The Heart Healers (23 page)

Dr. Albert Starr was only sixty-three years old when he abandoned the comfortable confines of East Coast academia to create the University of Oregon Hospital’s brand-new program in cardiac surgery.

Starr’s mandate was to achieve results comparable to those of Lillehei and Kirklin in Minnesota, and to do it as soon as possible. Soon after his arrival a fragile sixty-year-old retired engineer with early Parkinson’s disease dressed in crumpled slacks and a sports shirt without a tie made an appointment to see him. Lowell Edwards did not walk into Starr’s office with a heart murmur, he came with a proposal.

Edwards was a hydraulic engineer who held sixty-three patents on a spectrum of inventions from a hydraulic lumber-debarking system to a fuel injection system for World War II aircraft. The patents provided royalty income to support his Edwards Laboratories in Portland. Edwards wanted to propose that he use his knowledge of fluid dynamics to create an artificial heart. His motivation was personal. In his teens, Edwards had two harrowing episodes of rheumatic fever, the cause of most of the heart valve disease in that era. As he grew up, for years he feared that he would develop valve disease and die in early adulthood. Starr admired Edwards’s vision of creating a pumping device to help people with heart valve disease. But he had a different vision. Cardiology did not even have a satisfactory artificial valve, he told Edwards, let alone a whole heart. The first step in helping the failing heart was not an artificial heart, but an artificial valve. Asking the heart muscle to endlessly pump against the tremendous resistance offered by a severely narrowed valve was like trying to pump the Hudson River through a storm pipe. Conversely, when a valve failed to prevent backflow, the heart muscle was like a sailor trying to bail out a boat with a huge hole in its hull. Both tasks were ultimately impossible, and the heart failed. Starr proposed a direct solution for Edwards’s greatest fear. Instead of an artificial heart, they would create an artificial valve. Like Harken before him, Starr chose the mitral valve as his first target.

Edwards and Starr confronted three critical initial questions, each with the potential to consume years of research. Was it possible to design a valve simpler than Nature’s impossibly complex mitral valve with its two leaflets and attached cords? Could they find a material that could be flung back and forth sixty times a minute, twenty-four hours a day without respite for an entire lifetime? How could they possibly attach their device within a vigorously contracting and relaxing heart?

Edwards and Starr made their first critical decision. They would not use Nature’s sheets and cords design. In its place Edwards’s first valve consisted of a circular ring into which he inserted two thick semicircular “leaflets” hinged on a metal crossbar. For his materials, Edwards chose Teflon for the ring and Silastic (a portmanteau of silicon and plastic) for his leaflets. Starr inserted the valve in place of the normal canine mitral valve apparatus. His dogs survived the operation with good cardiac function. Their collaboration had proven that a rigid valve could function in place of Nature’s more flexible living structure.

But their design was still an abject failure. All the dogs died within two to three days from congested lungs. Puzzled, Starr performed autopsies. He found two devastating complications. First, the circular ring housing the valve partially detached from the heart, and second, the exposed metal hinges were covered with fragile blood clots waiting to be dislodged to the brain and other organs. Edwards went back to the drawing board. He returned with a thicker, more compliant sewing ring. He had solved the problem of valve detachment. But they made no progress on eliminating clotting on the central crossbar and the animals continued to die. Two bars sitting in the middle of a flowing bloodstream was too tempting a target for platelets to build fatal clots.

Starr and Edwards had to admit that their leaflet design was a failure. Their new design, although not original, certainly reflected outside-the-box thinking. They created a free-floating spherical poppet inside a cage. When the heart contracted the ball leaped forward, to be limited in its forward excursion by the arms of the cage. When the heart relaxed the ball fell back into its circular seat, preventing backflow. They hoped that the constantly moving, spinning ball would not induce clot formation. Their new design had the short-term effect of lengthening animal survival to about a month, but the clotting problem reemerged. Over time blood clots formed on the cloth sewing ring that attached the valve to the heart. The clots ultimately became huge, blocking blood flow.

Then in the spring of 1959, while bounding up the stairs to his research laboratory, Starr glanced outside at the cherry blossoms emerging from their sheaths. Intuition flashed: “All of a sudden I had a Eureka moment … my mind wandered, and suddenly I thought of the solution to the thrombosis problem. Why not create a Silastic shield that could be retracted during implantation of the valve, and then snapped into place covering the entire zone of tissue injury?” Edwards created a retractable shield. Starr sewed the valve in place, and then pulled the shield over the sewing ring. The “cherry blossom” refinement resulted in 80% long-term animal survival.

Starr soon had a kennel full of happy dogs, which he planned to follow for the next few years looking for late complications. Early in the summer of 1960 Starr’s vision of logical and progressive scientific development evaporated in an instant when Dr. Herbert Griswold, Oregon’s chief of cardiology, stopped by the lab. What Griswold saw was a kennel full of healthy dogs with prosthetic mitral valves clicking away. One of the dogs licked his hand. Griswold had many patients in the hospital in the terminal stages of heart failure with mitral valve disease. He insisted that Starr treat his patients. As Starr recalls, “we were suddenly thrust into the real world of informed consent, liability, and the need to separate manufacturing from scientific assessment, with the first potential patients already in the hospital.”

*   *   *

NORMA FORBES HAD
rheumatic fever as a child, and it damaged her heart valves. When she became incapacitated by mitral stenosis years later, a surgeon had treated it by opening the valve using the closed heart technique pioneered by Harken and Bailey. The surgeon opened the stenosed valve, but left her with a new complication. Now the mitral valve did not prevent backflow when the ventricle contracted. She now had the opposite of mitral stenosis: mitral insufficiency. Her lungs were flooded with back flow. At just thirty-three years of age, she lay in a University of Oregon Medical Center oxygen tent continuously gasping for air, dying of heart failure in the prime of her life.

Norma Forbes posed a wrenching ethical issue for Starr. Should he use the unshielded ball valve that achieved long-term survival in only one of fifty dogs, or the shielded valve that provided predictable long-term survival in 80% of the dogs? The answer seems obvious, but now uncertainty made its usual appearance. The much simpler unshielded valve was also much easier to implant. In addition, the human coagulation system is far more forgiving than the dog’s. Perhaps the simpler device would be sufficient. But since Starr had never used either valve in a human being, which should he choose?

“I grappled with it myself and decided on the simple device for the first implant. The thinking was this: If we started with the simple device and it worked, we would not have to use the more complicated valve. If it did not work, we could then go to the more complicated valve. However, if we started with the more complicated valve and it worked, how would we know whether the simple valve might also have worked?” he thought. Starr had set his own stage for the Pain of the Pioneer.

With Norma on the heart-lung machine, Starr opened her heart, and cut out all of her scarred mitral valve. In its place he sutured in the unsheathed valve that he and Edwards had created. He felt confident, having performed the same surgery with the same valve in many dogs. As he closed the chest, he mused how Norma’s surgery proved to be quite a bit easier than his operations in the animal laboratory because Norma’s heart was larger.

Postoperative care was a different issue. “We had no intensive care unit and had to create one for this project. We literally mobilized the whole medical school faculty for this project,” Starr wrote. Norma awakened from anesthesia in the late afternoon. By evening was she able to sit up in bed for a portable X-ray of her chest. Starr was elated. On his first case, he had succeeded far beyond what he could have imagined when he and Edwards first met two years earlier.

Later that night, tired from lying on her back all day, Norma asked for help in rolling onto her right side. Hovering beside his patient’s bed, Starr immediately reached out to help.

“While I was helping her, she gasped and died suddenly in my arms,” he wrote. Again imagine yourself as the young surgeon at that moment. For Starr, the devastating experience was almost beyond endurance. This could not be. How could this possibly have happened?

The cause of Norma’s death, Starr would learn at autopsy, was massive air embolism. Her change in body position had released a large air bubble trapped in her left ventricle. As he closed the heart, Starr had allowed air to remain within the ventricle. He had actually seen the air quite clearly on Norma’s post-op X-ray, but assumed that he was looking at air trapped between the heart and the pericardium, a finding that carried no risk, since the air would be absorbed without symptoms over a couple of days. Instead, Starr had joined Bailey and Lillehei as a surgical pioneer done in by the same, simple, avoidable technical error. Just ten hours after the miraculously return of life’s simplest function, a breath of fresh air, youthful Norma Forbes’s vitality had been snatched away by the simplest of human errors. How stupendously unfair. But tragedies precede successes. He pledged to himself, “I will never let that happen again.” He never did.

Starr’s second patient was a fifty-two-year-old truck dispatcher who had two prior surgical procedures on the mitral valve using the Harken and Bailey’s finger-fracture method. His second patient became his first survivor. Starr and Edwards had gone from design to implementation in just two years. But he had hardly solved his problems. Starr would have to endure the wrenching learning curve that I have seen plague all the great new cardiovascular interventions. His operative mortality in the first year was a devastating 50%. Five years later it was less than 10%. Starr had learned from his repeated failures. Starr’s experience represents an eternal, inescapable truth I know too well. In medical innovation, if we are unwilling to fail, we rarely succeed.

After Starr presented his results at the American Surgical Association in the spring of 1961, surgical programs around the country began demanding heart valves. The new field of prosthetic valve surgery exploded. Lowell Edwards, who had made Starr’s early valves by hand, formed a company in Santa Ana, California, to meet the demand, and to continue valve development and testing. At the same time, Starr and Edwards modified their mitral design to make it suitable for aortic valve replacement. The first aortic valve was implanted just a year later. In California, Edwards Laboratories conducted accelerated fatigue tests that demonstrated the durability of the ball to beyond forty years without damage. From one of my many visits to Edwards Laboratories a few years later, I still recall the bizarre experience of seeing and hearing a chamber full of Silastic balls, like caged crazed drummers, each whacking away their rhythm at sixty times a minute into eternity.

*   *   *

TODAY BOTH THE
diagnosis and treatment of valve surgery bears little resemblance to the early Starr-Edwards years. In those halcyon years, I focused on the intricacies of physical diagnosis, hoping to discern the correct diagnosis by palpation (touch) and listening for murmurs with my stethoscope; today I am a cardiovascular dinosaur: my trainees get an irrefutable answer from the echocardiogram.

We also have abandoned the bulky ball-in-cage structure for valves with synthetic leaflets strikingly similar to Starr and Edwards’s first failed designs. But in addition we fashion leaflets from pig, cow, or even human tissue. And in some patients with aortic valve disease, we use a cadaver aorta with intact valves. Best of all, in selected cases surgeons can “resculpt” the diseased valve without even replacing it. When we come later to the future of valve surgery I will show you the most amazing feat of all: restoration of valve function without surgery.

Among the many valve types how do we choose which one to use, what are the pros and cons? Valve resculpting is appealing because we preserve the strength of the heart muscle and do not need to use blood thinners after surgery. Pig or cow tissue valves also eliminate the need for long-term blood thinner treatment, but they deteriorate over ten to fifteen years and must be replaced. Synthetic valves have greater durability but require blood-thinning therapy. So when a valve cannot be repaired, in a younger patient we favor a synthetic valve to avoid repeat surgery, whereas in an older patient we may recommend a tissue valve.

We recommend surgery based predominantly on a patient’s symptoms and evidence of progressive deterioration of cardiac function. Fainting, angina, or shortness of breath is an indication for surgery for patients with aortic stenosis, since with medical therapy, survival after the appearance of these symptoms is only two to four years. I seldom see mitral stenosis since the virtual eradication of rheumatic fever in the West, but would choose surgery with these symptoms or with coughing blood. For insufficient or incompetent aortic and mitral valves, we use the same criteria, always juggling the uncertainties: the risk of the disease, the risk of surgery, and the benefit of therapy.

In the earliest days, assistants gave the head cardiac surgeon a clear view of the heart by continuously vigorously spreading the ribs with a metal device shaped like a curved hand (now we have mechanical devices). A half century after medical school, I still wince at the memory of “rib retraction,” which required neither brains nor skill, and was assigned to the lowest person on the totem pole. After surgery we hospitalized our surgery patients for two to three weeks, treating them as if they were the human equivalent of expensive, fragile Dresden china. Today, our nurses remorselessly insist on postoperative day one that patients cough, blow into a resistant tube to reinflate their lungs, flex their legs to prevent blood clots, and start walking to restore vitality. If all goes well, we aim for hospital discharge on postoperative day three to five.