Recent improvements in total parenteral alimentation have significantly extended the survival for victims of short-bowel or severe malabsorption syndromes. Nevertheless, the long-term prognosis remains poor for patients whose residual bowel does not hypertrophy enough to allow at least partial enteral nutrition. Limiting factors for these individuals are recurrent infections that develop at venous access sites and the almost inevitable progressive liver dysfunction that results from prolonged high-calorie intravenous feeding. In such patients, the only viable alternative may be the attempted replacement of the diseased or absent bowel with a functioning allograft.

The number of patients who could benefit from bowel transplantation has been estimated to be approximately 1 or 2 per million of population per year. Suitable adult candidates are primarily patients who have survived massive bowel resection for treatment of benign conditions such as extensive Crohn's disease, vascular compromise, or trauma. Bowel transplantation has occasionally been recommended, as part of the multivisceral transplant used for reconstruction in patients whose extensive neoplasia was removed by ‘upper abdominal exenteration’. In children, the short-gut syndrome typically results from conditions such as congenital neuromyopathy, extensive atresia, microvillus atrophy, or extensive resection following midgut volvulus or necrotizing enterocolitis.

Early attempts at clinical bowel transplantation using conventional immunosuppression with azathioprine, steroids, and antilymphocyte sera were extremely discouraging. Thus, by 1985, only seven such transplants had been reported. All recipients died, without adequate allograft function, within 3 months of the procedure. Unresolved obstacles, unique to bowel transplantation, were presented by the inescapable microbial colonization of the allograft, the immunological consequences of the large volume of immunocompetent donor lymphocytes necessarily transplanted into the recipient, an inadequate means of identifying rejection, and a high incidence of thrombotic complications. This bleak experience discouraged most transplant centres from pursuing additional clinical trials until further advances, including more effective and less toxic immunosuppressive agents, became available.

It was not until August of 1988 that the first successful clinical small-bowel transplant was performed in Kiel, Germany. The recipient had undergone total small-bowel resection for extensive mesenteric venous thrombosis. Ten weeks later, a 60-cm bowel segment was transplanted into the patient from an HLA-compatible sibling. Despite the excellent histocompatibility, multiple rejection episodes required aggressive immunosuppression with cyclosporin, steroids, and antithymocyte globulin. Lifethreatening infectious complications, not unexpectedly, were encountered. Nevertheless, the recipient eventually became independent of parenteral nutrition and remains well over 4 years later.

In November 1988, successful transplantation of the liver and incontinuity small intestine was first accomplished in London, Ontario, Canada. The patient had suffered extensive bowel infarction secondary to a hypercoagulable state. The liver allograft corrected the coagulopathy and parenteral nutrition could be withdrawn from this patient 8 weeks after the transplant procedure.

These two cases established that small-bowel transplantation could provide a successful solution for the short-gut syndrome in patients with or without concomitant liver dysfunction. Nevertheless, the fact that a mere handful of successful cases has since been reported, from less than ten centres worldwide, emphasizes that this procedure is still not a realistic therapy, except for highly selected patients in whom all other treatment options have been exhausted.

As is true with kidney or pancreas transplantation, a suitable allograft for small-bowel replacement can be removed from a living donor but is more frequently obtained, in combination with other donor organs, from a cadaveric source. Specific considerations related to intestinal-tract procurement have been directed toward minimising microbial contamination from the bowel lumen and reducing the volume of lymphoid tissue retrieved with the allograft. Thus, provided there is time prior to organ retrieval, the donor is pretreated with a hyperosmolar cathartic in combination with non-absorbable antibiotics and antifungal agents. Based upon preclinical studies emphasizing that graft-versus-host disease (GVHD) can be induced by functional T cells in the intestinal transplant, donor lymphocyte depletion, using antilymphocyte globulin or the monoclonal antibody OKT3 prior to organ retrieval or irradiation of the excised allograft, have also been recommended. Interestingly, as discussed below, GVHD in human bowel transplant recipients has been encountered infrequently, making the necessity of these measures questionable.

The surgical technique of bowel allograft procurement depends upon the organs that are required. If the intestine is to be transplanted alone, a suitable segment is isolated, dividing the mesenteric vessels near their origin in cadaver donors, or more distally in a living donor. In the typical multiorgan procurement, the liver with the coeliac axis and proximal portal vein is separated from the intestine which remains in continuity with the distal portal vein and superior mesenteric artery (Fig. 1) 720. For combined small bowel-liver transplantation, the donor jejunum, ileum, liver, and abdominal aorta containing the origins of the superior mesenteric and coeliac arteries are isolated. The liver and bowel, connected by the portal vein and aortic segment, are then perfused and removed, en bloc. The most complex procedure, in which a ‘cluster’ graft is required, involves en-bloc removal of the liver, pancreas, and variable amounts of duodenum and jejunum (Fig. 2) 721.

Little information has been accumulated regarding the optimal method for preservation of human bowel allografts. Earlier studies suggested that continuous pulsatile perfusion might be required to maintain acceptable viability during ex-vivo storage periods of even 6 h. It is now clear, however, that adequate bowel preservation, at least for the short term, can be achieved by using an intracellular type of cold-storage fluid, such as Collin's or the University of Wisconsin solutions, which are employed for the other donor organs. Currently, successful bowel transplantation is accomplished after simple in-situ perfusion of the donor vessels followed, optionally, by ex-vivo flushing of the lumen of the excised bowel. Simple storage in the cold preservation fluid is effective for as long as 10 h. Whether more effective approaches can be devised to better maintain the integrity and, thereby, the early functional capacity of the allograft remains to be determined.

The technical aspects of bowel replantation alone are quite straightforward. Revascularization of the allograft is preferentially accomplished by anastomosing the donor portal or superior mesenteric vein to the recipient portal vein in the hepatic hilum or retropancreatic area. If it is not possible to construct a satisfactory portal–portal anastomosis, venous drainage can alternatively be provided into the recipient's systemic circulation. Arterialization is provided by anastomosis of the donor superior mesenteric artery to recipient aorta or iliac artery. For grafts that include the liver, which has been the more commonly performed procedure, revascularization is more complex (Fig. 3) 722. The supra- and infrahepatic vena caval anastomoses are constructed end-to-end as for orthotopic liver transplantation. The transplant bowel portal venous drainage is through the non-transected donor portal vein; however, recipient portal venous drainage must also be provided. This may be accomplished by joining recipient and donor portal veins in end-to-side fashion, as depicted in Fig. 3 722, or by constructing a portacaval shunt between recipient vessels. Arterialization is provided by anastomosing the donor aortic conduit (Fig. 3) 722 or a Carrel patch encompassing the origins of both the coeliac axis and superior mesenteric artery to the recipient infrarenal aorta. Complete restoration of continuity of the bowel is usually not undertaken at the primary transplant procedure. The distal end of the bowel is typically exteriorized to provide postoperative decompression and convenient access for the serial postoperative biopsies which are performed to evaluate the allograft status. Enteric anastomoses are subsequently performed when the allograft is known to be well vascularized and free of rejection.

Many of the technical problems that previously limited the success of bowel transplantation have been resolved with the use of current vascular techniques. Nevertheless, convalescence for these recipients is difficult, with recurrent rejection and septic episodes typically leading to protracted stays in hospital, even if no major surgical complications arise.

Unlike other organ transplants, the bowel allograft does not immediately regain normal function. Thus, total parenteral nutrition must be continued, sometimes for months postoperatively, when the attendant risks of catheter sepsis are even further increased because of the need for concomitant immunosuppression. Further complicating the post-transplant management is the erratic absorption, during periods of allograft dysfunction, of orally administered immunosuppressive agents.

At this stage of development of clinical bowel transplantation, the most efficacious postoperative immunosuppressive regimen remains undefined. Current reports indicate that substitution of the new macrolide immunosuppressant FK506 for conventional suppression provides a significantly improved outlook following intestinal transplantation. More than 30 patients treated with FK506 since 1990 have had successful bowel transplants.

In addition, there is no simple assay available for reliably detecting early bowel allograft rejection. Since rejection in experimental models is characterized by oedema, loss of villi, bleeding and ulceration of the mucosa, and submucosal mononuclear cell infiltrates, most centres have performed serial biopsies to monitor allograft status. However, histopathological examination of these partial-thickness biopsies does not always provide a reliable assessment of early rejection. Thus, immunohistochemical investigations, sequential measurement of intestinal permeability, and various measures of intestinal absorption are also being evaluated. A significant diagnostic advantage enjoyed by recipients of combined liver–small-bowel transplantation is the availability of better-defined functional and morphological criteria indicative of early rejection in the hepatic allograft. As a result, initiation of rejection treatment is typically undertaken on the basis of these criteria. This would also be expected to reverse any occult rejection that might be occurring simultaneously in the bowel. Acute rejection episodes following bowel transplantation are currently treated with high-dosage steroids and antilymphocyte preparations, such as ALG or OKT3, as in recipients of other organ allografts.

Although acute rejection episodes in human bowel allograft recipients have been encountered frequently, the incidence of graft-versus-host disease has been low. Only one documented transient episode of acute graft-versus-host disease has been reported. This resolved without any specific treatment and with no subsequent evidence of chronic graft-versus-host disease. Thus, the need for intensive lymphocytolytic treatment of the donor or the excised allograft is not clear.

Recipients of bowel allografts remain susceptible to the same infectious complications as other transplant recipients who are treated with comparable immunosuppressive regimens. Two specific problems have a particular impact on this group of patients. The first of these is bacteraemia and fungaemia due to translocation of these organisms from the bowel lumen as a result of an inadequate barrier function of the allograft mucosa. The primary microbial species responsible for these events are Candida and the aerobic Gram-negative bacilli (Enterobacteriaceae and Pseudomonas aeruginosa) that normally populate the gut lumen. Of note, the anaerobic flora of the gut is rarely involved in such microbial translocation, and, indeed, provides some protection against superinfection—a phenomenon termed colonization resistance.

To counterbalance the inadequate barrier function of the bowel allograft, particularly in the first 6 months post-transplant, one of a variety of prophylactic regimens, which leaves the anaerobic flora intact, may be administered. Based upon experience in neutropenic cancer patients, either a fluoroquinolone (such as norfloxacin, ciprofloxacin, or ofloxacin), or a non-absorbable antibiotic combination (such as vancomycin, gentamicin, or polymyxin, plus fluconazole or amphotericin), is typically prescribed. Whatever the antimicrobial regimen chosen, the gut flora should be monitored, since any aerobic Gram-negative rods or yeast species colonizing the allograft will likely translocate to the bloodstream, necessitating revisions in the basic prophylactic regimen.

The second special infection-related problem in bowel transplant recipients is Epstein–Barr virus associated B-cell lymphoma, which occurs in over 10 per cent of long-term survivors of bowel allografts (as opposed to 1–5 per cent of other organ recipients). The pathogenesis of this process is complex and has to do with immunosuppression-induced Epstein–Barr virus reactivation (particularly by monoclonal or polyclonal antilymphocyte antibodies), secondary infection and immortalization of B cells, and failure of the normal surveillance mechanisms to eliminate these malignantly transformed cells. In addition, some investigators have suggested that the antilymphocyte regimens used to pretreat the donor in the effort to limit graft-versus-host disease might be causally related to the subsequent development of B-cell lymphomas in the recipient. The hope is that new protocols, including the addition of effective antiviral drugs to the immunosuppressive protocol, will create a therapeutic programme that is both safe and effective.

There has been a resurgence of interest in transplantation of the small intestine since the first successful cases were reported in 1988. No official registry of bowel transplants has yet been established, making a complete summary of the experience in humans difficult. Communication with centres currently active in the field indicate that less than 100 of these procedures have been performed to date. Most of these have been undertaken since 1986.

The number of surviving bowel grafts which are sufficiently functional to allow withdrawal of all parenteral nutrition is small at this point, and the majority of these have been simultaneously transplanted with the liver. Whether this is because the intestine is less vigorously rejected when accompanied by the liver or simply because rejection is more easily diagnosed in liver allografts and therefore treated earlier, has not been established. Nevertheless, sporadic reports of chronically functioning isolated bowel allografts, including the world's first successful human transplant, are appearing. These successes indicate that, with further advances, it may soon be possible reliably to transplant the small intestine without the liver. Recently, at the University of Pittsburgh, an intensive clinical trial of bowel (with or without simultaneous liver) transplantation has been undertaken. The results, although preliminary, are encouraging.

In summary, small-intestinal transplantation represents a logical treatment for patients with short-gut syndrome whose survival is dependent upon total parenteral nutrition. Many unresolved obstacles, including inadequate immunosuppression, lack of a reliable means to maintain the allograft's functional barrier to infection, and the insensitivity of currently available diagnostic assays for rejection, have limited the applicability of the procedure. Nevertheless, the recent progress in this field will undoubtedly stimulate more widespread clinical trials in the near future. If bowel transplantation without simultaneous liver replacement can be performed with reliable success, it would be anticipated that this procedure will become the preferred treatment for most patients with short-gut syndrome. Successful small-bowel transplantation would obviate the considerable morbidity, mortality, and costs of total parenteral nutrition, and it would allow the patients to return to a relatively normal lifestyle. Moreover, unlike the situation with other organ transplants, the availability of suitable bowel allografts, from both living and cadaveric donors, should be more than sufficient to satisfy the need of all potential recipients who might benefit from small-intestine transplantation.

Deltz E, Schroeder P, Gebhardt H, et al. Successful clinical small bowel transplantation: report of a case. Clin Transplantation 1989; 3: 89–91.

Grant DR. Immunosuppression for small bowel transplantation. Clin Transplantation 1991; 5: 563–7.

Grant D, et al. Intestinal permeability and bacterial translocation following small bowel transplantation in the rat. Transplantation 1991; 52: 221–4.

Grant D, et al. Successful small-bowel/liver transplantation. Lancet 1990; 335: 181–4.

Hansmann ML, Hell K, Gundlach M, Deltz E, Schroeder P. Immunohistochemical investigation of biopsies in a successful small-bowel transplantation. Transpl Proc 1990; 22: 2502–3.

Kirkman RL. Small bowel transplantation. Transplantation 1984; 37: 429–33.

Revillon Y, Jan D, Goulet O, Ricour C. Small bowel transplantation in seven children: preservation technique. Transpl Proc 1991; 23: 2350–1.

Schroeder P, Goulet O, Lear PA. Small-bowel transplantation: European experience. Lancet 1990; 336: 108–111.

Starzl TE, et al. The many faces of multivisceral transplantation. Surg Gynecol Obstet 1991; 172: 335–44.

Todo S, et al. Cadaveric small bowel and small bowel liver transplantation in humans. Transplantation 1992; 53: 369–76.

Todo S, Tsakis A, Abu-Elmagd K, Reyes J, Starzl TE. Current state of intestinal transplantation. Adv Surg, 1994; 27: 295–316.