The History of Liver Transplantation

Introduction: The Genesis of Liver Transplantation

The transplantation of all of the other major organs can be traced back to the early 1900s, but for liver transplantation the first reported liver transplant was in 1952 at the fifty-fourth Congress of the Italian Society of Surgery. In 1952 Vittorio Staudacher from the University of Milan ( Fig. 1-1 ) published a series of experiments in which the first description of the technique of liver transplantation in four dogs was outlined. This first liver transplant was an orthotopic liver transplant, where the host liver was removed and fully replaced by the donor allograft, and in his report Staudacher clearly describes the procedure in five steps that resemble the modern transplant operation. In the discussion Staudacher commented that no one had reported a liver transplant previously. Although Staudacher’s achievements were known by some colleagues in Italy, his work went essentially unnoticed for almost 6 decades.

In 1955 C. Stuart Welch of Albany Medical College reported the first heterotopic liver transplant in a one-page article published in Transplantation Bulletin , the forerunner of the present day journal Transplantation . For more than 50 years, Welch’s report was considered the first reported liver transplant until the recent discovery of Staudacher’s published work. In Welch’s “auxiliary liver transplant,” a hepatic allograft was implanted into the right paravertebral gutter of dogs without disturbing the native liver. Welch followed up this publication with a more complete description published in Surgery in 1956. These auxiliary livers were revascularized by anastomosing the allograft hepatic artery to the recipient aortoiliac system, and by an end-to-end anastomosis of the allograft portal vein to the host inferior vena cava ( Fig. 1-2 ). By including a short length of donor retrohepatic vena cava, Welch avoided anastomosing multiple hepatic veins and instead required just one anastomosis; the upper end of the caval segment of the graft was anastomosed to the recipient vena cava, and the lower end was ligated.

In contrast to other types of transplanted organs, an auxiliary liver transplant allograft underwent a marked shrinkage beginning within 3 to 4 days of the surgery. Initially the atrophy was attributed to liver rejection. The central dogma at the time was that liver size and regeneration were governed by the amount of portal venous inflow (known as the “flow hypothesis” of hepatic homeostasis). Because the portal vein of the auxiliary liver allograft received ample systemic blood from the host vena cava, it was felt that the atrophy was not related to blood flow but was instead ascribed to immunological factors. It would be more than 10 years before the cause of the auxiliary allograft atrophy was fully appreciated and the idea of rejection being the culprit was refuted. It ultimately became apparent that the atrophy was due to the absence of hepatotrophic factors such as insulin, which are present in high concentrations in the splanchnic circulation but were missing from the systemic blood from the vena cava that perfused the auxiliary liver.

In 1960 Michael Francis Addison Woodruff of the University of Otago Dunedin School of Medicine in New Zealand published a compendium of work in transplantation up to 1959, and at that time the only references to liver transplantation were Welch’s two articles on heterotopic liver transplantation and a brief report by Jack Cannon of University of California, Los Angeles (UCLA) published in 1956 that described the liver transplant activities in animals performed at the recently founded UCLA School of Medicine. This article by Cannon was considered for many years to be the first experimental description of an orthotopic liver transplant, until the recently discovered work of Staudacher.

However, by the time Woodruff’s book was published in 1960, there were already two centers—the Peter Bent Brigham Hospital in Boston and Northwestern University in Chicago —that both independently began studying liver transplantation in 1958, each center looking at the field from different vantage points. The investigations from the Brigham Hospital were done under the direction of Francis D. Moore, and because the focus came from a center with an established history with kidney transplantation, this group approached liver transplantation from an immunological perspective with a therapeutic objective. In contrast, the work from the Northwestern University group led by Thomas E. Starzl stemmed from work regarding the metabolic interrelationships of the liver with the pancreas and intestine, which evolved from earlier investigations done at the University of Miami in the field of hepatotrophic physiology. In this circumstance, liver replacement was being performed for the purpose of studying these metabolic relationships. In these investigations the Northwestern University group pioneered a new method of total hepatectomy in which the host’s retrohepatic inferior vena cava was preserved, (heralding the approach that would come to be known as the piggyback variation of liver transplantation in humans ). For liver replacement in the dog, it was simpler to excise the host retrohepatic vena cava along with the native liver and to replace it with the comparable caval segment of the donor. The vena caval anastomosis above and below the liver and the hepatic arterial and biliary tract anastomoses were performed with conventional methods ( Fig. 1-3 ). When different means of portal revascularization were systematically tested in the laboratory at the Northwestern University program ( Fig. 1-4 ), it was discovered that any deviation from the normal portal supply resulted in reduced survival.

The research teams at Northwestern University in Chicago and the Brigham Hospital in Boston were unaware of each other’s activities until late 1959, and direct contact between the programs was not established until the 1960 meeting of the American Surgical Association. By then the cumulative total of liver replacement procedures in nonimmunosuppressed dogs was 111 (80 at the Northwestern University program, 31 at the Brigham Hospital program ). The outcomes from these canine liver transplants were published in 1960 in separate papers and in different journals.

Animal Models: Prerequisites for Canine Replacement

The two prerequisites for perioperative survival of canine liver transplant were independently established in each laboratory, both at the Brigham Hospital in Boston and at Northwestern University in Chicago. The first requirement for a successful canine liver replacement was prevention of ischemic injury to the allograft. At the Brigham Hospital program this was accomplished by immersing the liver in iced saline. At the Northwestern University program the method of hypothermia was influenced by F. John Lewis, who along with Norman Shumway pioneered total body hypothermia for open heart surgery while at the University of Minnesota. The livers were cooled by the intravascular infusion of chilled lactated Ringer solution ( Fig. 1-5 ) and monitoring core temperature with thermal probes. This now-universal step in preservation of organs had never been used before, apparently because of the fear of damaging the microcirculation. In time, better liver preservation was obtained by altering the osmotic, oncotic, and electrolyte composition (i.e., Collins, Schalm, and University of Wisconsin solutions ).

The second prerequisite for successful canine liver replacement was avoiding damage to the recipient splanchnic and systemic venous beds when venous drainage was obstructed during the host hepatectomy and graft implantation. In both laboratories this was accomplished by using external venovenous bypasses to decompress the venous drainage, although the particular details of the bypasses differed at each center.

Animal Models: Pathology of Liver Rejection

Until 1960 the kidney had been the only organ allograft whose unmodified rejection had been systematically studied. With development of the canine liver replacement models at each of the two programs, the pathology of rejection in a transplanted liver could now be studied. These initial histopathological assessments were done by David Brock at the Northwestern University program and Gustav Dammin at the Brigham Hospital program. Most of the transplanted canine livers were destroyed in about 5 to 10 days. The pathological examination of the transplanted livers typically showed a heavy concentration of mononuclear cells, both in the portal triads and in and around the central veins, all with extensive hepatocyte necrosis.

A curious exception was noticed in the sixty-third liver replacement experiment. The serum bilirubin level reached a peak at 11 days but then progressively declined. The predominant histopathological findings in the allograft by day 21 were more those of repair and regeneration rather than rejection. This was the first recorded exception to the existing dogma that once rejection was initiated, it was an inescapable process. Five years later, similar observations were made by Ken A. Porter of St Mary’s Medical School in London, assessing the allografts of long-surviving canine liver recipients from experiments done at the University of Colorado program, where rejection had developed and then spontaneously reversed under stable daily doses of azathioprine.

Immunosuppression: Host Irradiation and Cytoablation

Just when the surgical research in nonimmunosuppressed dogs began to lose momentum, it was dramatically revitalized. From January 1959 to February 1962, there were seven successful human kidney transplantations performed, the first by Joseph Murray at the Brigham Hospital in Boston (work for which he received the 1990 Nobel Prize in medicine) then six more times by the independent teams led by Jean Hamburger and Rene Kuss, both of whom were in Paris ( Table 1-2 ). For these seven transplants the immunosuppression came from preconditioning the patients with sublethal doses of 4.5 Gy total body irradiation. The first two recipients (they received fraternal twin kidneys) had continuous graft function for more than 2 decades without any further posttransplant immunosuppression. They were the first examples of acquired immunological tolerance in humans.

Exploring a substitute for irradiation, Willard Goodwin, a urologist from UCLA, pretreated recipients with myelotoxic doses of cyclophosphamide and methotrexate. One recipient had a prolonged survival of 143 days and had rejection that was successfully reversed several times with prednisone. Despite these initial moderate successes with cytoablation, it quickly became apparent that cytoablation by medication was not going to be a feasible means through which liver transplantation might occur.

Immunosuppression: 6-Mercaptopurine and Azathioprine

The real advances needed for liver transplantation required the arrival of the era of drug immunosuppression, and 6-mercaptopurine (6-MP) is generally considered the drug that heralded in this era. Much of the initial research that would be crucial for immunosuppression for liver transplantation was studied in kidney transplant models. In 1950, working at Wellcome Research Laboratories, Gertrude Elion and George Hitchings used innovative drug development methods to create 6-MP (work for which they received the Nobel Prize in medicine in 1988). The researchers Robert Schwartz and William Dameshek at Tufts Medical School in Boston first established that 6-MP was immunosuppressive and not require overt bone marrow depression to be successful. Using a skin allograft model in rabbits, William Meeker Jr. and Robert Good at the University of Minnesota showed 6-MP provided a modest prolongation of skin allograft survival. Upon learning of the immunosuppressive potential of 6-MP, both Roy Calne (then a surgical trainee) in London and Charles Zukoski at the Medical College of Virginia in Richmond independently performed experiments using transplant models with canine kidney allograft, reporting survival of up to 40 days.

After developing 6-MP, Elion and Hitchings used their drug development techniques to synthesize an imidazole derivative of 6-MP called azathioprine, a prodrug of 6-MP that required processing in the liver to become active and thereby prolonged the effects of the drug. By the end of 1960, both Zukoski, working with David Hume in Richmond, and Calne, who had moved to Boston for a fellowship with Murray, were using azathioprine in kidney transplants with survival results that would sometimes reach 100 days.

Animal Models: Toward Liver Transplantation by Kidney Transplant Experience

Calne’s experiments showing transplant rejection could sometimes be substantially delayed with azathioprine encouraged the Brigham Hospital program in Boston to pursue human kidney transplant trials. When the trials of kidney transplant with azathioprine began in Boston in 1960-61, there were initially high expectations, and the idea of actually transplanting livers seemed less remote. In 1961 William R. Waddell left Massachusetts General Hospital to become chair of surgery at the University of Colorado, where he was joined by Starzl, coming from Northwestern University in Chicago. Their goal at that point was to pursue development of liver transplantation, especially considering the 3 years of experience Starzl had gained at Northwestern University working with the canine hepatic replacement models ( Fig. 1-6 ). Unfortunately, the plans for liver transplantation were shelved when reports of the Boston clinical trial of kidney transplantation described disappointing results. The report by Murray et al, published in the Annals of Surgery , did have one positive element, because it described a kidney allograft transplanted from an unrelated donor in April 1962 that was still functioning 120 days later using azathioprine immunosuppression. That kidney ultimately functioned for another 13 months after this report for a total of 17 months, and it was the first example of 1-year survival of a human organ allograft without host conditioning with total body irradiation. This positive element in the report was tempered by the fact it was the only recipient of the first 13 treated solely with drug immunosuppression that survived for more than 6 months.

In the spring of 1962 the University of Colorado group of Waddell and Starzl, working at the Denver Veterans Administration hospital, obtained a supply of azathioprine and began developing experience with the drug. Initially the plan had been to study azathioprine in a liver transplant model, but it became clear quickly that the operation of liver replacement in dogs was too difficult and fraught with technical challenges to use it to evaluate an immunosuppressive drug. So the group decided to use the simpler canine kidney model first as a precursor to liver transplantation. The results from this transplant model yielded similar results to other laboratories with survival that sometimes approached 100 days. However, two key observations came from these canine transplant models that would affect future immunosuppressive management strategies. The first observation was that the allograft rejection that occurred after azathioprine monotherapy could be reversed by delayed addition of large doses of prednisone. The second observation was that pretreatment of the animals with azathioprine for 7 to 30 days before transplant doubled their mean survival, which to that point had been 36 days.

Human Trials: The Human Kidney Transplant Trials

Beginning in late 1962, the long-standing kidney transplant program at Brigham Hospital in Boston was joined by two other centers in performing human kidney transplantation: the group at the University of Colorado in Denver comprising Starzl and Waddell and the group at the Medical College of Virginia in Richmond led by Hume ( Fig. 1-7 ). The groups at Colorado and Virginia were in close contact with each other, collaborating on ideas, and both realized early that a combination of “azathioprine and steroids” was key to a successful outcome; however, they approached the strategy from different directions. The University of Colorado group reserved steroids for when rejection occurred, which invariably happened with azathioprine monotherapy. The Medical College of Virginia group used reduced-dose steroids from the time of the transplant as part of a dual drug combination.

The University of Colorado group began human kidney transplants in 1962 using a protocol that gave daily doses of azathioprine 1 to 2 weeks before transplant, as well as continuing it after, and added high doses of prednisone to treat any rejection. The successful results of the first 10 kidney cases using this protocol were described in the report “The Reversal of Rejection in Human Renal Homografts With Subsequent Development of Homograft Tolerance.” The term tolerance referred to the time-related decline of need for maintenance immunosuppression. Based on their results using this protocol, Starzl and the University of Colorado group concluded that renal transplantation had reached the level of a bona fide (albeit still flawed) clinical service.

In 1963 a small conference organized by the National Research Council ultimately became a landmark event in transplantation. Twenty-five of the leading transplant clinicians and scientists from around the world assembled to review the current status of human kidney transplantation. The results were very discouraging because less than 10% of the several hundred human allograft recipients had survived more than 3 months. Of those treated with total body irradiation for immunosuppression, only 6 patients had survival close to 1 year. The results of those with drug-based immunosuppression were equally poor, as Murray reported that of his first 10 patients treated with 6-MP or azathioprine, only the one survived a year, whereas the others died within 6 months. Some participants at the conference began to question whether human transplantation could still be justified. Ultimately the Colorado group described their success with their immunosuppressive protocol of using azathioprine and adding large doses of prednisone with any rejection, which allowed a 1-year survival rate that exceeded 70%. Because the Colorado group, which had been a late invite to the meeting, reported more surviving recipients than the rest of the world’s other centers combined, the audience was incredulous, and it provoked intense discussions. However, the fact that Starzl brought with him the wall charts (on the advice of Goodwin, who was aware of the results) that detailed the daily progress, urine output, and laboratory work of each patient, quelled the debate ( Fig. 1-8 ). As Clyde Barker of the University of Pennsylvania described the events: “The gloom was dispelled by only one presentation given by Tom Starzl, a virtually unknown newcomer to the field, who was invited to the conference as an afterthought…. The outlook for renal transplantation was completely changed by Starzl’s report.”

Before the 1963 National Research Council conference there were only the three active kidney transplant centers in the United States (Brigham Hospital, University of Colorado, and Medical College of Virginia). Within a year of the conference, and as word of the effectiveness of this new immunosuppression protocol spread, 50 new transplant programs began in hospitals throughout the United States, with a similar proliferation of transplant centers across Europe. Some of the benefits of kidney transplantation proved to be truly long lasting in some cases, because eight of the recipients from the University of Colorado program from 1962 to 1963 still had their kidney transplants 40 years later (making them the longest-surviving organ allograft recipients in the world) and some of them have lasted 50 years.

Human Trials: The Human Liver Transplant Trials of 1963

Although the follow-up evaluations of the kidney transplant trials were still short, the successful human kidney transplant experience at the Colorado program encouraged the decision to go forward with the exponentially more difficult initiative of liver transplantation ( Fig. 1-9 ). The first attempted human liver transplant was on March 1, 1963, in a 3-year-old boy with biliary atresia named Bennie Solis. Bennie had been operated on numerous times previously and had deteriorated to the point of being unconscious and ventilated. Unfortunately, Bennie bled to death during the actual transplant operation, because of the many high-pressure venous collaterals that had formed and an uncontrollable coagulopathy. This result occurred despite the fact that the operative team had performed more than 200 similar transplant operations in animal models. The complexity and difficulty was so extreme, it took the team several hours just to make the incision and enter the abdomen, because of the significant collateralized adhesions.

Two more liver transplantations were performed over the next 4 months in two adults, one transplanted May 5, 1963, for a hepatoma, and the second transplanted June 3, 1963, for a cholangiocarcinoma. The donor procurement for these transplants had successful allograft preservation accomplished by transfemoral infusion of a chilled perfusate into the aorta of the non–heart-beating donors after cross-clamping the aorta at the diaphragm ( Fig. 1-10 )—in much the same way as the first stage of the multiple organ procurement operation still performed today. The cold ischemia time for the two procurements was 2.5 hours and 8 hours, respectively, and neither recipient had any significant ischemic damage as evidenced by modest increases in the liver enzyme levels after transplant. For the operative procedure the various anastomoses were performed in the same way as in the dog experiments except for the biliary tract reconstruction. (The complete operation was drawn in 1963 [ Fig. 1-11 ], and that picture could still be used today to depict a human liver transplantation.) The immunosuppression protocol for the recipients in the University of Colorado group’s liver transplantation trials derived from that center’s experience in the human kidney transplant trials, with azathioprine administered both before and after transplantation, adding a high-dose course of prednisone with the onset of rejection.

Although both procedures seemed satisfactory, these recipients—the second and third recipients of the trial, died after 22 and 7.5 days, respectively. Both patients died in part because of pulmonary emboli, although interestingly, both were also found to have extrahepatic micrometastasis of their cancers at autopsy, although with no rejection of the allograft. The strategy of controlling the coagulation using transfusion of blood products and ε-aminocaproic acid for fibrinolysis, which was adopted following the uncontrolled coagulopathy of the first transplant, had unintentionally backfired. During the implantation of the livers, passive venovenous bypass with plastic tubing was used, similar to the technique used in the canine model. However, in the humans who had been given coagulation-promoting therapy, clots formed in the bypass tubing and passed to the lungs, causing abscesses and lung damage that contributed to their deaths (and to the next two recipients to follow). Ironically, the use of the venovenous bypass to decompress the venous system—something that was so crucial to survival in the canine experiments—was not necessary for most human recipients. (A motor-driven venovenous bypass system introduced in Pittsburgh in the 1980s and later use of percutaneous catheters have made the procedure easier, but in many centers bypass is only used selectively, if at all, and never in infants or small children). Ultimately, venous decompression was shown to be expendable in dogs submitted to common bile duct ligation several weeks in advance of transplantation—an animal model of cirrhosis and portal hypertension—and the venous collaterals that developed enabled transplantation without venovenous bypass.

Human Trials: The Liver Transplant Moratorium

During the last half of 1963, two more liver transplantations were performed by Starzl’s group at the University of Colorado, and one each at the Brigham Hospital in Boston by Moore and at the Hospital St Antoine in Paris by Jean Demirleau ( Table 1-3 ). The transplant in Paris was the first liver transplant in Europe, and used a 71-year-old donor into a 75-year-old recipient, making this also the first transplant using what would be today called a “marginal donor.” The operation lasted 4 hours, but the patient died 3 hours after transplant from uncontrollable fibrinolysis.

After the deaths of these seven patients in three different centers, there was great pessimism worldwide that the operation was too difficult to be practical, that the methods of organ preservation were inadequate for an organ so sensitive to ischemic damage, and that the available immunosuppression options were too primitive to allow success. This sentiment was reinforced by the fact that long-term survival following liver transplantation had not yet even been achieved in the experimental animal models. Clinical activity in liver transplantation ceased for 3.5 years between January 1964 and the summer of 1967. The worldwide moratorium was voluntary, but the decision to stop was reinforced by widespread criticism that transplantation was too formidable to be practical. During the moratorium on liver transplantation, the field did not stay still; problems that contributed to the failures of the transplants of 1963 were addressed, and advances were made across the field, in immunosuppression, organ preservation, and operative techniques.