Kantrowitz and his research team worked on a wide variety of assistive devices during the 1960s, but his greatest innovation was the left ventricular assist device (LVAD) for providing diastolic augmentation. Working once again with his brother Arthur, he designed a U-shaped device consisting of an expandable chamber housed within a silicone rubber container. An external pump pulsed air to the chamber through a tube inserted through the chest wall. The device was attached to the aorta at two points and the aortic arch was tied off between them, so that all the blood pumped from the left ventricle was shunted through the device. Timed by a lead to the patient's EKG monitor, the air chamber inflated during the relaxed phase of the cardiac cycle (diastole), providing an extra push to the blood flowing down to the coronary arteries at the base of the aorta, and out into general circulation through the descending aorta. This enhanced the oxygenation of the heart muscle, and by boosting the general circulation, took some of the workload off the heart. After extensive laboratory experience and successful implantation in several dogs, Kantrowitz was ready to use the device in a human patient.

In February 1966, he implanted the LVAD in a 33-year-old man with cirrhosis of the liver and chronic cardiomyopathy, in severe congestive heart failure. The operation went smoothly, and the patient's circulation appeared to respond, but he bled extensively afterward and died after 20 hours. Kantrowitz did a second implantation in May 1966, with somewhat better results. Prior to the operation, the 63-year-old diabetic woman had had two heart attacks, and had been bedridden for several years. Afterward, she seemed to be recovering well for over a week, then died from a series of small strokes. Postmortem examination showed that there were blood clots in one arm of the LVAD. The attempt to implant a permanent partial left heart device was widely covered in the press, as was Michael DeBakey's trial with a different kind of LVAD. Kantrowitz decided to modify the LVAD design, and would soon develop a device--the intraaortic balloon pump--to be placed inside the aorta itself for temporary cardiac assistance. His second generation permanent LVAD would be based on this design. (See exhibit section "Further Innovations in Diastolic Augmentation.")

Kantrowitz was committed to finding mechanical means to help those with heart failure, but, like many other surgical innovators, he recognized that heart transplants would also provide a chance for patients with irreparable defects or damage. This was especially true for children born with multiple defects because no mechanical device with sufficient power would fit into the small space of a child's chest, nor could such a device grow with the child. By the early 1960s, reliable heart-lung bypass machines were making open-heart surgery easier, and surgeons had had some success in transplanting kidneys, and were attempting to transplant other organs. There were several key challenges in transplant surgery, beyond achieving the best technique for attaching the transplanted organ. Immunological rejection of the foreign tissue by the recipient was always a danger, even when efforts were made to match tissue types closely; immunosuppressant drugs such as cortisone often helped, but left the patient open to infection. Hearts posed another problem: unlike kidneys, which can be taken directly from living donors (because they have a second one to carry on with), human hearts would only be available when their owners had died. In the United States death was legally defined as the cessation of cardiac activity, not just brain death, so surgeons needed to preserve the donor heart's viability from the time it stopped beating on its own to its insertion into the recipient. Two Stanford surgeons, Norman Shumway and Richard Lower, had started transplanting dog hearts while experimenting with hypothermia during the late 1950s. They had worked out many of the surgical and preservation techniques, and Kantrowitz built on their work as he began his own transplant research in 1962. He worked on puppies rather than adult dogs, reasoning that rejection might be less intense with their less mature immune systems.

Over the next four years, Kantrowitz adapted Shumway and Lower's surgical transplant techniques to the smaller puppies, working with hypothermia rather than a heart-lung bypass machine. By 1966, he and his team had done over 400 canine transplants, with a good number of the puppies surviving and growing normally for many months (several transplant dogs would survive for over four years). They felt ready to attempt a transplant in an infant with irreparable heart defects, using a donor heart obtained from an anencephalic baby. Such infants, born without the vital cerebral cortex structures above the midbrain, rarely lived more than 48 hours, yet in most cases, their hearts were normal. In May 1966, an infant with an incurable complex of congenital heart defects was born at Maimonides. As he was unlikely to live more than a few weeks without a transplant, his family agreed to let the surgeons attempt this experimental procedure. The chief of the hospital's obstetric service initiated a national search for an anencephalic donor, writing to 500 hospitals across the country. A donor was found in Oregon, and after tissue compatibility was established, the donor baby was flown to New York, and both donor and recipient were prepared for surgery. Kantrowitz's team had to wait until the donor's heart stopped beating to extract it, and by that time the heart was found to be too weak to support the recipient, so Kantrowitz did not attempt to complete the transplant. The intended recipient died of congestive heart failure several weeks later.

Not until November of 1967 did another suitable recipient appear. Again, the newborn had an uncorrectable combination of congenital defects. Several days before a compatible donor was found, South African surgeon Christiaan Barnard startled the world by completing the first human heart transplant in Cape Town on December 3. Kantrowitz, along with Shumway, Lower, and several others who hoped to accomplish the first such procedure, were stunned and disappointed. Nevertheless, when an anencephalic donor was located for his small patient, Kantrowitz went ahead, performing the world's second, and America's first, heart transplant on December 6. Although forced once again to wait until the donor's heart stopped, Kantrowitz's operation went smoothly, and the donor's heart took over the recipient's circulation. The patient awoke and moved, but the new heart failed less than seven hours later.

Kantrowitz had hoped to keep his work unpublicized until he could report his results to the medical community. However, the media frenzy following Barnard's success (his adult patient survived for 13 days before succumbing to infection) made it impossible to keep the infant transplant secret. Media interest in this exciting work was intense and reporters were camped on the hospital steps even before the operation was done. Kantrowitz was forced to confront the press after his patient's death and report that the operation had been a failure. In January 1968, he did another transplant, this time taking the heart from a young woman brain-dead from a cerebral hemorrhage to save a 58-year-old man with congestive heart failure. The donor's heart seemed healthy but proved too small to support the much larger recipient, who died 10½ hours after surgery.

The first few heart transplants were followed by another 100 worldwide during 1968. Many of these were done by surgeons who, though experienced open-heart operators, had virtually no experience with transplants (or the animal research that informed the work of Kantrowitz, Shumway, and others). The procedure itself was not unduly complicated, but the post-operative management of transplant patients was quite challenging; many recipients were severely debilitated by their heart disease, and rejection responses had to be monitored and controlled, while still protecting the patient from infection. Only the teams with considerable laboratory experience did this well, and the overall survival record in the first enthusiastic wave of transplants was dismal--two thirds died within three months. Kantrowitz, despite his years of preparation, finally decided that the immunological challenges involved were too great, as did many other surgeons. Within several years, only Shumway's team at Stanford was attempting transplants. It was only when the immunosuppressant cyclosporin became available in the late 1970s that long-term survival became possible.

Although the first attempted transplants did not lead directly to routine clinical success, they did stimulate much discussion about how to define death and how to evaluate candidates for transplants. Many surgeons, including Kantrowitz, traced at least some failed transplants to their inability to remove the hearts from brain-dead donors before those hearts shut down. Because South Africa used different criteria to determine death, Christiaan Barnard was able to use living donor hearts, and many of his transplants survived for many months. In the United States, the definition of death was tested in 1972 when Lower, then at the University of Virginia, was sued for wrongful death for taking the heart of a brain-dead donor for a 1968 transplant. The judge and jury cleared him, concluding that clinical brain death was as valid a definition as arrested heart function. By 1980, many states and the AMA had officially revised their definitions, vastly improving the odds for transplant recipients in years to come. As one of the small group of heart transplant pioneers, Kantrowitz continued to participate in these discussions and to help formulate common criteria and protocols for transplants, but decided to return to working on ventricular assist devices.