking ways to convert power in muscles to drive a circulatory support device. "This would eliminate all of the electricity and batteries needed in electromechanical systems," says Farrar.

Other teams have tried to accomplish this goal by wrapping the latissimus dorsi around the damaged heart. "We have invented a skeletal muscle energy converter," says Farrar. "Instead of wrapping the muscle around the heart like some people do, we have a way of attaching the muscle to our energy converter. Then the energy converter changes the mechanical power in muscle to hydraulic power that can drive the VAD."

Like other VADs, the Thoratec system can function as a bridge to heart transplant. But Farrar sees them as one day becoming the primary support for persons with serious heart problems. "Bridging to transplant is the proving ground for showing that this technology works," he says. "The ultimate goal would be to use these as an alternative to transplant and in lieu of transplant."

Going for the whole magilla. Several generations of engineers have tried to build complete artificial hearts. "The artificial heart is the Holy Grail of cardiac medicine," says David M. Lederman, president and CEO of ABIOMED Inc., Danvers, MA. "It's the next frontier." Again and again engineers have believed that the artificial heart was near at hand. Again and again, projects have failed. "This time it's for real." says Lederman.

By the year 2000, Lederman believes ABIOMED will be in clinical trials with its full artificial heart. It consists of three hermetically sealed units-two blood-pumping chambers, separated by an active artificial septum (partition). The centrally located septum and two left and right artificial ventricles pump blood.

A small electric motor drives a centrifugal pump at 5,000 to 9,000 rpm. The pump continuously moves a special fluid in one direction, and a fluid switching rotary valve converts that unidirectional flow into a reciprocating action. The flow alternately fills and empties each half of the artificial septum. Two lateral flexible membranes form the left and right walls of the septum, and they transfer pumping action to the artificial ventricles, which contain blood. As the septal walls move, the ventricles alternately pump blood.

Each ventricle has two tri-leaflet polymeric heart valves that mimic the action of native heart valves. The blood-pumping ventricles and associated valves are manufactured in one piece from ABIOMED's proprietary Angioplex^TM elastomer. The seamless design of the ventricles arises from the conviction of ABIOMED's engineers that the smooth surface approach to avoiding clots works only if the entire path seen by blood is smooth. "If you interpose mechanical valves, steps or junctions, then you have a problem," says Lederman. "We don't think it's possible to make a smooth-surface device that will not require any kind of anticoagulation unless you make the whole thing in one piece. And we have."

Rechargeable batteries worn externally provide power to the heart via what Lederman calls, "now the most sophisticated skin transformer available." In addition to powering the heart, the transformer provides a means of collecting data on its performance. That information can be sent to a cardiac center by telemetry or modem.

Pressure sensors in the heart provide it with the information needed to determine heart rate. The pressure sensors feed a microprocessor built into the heart. It then controls the rate at which the heart operates. "When a patient exercises," says Lederman, "it will beat faster like a natural heart."

There is no single-point failure in the heart's software, according to Lederman, and the fault safety mechanism is hardware driven. A hardware backup always exists for the electronics. In the event of some sort of massive control failure, the heart would continue to run at a constant rate and trigger alarms.

An implanted battery allows about two hours of operation without the external battery pack. That time interval can allow a patient to shower, or take a walk free of the batteries. About a half pound of batteries must be carried for each hour of operation. ABIOMED supplies a backpack, shoulder harness, and purse that can hold the batteries.

As an interesting aside, the ventricles used on the ABIOMED artificial heart are transparent. "During the initial implantation, we want to make absolutely sure there are no air bubbles trapped anywhere. We want to be confident that the perfusionist can make sure there's absolutely nothing inside," Lederman explains.

A rising tide. Each year, physicians identify from 30,000 to 50,000 people in the United States as candidates for heart transplants. About 2,000 of those patients actually receive hearts. The devices described in this article offer hope to thousands of persons who otherwise face certain death.

A rising tide lifts all the boats. As many years of effort by dedicated scientists engineers come to fruition, as more and more sophisticated equipment becomes available to treat heart disease, millions of people all over the world will benefit.

First of the Mohicans

In 1993, after five years of trials, the first "heart" was approved for commercial sale by the FDA. The BVS-5000 from ABIOMED Inc. was intended to be relatively inexpensive and easy to operate. A cart-sized unit, it can actuate either one or two pumps. ABIOMED's BVS-5000 system uses pneumatics to pump blood. Blood enters a flexible sleeve, then air collapses the sleeve and pumps the blood.

The technology behind the BVS-5000 is transparent to the caregiver. Operating instructions are printed on the machine's control panel. Once calibrated, the BVS-5000 uses internal diagnostics to monitor its performance. If a fault occurs, for example a blood line becomes plugged or kinked, the machine will trigger an alarm. The operator can then read an explanation of the cause of the alarm condition from the control panel and take corrective action.

To build their company's new implantable artificial heart, design engineers at ABIOMED had to fold this machine's functions into an object a surgeon could hold in his or her hands.

Building a heart

Whether a design engineer works on a ventricular assist device or a full artificial heart, he or she must overcome certain problems. Here are a few of the most noteworthy: Clot formation: blood clots travel and can kill the patient. Clotting must be prevented. Blood damage: pumps must be designed to avoid destroying fragile blood cells. Reliability: as life-support equipment, these devices can't fail. Durability: the body is a hostile, very corrosive environment, so materials choices become critical. Materials: suppliers often fear legal liability, thus many companies have decided to manufacture proprietary plastics for use in their products. Obtaining FDA approval: FDA requirements are--quite understandably--rigorous; companies must submit thousands of pages of documentation and take part in lengthy animal tests and clinical trials before approval will be granted.

Other applications