Tour de force

Engineers at Arrow International face this daunting challenge as they prepare left ventricular assist devices (LVADs)— conceived by teams at Penn State University and the Cleveland Clinic Foundation—for clinical trials, volume manufacturing, and approval by regulatory agencies around the world. These implanted pumps attach to the native heart, bringing improved circulation to patients suffering from congestive heart failure.

In their labs in this western suburb of Reading, the Arrow team performs a seemingly unending parade of tests involving individual electronic and mechanical components, as well as in vitro reliability tests on entire LVAD systems in custom-designed glass tanks. In addition, the engineers must oversee the work of outside vendors who put the devices through still more tests, ranging from environmental stresses to gas chromatography.

Project engineer Jeff Ravel performs highly-accelerated life testing on the new external battery pack module for the LionHeart LVAD. These HALT tests give early warnings of trouble spots in components.             James Thompson, Arrow's program manager for CorAide, demonstrates the simple, compact design of this LVAD, initially intended for short-term patient support.

Arrow, however, is no neophyte among the handful of teams now trying to build effective LVADs that will extend the life of patients suffering from congestive heart failure—a disease that kills 100,000 people each year worldwide, 40,000 in the US alone. A leading manufacturer of catheters and other critical care devices, Arrow as far back as 1993 was working on its own intra-aortic balloon pump for short-term patient support after heart attacks or surgery. At that time, the company learned that a Penn State team of engineers and clinicians based at the nearby Hershey medical center was looking for a commercial partner to take its LionHeart LVAD to market.

"Penn State was impressed with Arrow's ability to do the machining and other manufacturing and test tasks needed to advance the technology," recalls Jeffrey Lewis, one of the original team of Arrow engineers to work with Penn State and now the manager of quality and reliability for the company's Cardiac Assist Division. "And we were very excited to work with Gus Rosenberg, Dr. Pierce, and the others at Penn State who had pioneered this technology."

One of six teams to receive funding from the National Institutes of Health (NIH) to develop artificial hearts, Penn State had devised a novel "transcutaneous" approach to powering the implanted heart pump. Instead of using a cable that penetrates the skin to carry power from the external battery pack to the implant, LionHeart features an external coil to transmit power to an implanted induction coil directly under it (Figure 1). This approach eliminates the risk of infection found in percutaneous systems where a power cable penetrates the skin, and it gives patients a greater sense of freedom and mobility.

Figure 1. Implanted components of the LionHeart. An external coil placed directly over the internal induction coil transmits power transuctaneously to the implanted system.

In addition, the system's transmitter, power pack, and battery charger were very rudimentary. Design enhancements also were needed to improve the performance and corrosion resistance of the implanted motor. And a whole range of manufacturing processes—most notably the production of the pump's polymer blood sac—had to be devised. Arrow engineers had to tackle all these tasks—and more. "Arrow took on a monumental task in doing all the essential tests and in improving on the designs conceived here," says Gerson Rosenberg, the PhD engineer who heads the Division of Artificial Organs at Penn State's College of Medicine in Hershey.

One of the biggest early challenges for Arrow was creating the in vitro test protocol for the LionHeart. This was uncharted territory, recalls Lewis, who in the mid-'90s participated in the FDA-led team to establish the first rules and guidelines for conducting a continuous bench test for LVAD systems in a simulated human environment.

The objective of this reliability life test would be to demonstrate a minimum level of reliability and confidence in advance of gaining an FDA Investigational Device Exemption (IDE) for clinical trials involving patients with end stage heart failure. The tests would need to show a calculated 80% reliability, with at least a 60% confidence for a one-year operating life. In addition, the teams developing these devices would have to run at least eight systems for an entire year without a failure before starting any IDE clinical trials.

Setting the objective was one thing; actually developing the test bed was quite another. "There was no off-the-shelf test cell that we could buy," recalls Lewis. "So we had to design it from scratch, which involves fully understanding how your system works under various conditions, including worst-case scenarios."

The in vitro test cell required that the entire implanted system be submerged in a 37°C saline solution to simulate the harsh environment of the body. The engineers chose conventional glass fish tanks, which they modified to hold two LVADs. A plexiglass inlet pressure column in the tank propels saline to the inlet cannula of the LVAD at varying pressures to simulate the functioning of the device while a person is at rest, in normal activity, or exercising (Figure 2). Just as in a human implant, an external coil transmits power to the submerged LVAD through a Delrin plastic port in the glass—simulating the skin.
 

Figure 2. This in vitro test cell tank holds two LionHeart systems, immersed in a saline solution heated to body temperature. The tank links to a computer and a data-acquisition system that gives real-time readouts on temperature, pressure, flow, and power. Just as in a human implant, power is transmitted to the implanted system from an external transmitting coil.

It took the Arrow team nearly a year to devise the test cell tank, dealing with a raft of vendors, including IOtech (data-acquisition boards), DigiFlow (flow meters), Watlow (temperature controllers), DasyLab (data-acquisition software), Osmonics (filters), Wedeco (fluid sterilizers), Moyno (recirculation system), and Gems Sensors (flow switches). Each of the two LVADs in a tank has its own PC. The DasyLab software with a chart recorder yields real-time readings on such parameters as power, flow, pressure, LVAD pulse rate, and solution temperature. The system automatically generates printouts if the test cell operates outside of expected ranges.

Lewis, who relies on software tools from Reliasoft and Minitab to track LionHeart's complex test regimens, notes that the test cells have worked quite well. "We built 12 systems for the original LionHeart design and ran them until they wore out," he says. Now, the engineers are planning new in vitro cells to test the next generation of LionHeart components, such as an improved battery located inside the implanted motor controller.

Arrow will also rely on these cells for in vitro testing of another LVAD—the CorAide—which it is developing in conjunction with the Cleveland Clinic Foundation. Intended initially for shorter-term applications, such as assisting patients awaiting heart transplant, the device is a lighter, simpler design than the LionHeart. It features an inverted DC motor that rotates a suspended assembly to produce continuous blood flow, rather than pulsed flow as in the LionHeart.

Figure 3. In this test cell for the CorAide system, engineers incorporate a LionHeart LVAD, maintained at a low level of output to simulate a debilitated native left ventricle. This creates a pulsatile fluid stream as the inflow to the CorAide centirufgal pump. Acting in parallel with the LionHeart, the CorAide boosts the outflow, pumping the fluid throughout the mock circulatory loop. The other components in this system simulate systemic vascular compliance and resistance.

The LionHeart in vitro tests have demonstrated at least a 4-year durability for most of the components, but they have also pointed to shortcomings in the polymer blood sac housed within the pump assembly and the internal battery located in the implanted motor controller.

Made of seamless segmented polymer (ether polyurethaneurea), the Spandex-like blood sac must withstand constant, repetitive pressure and abrasion, as a roller screw pushes a metal plate up against it 100 times a minute, compressing the sac and pushing out the fluid (Figure 4). In the course of a year, that amounts to some 50 million flexes of the sac. As a result of this punishment, sac failures in the in vitro tests typically occurred in the 2–3-year range.

To improve performance, the Arrow engineers have adopted accelerated life testing, using mechanical test beds with rods that propel the pusher plates against fluid filled sacs at up to 2.5 times the normal rate. Using finite element analysis software, engineers compare the results of varying the geometry and thickness of the material, as well as different coatings on the pusher plate. "Based on these tests, we're looking at design changes that could take the expected life of the sac to a minimum of 3 years," notes Troy Werley, senior project engineer.

To ensure greater uniformity in the sacs, Werley has introduced a robot-based manufacturing system. He sends periodic samples of the raw materials used to make the sacs to Penn State for chemical analysis, using gas chromatography. Penn State also employs scanning electron microscopes to analyze abrasion and wear in blood sacs following accelerated tests.

Figure 4. Within LionHeart's titanium pump housing, a roller screw energy converter pushes a metal plate up against a blood-filled polyurethane sac, typically at the rate of 100 times a minute. With sac failure one of the chief concerns, engineers have devised accelerated tests that propel the metal plate against the sac at rates as high as 250 times a minute.

Arrow has devised other accelerated tests for flexing power cables in the LionHeart, using Keithley DMMs to analyze electrical properties in cables that have been repeatedly bent to angles up to 45° and pulled by weights up to 2 lbs. To evaluate the performance of implanted valves, engineers rely on a tester from Dynatek Dalta that activates the valves up to 10 times faster than normal. The blood pump also undergoes tests on a vibrating table.

Accelerated life-cycle testing, with equipment from Arbin Instruments, also simulates loads on power supplies, as well as the discharge cycles in the implanted battery and the external battery pack that patients wear. Here, too, the tests have contributed to design improvements. For the implanted battery, Arrow plans to shift to a new lithium ion chemistry that will double or triple the time that the system can operate after a charge. With the new battery, which will soon undergo in vitro tests, LionHeart recipients should be able to unplug all external systems for up to an hour every day for bathing, swimming, or other activities. Currently, patients are instructed to limit their time off external power to 20 minutes every other day.

Arrow's engineers tune the induction coils for LionHeart with Hewlett-Packard (Agilent) LCR meters, and they perfom hipot testing inhouse, using a Vitrek unit. They also perform some initial EMC screening with spectrum analyzers from Com-Power and Rohde & Schwarz, but the company relies on certified test houses for most EMC and ESD testing of the electronic components in this low-power device (4 to 8 W in the implanted system).

An outside test services company, using QualMark environmental chambers, performs highly accelerated life testing (HALT) on LionHeart's external and implanted electronic parts, exposing them to as much as a 100° variation in temperature and to vibrations exceeding 20 g. Such tests take place while the device is operating. For example, the LionHeart's new external power control module, which incorporates the transmitter coil, battery pack, and power supply in a 3-lb package, might be transmitting power to a functioning heart pump. Prior to sending components to the test house, Arrow engineers must add slots and other test fixtures to allow the temperature changes to reach boards and other electronic parts.

Says engineering manager Dan Frank, a 10-year veteran of the LionHeart program: "These HALT tests give us qualitative feedback so that we have the confidence that a component won't have an early failure in the in vitro tests. So, HALT can significantly speed the development process."

To ensure even faster test turnaround and exercise more control over the accelerated testing of electronic components, Arrow recently purchased two Test Equity temperature chambers and is investigating buying a QualMark chamber to do HALT in-house. Arrow engineers already perform required tests (EN 45502-1; Ref. 2) to ensure that the implant will work properly if a patient undergoes special medical procedures. For example, to simulate an ultrasound exam, the implantable parts of LionHeart are immersed in a water bath at room temperature and subjected for 1 hr to ultrasonic energy. Another test demonstrates the device's ability to withstand a defibrillation procedure (see "Can the LVAD survive defibrillation?").

In addition, Arrow engineers routinely do functional tests on incoming boards and other electronic components supplied by their contract manufacturers. They also perform a long list of inspection tasks on other supplied parts. For example, they use a Torquemeter dynamometer to test LionHeart's implanted pump motors as they come in from suppliers. To verify the dimensional accuracy of the LVAD's machined parts, they rely on a DEA Mistral coordinate measuring machine. A Deltronic DV 114 comparator inspects O-rings and other parts that are subject to distortion. The lab equipment for the LVAD projects ranges from simple gauges to ruggedized laptops to microscopes for inspection of connectors.

All told, Arrow's decade-long work on the LionHeart and CorAide devices represent an impressive tour de force of test, measurement, and inspection technologies. Yet, the real measure of the company's success will be the performance of these LVADs in the real world.

Since the first human implant in February 2001, 35 patients have received the LionHeart. The first patient lived for 2 years. Another was on the system for over 1000 days. One German patient was so comfortable with the system that he routinely removed the external power pack and took regular swims.

Arrow expects to receive Europe's CE Mark for the LionHeart this year, which will open the door to a substantial increase in implants. Clinical trials in Europe also began this year for CorAide, although results from the first implant indicated the need for further pump refinements.

Even so, Arrow, the largest company in the field, still faces uphill battles, as do the other LVAD teams. While clinical trials in the US continue for LionHeart under an IDE, Arrow officials concede that they need more clinical evidence to win FDA approval and also to convince cardiologists that LVADs are a better alternative for heart-failure patients than traditional treatment with medication.

In that regard, LVAD developers got a major boost in 2002, when results were released from a long-awaited study called REMATCH (Randomized Evaluation of Mechanical Assistance for the Treatment of Congestive Heart Failure). Conducted by Columbia University, NIH, and Thoratec, which makes a percutaneous LVAD, the study tracked 129 patients—half of whom received the Thoratec LVAD and half treated with medication. Among other findings, REMATCH showed that 52.1% of the LVAD patients had a survival rate of 1 year, versus a 24.7% survival rate for those on medication. What's more, LVAD patients generally were found to have greater mobility and better emotional health.

"REMATCH was a positive development for the entire LVAD field, because it showed that this technology can provide a higher quality of life for patients," says Carl Botterbusch, GM of Arrow's Cardiac Assist Division. "It also laid the groundwork for insurance reimbursement, which is very important in an expensive procedure like this." The total cost of the implant and surgery can reach $200,000.

Some cardiologists and heart surgeons also believe that the track record of LVAD implants will improve substantially if more younger patients begin receiving the devices. By their nature, clinical trials tend to focus on very sick older patients, many of whom are suffering from multi-organ failure.

One thing is sure: The need is great. In the US alone, some four million people suffer from chronic heart failure, with about 400,000 new cases diagnosed annually. Heart transplants fill only a small portion of the need, since just 2500 donor hearts become available each year.

"If Arrow could meet the needs of just a small percentage of people who are dying from heart failure, it would represent a very significant market for us," notes company president Philip Fleck. "And with two very different technologies—the LionHeart and CorAide—we are well positioned to be the leader in the field."

Fleck adds, however, that he views both LVAD projects as long-term R&D efforts, as his engineers strive to introduce still more improvements that will make the systems more durable and user friendly for patients. And that, of course, means there's no end in sight for Arrow's testing mission.

Mehta, S., et al. "The LionHeart LVD-2000: A Completely Implanted Left Ventricular Assist Device for Chronic Circulatory Support," Annals of Thoracic Surgery, March 2001 Supplement. ats.ctsnetjournals.org.

Ochiai, Y., et al. "In Vivo Hemodynamic Performance of the Cleveland Clinic CorAide Blood Pump in Calves," Annals of Thoracic Surgery , September 2001. ats.ctsnetjournals.org.