- Email: [email protected]
Transplantation of the abdominal wall
ARTICLES
Articles
Transplantation of the abdominal wall
David M Levi, Andreas G Tzakis, Tomoaki Kato, Juan Madariaga, Naveen K Mittal, Jose Nery, Seigo Nishida, Phillip Ruiz
Summary Background Closure of the abdomen in patients undergoing intestinal transplantation can be extremely difficult, if not impossible. We describe our initial experience with abdominal wall allotransplantation to facilitate abdominal closure. Methods We undertook nine cadaveric abdominal wall composite allograft transplants in eight patients. The graft’s blood supply was based on the inferior epigastric vessels left in continuity with the donor femoral and iliac vessels. Skin biopsies were undertaken randomly and when rejection was suspected. Vessel patency was monitored by doppler ultrasound. Findings Six patients have survived, five of whom have intact, viable abdominal wall grafts. Two patients have had a clinically mild episode of acute rejection of the skin of the abdominal wall that resolved with corticosteroid therapy. No clinically apparent graft-versus-host disease has been noted. Interpretation Transplantation of an abdominal wall composite allograft can facilitate reconstruction and closure of the abdominal compartment in intestinal transplant recipients with complex abdominal wall defects. Lancet 2003; 361: 2173–76
Introduction Intestinal failure—the permanent loss of sufficient functional intestinal absorptive surface area to sustain life—remains a catastrophic disease entity. Although total parenteral nutrition is essential in the management of affected patients, such treatment can be associated with serious complications, especially when used long term. Complicated intestinal failure is characterised by cholestatic liver disease induced by total parenteral nutrition, a loss of all central venous access sites for total parenteral nutrition, or both. Intestinal transplantation is reserved for patients with complicated intestinal failure. Recent advancements in immunosuppression, graft surveillance, and surgical technique have resulted in improved survival for these patients.1,2 We have previously reported a technical description of intestinal and multivisceral transplant procedures.2 One great technical problem is closure of the abdomen at the end of transplantation. Since 1994, more than 160 intestinal failure patients have received transplants at our institution. Many of these patients have undergone complete resection of the midgut and have lost the domain of the abdominal compartment. Additionally, some had severely damaged abdominal walls from injuries, repeated laparotomies, ostomy construction, wound infections, tumours, and enterocutaneous fistulae. The use of plastic surgical techniques to recruit tissue for abdominal closure can be helpful,3 but in about 20% of patients in our series adequate tissue for closure does not exist. Failure to close the abdomen at the time of transplant leaves the recipient with a large, open wound and, at best, a staged closure needing several surgical procedures and a long period of rehabilitation. Complications can be life threatening, and include intraabdominal infection, injury to unprotected organs, bleeding, and fistula formation. We propose that transplantation of a cadaveric, composite allograft consisting of the abdominal wall could be an effective way to address this problem. We describe our initial experience with transplantation of the abdominal wall to facilitate reconstruction of the abdominal compartment during intestinal transplantation.
Methods
Division of Transplantation, Department of Surgery (D M Levi MD, A G Tzakis MD, T Kato MD, J Madariaga MD, N K Mittal MD, J Nery MD, S Nishida MD), and Department of Pathology (P Ruiz MD), University of Miami/Jackson Memorial Medical Center, Miami, Florida, USA Correspondence to: Dr David M Levi, University of Miami School of Medicine, Daughtry Family Department of Surgery, PO Box 015809 (M840), Miami, FL 33101 USA (e-mail: [email protected])
The abdominal wall composite graft is a full thickness, vascularised, myocutaneous free flap. It consists of one or both rectus abdominus muscles, with the investing fascia, the overlying subcutaneous tissue, and skin. The blood supply is derived from the donor inferior epigastric vessels, left in continuity with the larger femoral and iliac vessels (figure 1). Intestinal graft recipients received either an isolated small bowel graft or a multivisceral graft. The latter includes the donor stomach, duodenum, liver, pancreas, and small intestine. The decision to include an abdominal wall composite graft in patients undergoing intestinal transplantation was made if conventional abdominal
THE LANCET • Vol 361 • June 28, 2003 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet.
2173
ARTICLES
Age (years) Sex
Diagnosis
53 1 2 21 41 6 15 19
Midgut strangulation MV Gastroschisis MV Hirschsprung’s disease MV Blunt trauma ISB Blunt trauma ISB Midgut volvulus MV Blunt trauma ISB Penetrating trauma ISB
Female Female Male Male Male Female Male Male
Transplant Patients’ status type (months of follow-up) Alive (23) Died Alive* Alive (13) Died Alive (9) Alive (3) Alive (1)
MV=multivisceral. ISB=isolated small bowel. *Without abdominal wall graft.
Patients’ characteristics
Figure 1: Implantation of abdominal wall composite graft in the recipient
closure was not possible. All transplants were done with informed consent. Intestinal transplantation has been undertaken at our institution in accord with protocols approved by the investigational review board since 1994. As for many of the technical alterations to these procedures that have been developed, a separate protocol for the inclusion of the abdominal wall graft has not been submitted to the investigational review board. Induction immunosuppression consisted of the anti-CD52 monoclonal antibody alemtuzumab, which was given intravenously (0·3 mg/kg) immediately preoperatively, immediately postoperatively, and on postoperative days 3 and 7. Maintenance immunosuppression was based on tacrolimus without corticosteroids. The target trough concentration of tacrolimus in serum was 10 g/L. One child received a transplant before we started using alemtuzumab in paediatric patients; instead, they received daclizumab for induction with tacrolimus and methylprednisolone for maintenance immunosuppression. Intravenous corticosteroids were used for the treatment of rejection. Procurement of the graft was done as part of the cadaveric, heart-beating donor, multiorgan procurement procedure.4,5 The procedure began with a median sternotomy and a bisubcostal incision. Longitudinal incisions were made following both lateral edges of the rectus sheath. These incisions were continued into the groins bilaterally. The common femoral vessels were identified. Finally, a transverse, suprapubic incision was made, connecting the two longitudinal incisions. The assessment and dissection of the rest of the abdominal and thoracic organs proceeded as previously described.4 The abdominal wall graft was packed with ice in situ while the other organs were flushed and removed. When the other organs had been procured, the distal aorta was cannulated and the graft was flushed with cold preservation solution. The graft was removed in one piece with the femoral and iliac vessels, with a short segment of distal aorta and inferior vena cava. Closure of the donor’s abdomen was facilitated by mobilising skin and subcutaneous tissue flaps from the lateral abdomen and flanks. After reperfusion of the isolated intestinal or multivisceral graft in the recipient, we transplanted the abdominal wall graft as a separate organ. The inclusion of an abdominal wall graft added about 2 h to the procedure’s operative time. Much like a kidney allograft,
2174
the vessels of the abdominal wall graft were implanted into the recipient’s common iliac artery and vein. Alternatively, the infrarenal aorta and inferior vena cava were used in one child. In two children who received multivisceral grafts, the arterial blood supply to the abdominal wall graft was from the donor distal aorta and the venous drainage was via the donor infrahepatic vena cava. The graft was incorporated into the recipient’s abdominal wall during closure of the abdomen. The graft, with its long vascular pedicle, was rotated and positioned according to location of the abdominal wall defect. The skin of the abdominal wall graft was left intact and was given no special treatment; normal skin colour indicated adequate perfusion. The flow through the inferior epigastric vessels of the graft was monitored with a handheld doppler ultrasound device. We undertook biopsies of the skin of the graft at randomly and when rejection was suspected on clinical grounds. The biopsy tissue was placed in a 10% buffered formalin solution, processed, and paraffin-embedded. 5 m sections were stained with haematoxylin and eosin. We recorded the patients’ demographics and diagnosis, type of transplant, graft and patient survival, rejection episodes, and reason for graft loss.
Results In eight patients (four adults and four children) we used an abdominal wall composite graft to facilitate coverage of the isolated intestinal or multivisceral graft and closure of the abdomen. All patients had small abdominal compartments from previous surgical resections and severely damaged abdominal walls from previous incisions, wound infections, fistulae, and stomata. Patients’ characteristics are shown in the table. All transplants were undertaken without HLA matching. Seven recipients received ABO identical grafts; one child (blood type A) received an ABO
Figure 2: Abdominal wall composite graft implanted during intestinal transplant procedure
THE LANCET • Vol 361 • June 28, 2003 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet.
ARTICLES
Figure 3: Photomicrographs of a skin biopsy specimen showing acute rejection Haematoxylin and eosin stain. (A) Low-power (⫻100) magnification of skin with perivascular chronic inflammatory cell infiltrates in the superficial dermis and some lymphocytic exocytosis. (B) Magnification ⫻200. Perivascular infiltrate (solid dark arrow) with reactive lymphoid cells, erythrocyte extravasation (dotted arrow), and exocytosis (white arrow).
identical multivisceral graft followed 6 days later with a type O abdominal wall graft from a second donor. The abdominal wall from the first donor was not available in this case because consent for its procurement was not obtained from the donor’s family. One other patient received an ABO identical abdominal wall graft 3 days after the initial intestinal transplant procedure. This time interval enabled us to find a larger donor with a greater surface area abdominal wall than the initial donor. With the exception of these two recipients, the abdominal wall grafts originated from the same donor as the intestinal graft and were transplanted concurrently with the intestinal graft. Figure 2 depicts an abdominal wall graft at the conclusion of the transplant procedure. The size of the abdominal wall graft depended on the size of the donor and the width of the rectus muscles. The elasticity of the graft, especially the skin, and the tendency of the graft to shrink in cold, hypertonic preservation solution, made accurate measurement of the graft’s surface area difficult. The surface area of coverage provided by the abdominal wall graft ranged from 150 cm2 to 500 cm2. Six of the eight patients are alive; five have functioning, viable abdominal wall composites grafts. These five patients have been followed up for 1, 3, 9, 13, and 23 months post-transplant. Wound healing was uneventful and proceeded at a rate similar to that for the recipient’s native skin. The intestinal grafts have been protected expediting patient recovery. In one patient, a 2-year-old child, we removed the abdominal wall graft because of infarction. This infarct was probably related to hypovolaemia and hypotension, although a mechanical problem cannot be excluded with certainty. The child eventually recovered after the abdominal wall defect closed by secondary intention. Another child died after sepsis and multiorgan failure with an intact abdominal wall graft. One adult patient died in the early postoperative period after receiving two isolated intestinal and abdominal wall grafts. The first graft infarcted because of hypoperfusion from intraabdominal haemorrhage. The patient eventually died from sepsis after retransplantation. Seven of the eight patients received alemtuzumab and tacrolimus maintenance according to an established protocol.6 The other, a child, received daclizumab as an induction agent instead of alemtuzumab. The five
surviving patients with abdominal wall grafts received alemtuzumab induction. Of these five, one adult has had two episodes of mild rejection of the intestine, proven by mucosal biopsy, without concurrent rejection of the skin of the abdominal wall graft. One recipient of an isolated small bowel graft had moderate rejection of the intestine without involvement of the abdominal wall graft. Additionally, two patients have had an episode of rejection of the abdominal wall graft without involvement of the intestinal graft. This manifested clinically as erythema of the skin of the composite graft. No other signs or symptoms were noted. One patient had homogeneous graft skin involvement and the other had a patchy, macular rash. Both cases arose about 1 month after the transplant procedure. The histopathological alterations, shown in figure 3, included focal spongiosis with lymphocytic exocytosis and focal epidermal necrosis, a substantial perivascular lymphocytic infiltrate with eosinophils and neutrophils, and venulitis. The lymphocytes were enlarged and reactive. These changes were consistent with acute rejection. The patients were treated with intravenous corticosteroids, and complete resolution took about 10 days and was confirmed by biopsy. Notably, one of these patients had received an ABO non-identical abdominal wall graft 6 days after the multivisceral transplant procedure. The other also received an abdominal wall graft several days after the intestinal transplant but from a different ABO identical donor. No patient in this series had signs or symptoms of graftversus-host disease, and none died as a results of complications of the abdominal wall graft.
Discussion We have shown that transplantation of the abdominal wall is feasible and safe; and for five patients this procedure has proven instrumental in the success of their transplant. It allowed for primary closure of the abdomen, avoided the potential morbidity of exposed viscera, and permitted early mobilisation and rehabilitation of these patients. The procurement of the abdominal wall graft does not interfere with the procurement of other organs and tissues. If complications arise, the graft can be removed fairly easily. The abdominal wall graft is procured after the other organs have been removed from the cadaver. Consequently, the quality of those organs is not
THE LANCET • Vol 361 • June 28, 2003 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet.
2175
ARTICLES
jeopardised. Since the iliac vessels are included with the abdominal wall graft, the carotid and subclavian vessels can serve as excellent vascular grafts for the other organs to be transplanted as needed. Graft implantation is straightforward and the large diameter of the anastamosed area diminishes the risk of technical failure. Careful attention needs to be directed to the three-dimensional orientation of the vascular pedicle of the abdominal wall graft, to avoid mechanical occlusion. Transplantation of the abdominal wall composite graft can take place during the intestinal transplant procedure, or several days later with a graft from a different donor. Delaying implantation of the abdominal wall graft allows perioperative oedema to diminish before abdominal closure and the patient’s condition to stabilise. This strategy may be preferable when the recipient has an especially large defect in the abdominal wall. Ideally, a small intestinal graft would be transplanted, and oedema would decrease over several days, followed by the implantation of an abdominal wall graft obtained from a larger donor. In our experience, HLA matching is of little clinical significance in intestinal transplantation. Therefore, we did not consider HLA matching in this group of patients. Matching the donor and recipient size and ABO compatibility are of much greater importance. Rejection can be effectively prevented and treated with modern immunosuppression. Morevoer, because the skin of the abdominal wall graft is accessible for visual inspection and biopsy, monitoring for rejection is easy. Induction with alemtuzumab seems to be effective in combination with low-dose tacrolimus maintenance. With this corticosteroid-sparing regimen, number and severity of rejection episodes have been reduced in intestinal transplant patients; many such patients have never had rejection.6,7 Although not yet observed, we speculate that chronic rejection of the abdominal wall graft might not be a serious event, because a fibrotic graft may still provide satisfactory coverage of the abdominal organs. For five patients, transplantation of the abdominal wall proved instrumental in their recovery. For the survivor in whom the abdominal wall graft thrombosed, we were able to remove the graft without compromising the intestinal graft and revert to conventional techniques for abdominal closure. Removal of the graft did, however, lead to a delay in feeding and mobilisation of the patient, and contributed to a hospital stay of over 2 months. Two deaths in this small series were not related to the inclusion of the abdominal wall graft. We have described initial experiences with a novel technique. It is difficult to make a direct comparison between this small group and other intestinal and multivisceral transplant recipients. In our experience, which encompasses some 160 intestinal and multivisceral transplants, some patients with intestinal failure have devastating tissue loss that makes primary closure after transplant impossible. We propose that the addition of an abdominal wall composite graft is a feasible solution to this difficult problem. Human composite allografts have been previously described in the form of transplantation of the hand. To date, 14 cadaveric hand transplants have been undertaken in 11 patients, with encouraging early results.8,9 By contrast with the hand transplant recipients, who are
2176
generally in good health, patients with intestinal failure suffer from a life-threatening disease and need much treatment in hospital. Furthermore, because they are already receiving an intestinal graft—and, hence, therapeutic immunosuppression—the inclusion of an abdominal wall graft carries no additional risk of complications related to immunosuppression. Unlike the abdominal wall graft, hand composite grafts include an obligatory, small amount of donor bone-marrow cells, which may assert an immune modulatory effect.10 Finally, the function of the abdominal wall graft is different to that of the hand graft; the purpose of the abdominal wall graft is to provide biological coverage of the newly transplanted abdominal organs. Recent improvements in immunosuppression, graft surveillance, and surgical technique have made intestinal transplantation a potentially life-saving option for patients with complicated intestinal failure. In intestinal failure patients with irreparable abdominal wall defects, the idea of transplanting the abdominal wall from a cadaveric donor is a logical step. Future applications might include transplantation of the abdominal wall composite graft as an isolated organ transplant in patients with massive abdominal wall defects after trauma or complicated surgery. Contributors D Levi obtained the data, and wrote and revised the manuscript; A Tzakis developed the surgical procedure; T Kato, N Mittal, and S Nishida directed patient management and graft surveillance; P Ruiz provided the pathological interpretation of the skin biopsy specimens; D Levi, A Tzakis, T Kato, J Madariaga, J Nery, and S Nishida participated in the transplant procedures; and all investigators made revisions to the manuscript.
Conflict of interest statement None declared.
Acknowledgments We thank Kelly M Moore for supplying the illustration in figure 1. There was no funding source for this investigation.
References 1
Nishida S, Levi D, Kato T, et al. Ninety-five cases of intestinal transplantation at the University of Miami. J Gastrointest Surg 2002; 6: 233–39. 2 Kato T, Ruiz P, Thompson JF, et al. Intestinal and multivisceral transplantation. World J Surg 2002; 26: 226–37. 3 Alexandrides IJ, Liu P, Marshall DM, et al. Abdominal wall closure after intestinal transplantation. Plast Reconstr Surg 2000; 106: 805–12. 4 Starzl TE, Miller C, Broznick B, et al. An improved technique for multiple organ harvesting. Surg Gynecol Obstet 1987; 165: 343–48. 5 Abu-Elmagd K, Fung J, Bueno J, et al. Logistics and technique for procurement of intestinal, pancreatic, and hepatic grafts from the same donor. Ann Surg 2000; 232: 680–87. 6 Tzakis AG, Kato T, Nishida, et al. Campath-1H in intestinal and multivisceral transplantation: preliminary data. Transplant Proc 2002; 34: 937. 7 Nishida S, Levi D, Kato, et al. Induction therapy for adult small bowel transplant with Campath-1H. Transplant Proc 2002; 34: 1889. 8 Jones JW, Gruber SA, Barker JH, et al. Successful hand transplantation. One-year follow-up. Louisville Hand Transplant Team. N Engl J Med 2000; 343: 468–73. 9 Hettiaratchy S, Butler PEM, Lee WPA. Lessons from hand transplantations. Lancet 2001; 357: 494–95. 10 Fontes P, Rao AS, Demetris AJ, et al. Bone marrow augmentation of donor-cell chimerism in kidney, liver, heart, and pancreas islet transplantation. Lancet 1994; 344: 151–55.
THE LANCET • Vol 361 • June 28, 2003 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet.
Articles
Transplantation of the abdominal wall
David M Levi, Andreas G Tzakis, Tomoaki Kato, Juan Madariaga, Naveen K Mittal, Jose Nery, Seigo Nishida, Phillip Ruiz
Summary Background Closure of the abdomen in patients undergoing intestinal transplantation can be extremely difficult, if not impossible. We describe our initial experience with abdominal wall allotransplantation to facilitate abdominal closure. Methods We undertook nine cadaveric abdominal wall composite allograft transplants in eight patients. The graft’s blood supply was based on the inferior epigastric vessels left in continuity with the donor femoral and iliac vessels. Skin biopsies were undertaken randomly and when rejection was suspected. Vessel patency was monitored by doppler ultrasound. Findings Six patients have survived, five of whom have intact, viable abdominal wall grafts. Two patients have had a clinically mild episode of acute rejection of the skin of the abdominal wall that resolved with corticosteroid therapy. No clinically apparent graft-versus-host disease has been noted. Interpretation Transplantation of an abdominal wall composite allograft can facilitate reconstruction and closure of the abdominal compartment in intestinal transplant recipients with complex abdominal wall defects. Lancet 2003; 361: 2173–76
Introduction Intestinal failure—the permanent loss of sufficient functional intestinal absorptive surface area to sustain life—remains a catastrophic disease entity. Although total parenteral nutrition is essential in the management of affected patients, such treatment can be associated with serious complications, especially when used long term. Complicated intestinal failure is characterised by cholestatic liver disease induced by total parenteral nutrition, a loss of all central venous access sites for total parenteral nutrition, or both. Intestinal transplantation is reserved for patients with complicated intestinal failure. Recent advancements in immunosuppression, graft surveillance, and surgical technique have resulted in improved survival for these patients.1,2 We have previously reported a technical description of intestinal and multivisceral transplant procedures.2 One great technical problem is closure of the abdomen at the end of transplantation. Since 1994, more than 160 intestinal failure patients have received transplants at our institution. Many of these patients have undergone complete resection of the midgut and have lost the domain of the abdominal compartment. Additionally, some had severely damaged abdominal walls from injuries, repeated laparotomies, ostomy construction, wound infections, tumours, and enterocutaneous fistulae. The use of plastic surgical techniques to recruit tissue for abdominal closure can be helpful,3 but in about 20% of patients in our series adequate tissue for closure does not exist. Failure to close the abdomen at the time of transplant leaves the recipient with a large, open wound and, at best, a staged closure needing several surgical procedures and a long period of rehabilitation. Complications can be life threatening, and include intraabdominal infection, injury to unprotected organs, bleeding, and fistula formation. We propose that transplantation of a cadaveric, composite allograft consisting of the abdominal wall could be an effective way to address this problem. We describe our initial experience with transplantation of the abdominal wall to facilitate reconstruction of the abdominal compartment during intestinal transplantation.
Methods
Division of Transplantation, Department of Surgery (D M Levi MD, A G Tzakis MD, T Kato MD, J Madariaga MD, N K Mittal MD, J Nery MD, S Nishida MD), and Department of Pathology (P Ruiz MD), University of Miami/Jackson Memorial Medical Center, Miami, Florida, USA Correspondence to: Dr David M Levi, University of Miami School of Medicine, Daughtry Family Department of Surgery, PO Box 015809 (M840), Miami, FL 33101 USA (e-mail: [email protected])
The abdominal wall composite graft is a full thickness, vascularised, myocutaneous free flap. It consists of one or both rectus abdominus muscles, with the investing fascia, the overlying subcutaneous tissue, and skin. The blood supply is derived from the donor inferior epigastric vessels, left in continuity with the larger femoral and iliac vessels (figure 1). Intestinal graft recipients received either an isolated small bowel graft or a multivisceral graft. The latter includes the donor stomach, duodenum, liver, pancreas, and small intestine. The decision to include an abdominal wall composite graft in patients undergoing intestinal transplantation was made if conventional abdominal
THE LANCET • Vol 361 • June 28, 2003 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet.
2173
ARTICLES
Age (years) Sex
Diagnosis
53 1 2 21 41 6 15 19
Midgut strangulation MV Gastroschisis MV Hirschsprung’s disease MV Blunt trauma ISB Blunt trauma ISB Midgut volvulus MV Blunt trauma ISB Penetrating trauma ISB
Female Female Male Male Male Female Male Male
Transplant Patients’ status type (months of follow-up) Alive (23) Died Alive* Alive (13) Died Alive (9) Alive (3) Alive (1)
MV=multivisceral. ISB=isolated small bowel. *Without abdominal wall graft.
Patients’ characteristics
Figure 1: Implantation of abdominal wall composite graft in the recipient
closure was not possible. All transplants were done with informed consent. Intestinal transplantation has been undertaken at our institution in accord with protocols approved by the investigational review board since 1994. As for many of the technical alterations to these procedures that have been developed, a separate protocol for the inclusion of the abdominal wall graft has not been submitted to the investigational review board. Induction immunosuppression consisted of the anti-CD52 monoclonal antibody alemtuzumab, which was given intravenously (0·3 mg/kg) immediately preoperatively, immediately postoperatively, and on postoperative days 3 and 7. Maintenance immunosuppression was based on tacrolimus without corticosteroids. The target trough concentration of tacrolimus in serum was 10 g/L. One child received a transplant before we started using alemtuzumab in paediatric patients; instead, they received daclizumab for induction with tacrolimus and methylprednisolone for maintenance immunosuppression. Intravenous corticosteroids were used for the treatment of rejection. Procurement of the graft was done as part of the cadaveric, heart-beating donor, multiorgan procurement procedure.4,5 The procedure began with a median sternotomy and a bisubcostal incision. Longitudinal incisions were made following both lateral edges of the rectus sheath. These incisions were continued into the groins bilaterally. The common femoral vessels were identified. Finally, a transverse, suprapubic incision was made, connecting the two longitudinal incisions. The assessment and dissection of the rest of the abdominal and thoracic organs proceeded as previously described.4 The abdominal wall graft was packed with ice in situ while the other organs were flushed and removed. When the other organs had been procured, the distal aorta was cannulated and the graft was flushed with cold preservation solution. The graft was removed in one piece with the femoral and iliac vessels, with a short segment of distal aorta and inferior vena cava. Closure of the donor’s abdomen was facilitated by mobilising skin and subcutaneous tissue flaps from the lateral abdomen and flanks. After reperfusion of the isolated intestinal or multivisceral graft in the recipient, we transplanted the abdominal wall graft as a separate organ. The inclusion of an abdominal wall graft added about 2 h to the procedure’s operative time. Much like a kidney allograft,
2174
the vessels of the abdominal wall graft were implanted into the recipient’s common iliac artery and vein. Alternatively, the infrarenal aorta and inferior vena cava were used in one child. In two children who received multivisceral grafts, the arterial blood supply to the abdominal wall graft was from the donor distal aorta and the venous drainage was via the donor infrahepatic vena cava. The graft was incorporated into the recipient’s abdominal wall during closure of the abdomen. The graft, with its long vascular pedicle, was rotated and positioned according to location of the abdominal wall defect. The skin of the abdominal wall graft was left intact and was given no special treatment; normal skin colour indicated adequate perfusion. The flow through the inferior epigastric vessels of the graft was monitored with a handheld doppler ultrasound device. We undertook biopsies of the skin of the graft at randomly and when rejection was suspected on clinical grounds. The biopsy tissue was placed in a 10% buffered formalin solution, processed, and paraffin-embedded. 5 m sections were stained with haematoxylin and eosin. We recorded the patients’ demographics and diagnosis, type of transplant, graft and patient survival, rejection episodes, and reason for graft loss.
Results In eight patients (four adults and four children) we used an abdominal wall composite graft to facilitate coverage of the isolated intestinal or multivisceral graft and closure of the abdomen. All patients had small abdominal compartments from previous surgical resections and severely damaged abdominal walls from previous incisions, wound infections, fistulae, and stomata. Patients’ characteristics are shown in the table. All transplants were undertaken without HLA matching. Seven recipients received ABO identical grafts; one child (blood type A) received an ABO
Figure 2: Abdominal wall composite graft implanted during intestinal transplant procedure
THE LANCET • Vol 361 • June 28, 2003 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet.
ARTICLES
Figure 3: Photomicrographs of a skin biopsy specimen showing acute rejection Haematoxylin and eosin stain. (A) Low-power (⫻100) magnification of skin with perivascular chronic inflammatory cell infiltrates in the superficial dermis and some lymphocytic exocytosis. (B) Magnification ⫻200. Perivascular infiltrate (solid dark arrow) with reactive lymphoid cells, erythrocyte extravasation (dotted arrow), and exocytosis (white arrow).
identical multivisceral graft followed 6 days later with a type O abdominal wall graft from a second donor. The abdominal wall from the first donor was not available in this case because consent for its procurement was not obtained from the donor’s family. One other patient received an ABO identical abdominal wall graft 3 days after the initial intestinal transplant procedure. This time interval enabled us to find a larger donor with a greater surface area abdominal wall than the initial donor. With the exception of these two recipients, the abdominal wall grafts originated from the same donor as the intestinal graft and were transplanted concurrently with the intestinal graft. Figure 2 depicts an abdominal wall graft at the conclusion of the transplant procedure. The size of the abdominal wall graft depended on the size of the donor and the width of the rectus muscles. The elasticity of the graft, especially the skin, and the tendency of the graft to shrink in cold, hypertonic preservation solution, made accurate measurement of the graft’s surface area difficult. The surface area of coverage provided by the abdominal wall graft ranged from 150 cm2 to 500 cm2. Six of the eight patients are alive; five have functioning, viable abdominal wall composites grafts. These five patients have been followed up for 1, 3, 9, 13, and 23 months post-transplant. Wound healing was uneventful and proceeded at a rate similar to that for the recipient’s native skin. The intestinal grafts have been protected expediting patient recovery. In one patient, a 2-year-old child, we removed the abdominal wall graft because of infarction. This infarct was probably related to hypovolaemia and hypotension, although a mechanical problem cannot be excluded with certainty. The child eventually recovered after the abdominal wall defect closed by secondary intention. Another child died after sepsis and multiorgan failure with an intact abdominal wall graft. One adult patient died in the early postoperative period after receiving two isolated intestinal and abdominal wall grafts. The first graft infarcted because of hypoperfusion from intraabdominal haemorrhage. The patient eventually died from sepsis after retransplantation. Seven of the eight patients received alemtuzumab and tacrolimus maintenance according to an established protocol.6 The other, a child, received daclizumab as an induction agent instead of alemtuzumab. The five
surviving patients with abdominal wall grafts received alemtuzumab induction. Of these five, one adult has had two episodes of mild rejection of the intestine, proven by mucosal biopsy, without concurrent rejection of the skin of the abdominal wall graft. One recipient of an isolated small bowel graft had moderate rejection of the intestine without involvement of the abdominal wall graft. Additionally, two patients have had an episode of rejection of the abdominal wall graft without involvement of the intestinal graft. This manifested clinically as erythema of the skin of the composite graft. No other signs or symptoms were noted. One patient had homogeneous graft skin involvement and the other had a patchy, macular rash. Both cases arose about 1 month after the transplant procedure. The histopathological alterations, shown in figure 3, included focal spongiosis with lymphocytic exocytosis and focal epidermal necrosis, a substantial perivascular lymphocytic infiltrate with eosinophils and neutrophils, and venulitis. The lymphocytes were enlarged and reactive. These changes were consistent with acute rejection. The patients were treated with intravenous corticosteroids, and complete resolution took about 10 days and was confirmed by biopsy. Notably, one of these patients had received an ABO non-identical abdominal wall graft 6 days after the multivisceral transplant procedure. The other also received an abdominal wall graft several days after the intestinal transplant but from a different ABO identical donor. No patient in this series had signs or symptoms of graftversus-host disease, and none died as a results of complications of the abdominal wall graft.
Discussion We have shown that transplantation of the abdominal wall is feasible and safe; and for five patients this procedure has proven instrumental in the success of their transplant. It allowed for primary closure of the abdomen, avoided the potential morbidity of exposed viscera, and permitted early mobilisation and rehabilitation of these patients. The procurement of the abdominal wall graft does not interfere with the procurement of other organs and tissues. If complications arise, the graft can be removed fairly easily. The abdominal wall graft is procured after the other organs have been removed from the cadaver. Consequently, the quality of those organs is not
THE LANCET • Vol 361 • June 28, 2003 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet.
2175
ARTICLES
jeopardised. Since the iliac vessels are included with the abdominal wall graft, the carotid and subclavian vessels can serve as excellent vascular grafts for the other organs to be transplanted as needed. Graft implantation is straightforward and the large diameter of the anastamosed area diminishes the risk of technical failure. Careful attention needs to be directed to the three-dimensional orientation of the vascular pedicle of the abdominal wall graft, to avoid mechanical occlusion. Transplantation of the abdominal wall composite graft can take place during the intestinal transplant procedure, or several days later with a graft from a different donor. Delaying implantation of the abdominal wall graft allows perioperative oedema to diminish before abdominal closure and the patient’s condition to stabilise. This strategy may be preferable when the recipient has an especially large defect in the abdominal wall. Ideally, a small intestinal graft would be transplanted, and oedema would decrease over several days, followed by the implantation of an abdominal wall graft obtained from a larger donor. In our experience, HLA matching is of little clinical significance in intestinal transplantation. Therefore, we did not consider HLA matching in this group of patients. Matching the donor and recipient size and ABO compatibility are of much greater importance. Rejection can be effectively prevented and treated with modern immunosuppression. Morevoer, because the skin of the abdominal wall graft is accessible for visual inspection and biopsy, monitoring for rejection is easy. Induction with alemtuzumab seems to be effective in combination with low-dose tacrolimus maintenance. With this corticosteroid-sparing regimen, number and severity of rejection episodes have been reduced in intestinal transplant patients; many such patients have never had rejection.6,7 Although not yet observed, we speculate that chronic rejection of the abdominal wall graft might not be a serious event, because a fibrotic graft may still provide satisfactory coverage of the abdominal organs. For five patients, transplantation of the abdominal wall proved instrumental in their recovery. For the survivor in whom the abdominal wall graft thrombosed, we were able to remove the graft without compromising the intestinal graft and revert to conventional techniques for abdominal closure. Removal of the graft did, however, lead to a delay in feeding and mobilisation of the patient, and contributed to a hospital stay of over 2 months. Two deaths in this small series were not related to the inclusion of the abdominal wall graft. We have described initial experiences with a novel technique. It is difficult to make a direct comparison between this small group and other intestinal and multivisceral transplant recipients. In our experience, which encompasses some 160 intestinal and multivisceral transplants, some patients with intestinal failure have devastating tissue loss that makes primary closure after transplant impossible. We propose that the addition of an abdominal wall composite graft is a feasible solution to this difficult problem. Human composite allografts have been previously described in the form of transplantation of the hand. To date, 14 cadaveric hand transplants have been undertaken in 11 patients, with encouraging early results.8,9 By contrast with the hand transplant recipients, who are
2176
generally in good health, patients with intestinal failure suffer from a life-threatening disease and need much treatment in hospital. Furthermore, because they are already receiving an intestinal graft—and, hence, therapeutic immunosuppression—the inclusion of an abdominal wall graft carries no additional risk of complications related to immunosuppression. Unlike the abdominal wall graft, hand composite grafts include an obligatory, small amount of donor bone-marrow cells, which may assert an immune modulatory effect.10 Finally, the function of the abdominal wall graft is different to that of the hand graft; the purpose of the abdominal wall graft is to provide biological coverage of the newly transplanted abdominal organs. Recent improvements in immunosuppression, graft surveillance, and surgical technique have made intestinal transplantation a potentially life-saving option for patients with complicated intestinal failure. In intestinal failure patients with irreparable abdominal wall defects, the idea of transplanting the abdominal wall from a cadaveric donor is a logical step. Future applications might include transplantation of the abdominal wall composite graft as an isolated organ transplant in patients with massive abdominal wall defects after trauma or complicated surgery. Contributors D Levi obtained the data, and wrote and revised the manuscript; A Tzakis developed the surgical procedure; T Kato, N Mittal, and S Nishida directed patient management and graft surveillance; P Ruiz provided the pathological interpretation of the skin biopsy specimens; D Levi, A Tzakis, T Kato, J Madariaga, J Nery, and S Nishida participated in the transplant procedures; and all investigators made revisions to the manuscript.
Conflict of interest statement None declared.
Acknowledgments We thank Kelly M Moore for supplying the illustration in figure 1. There was no funding source for this investigation.
References 1
Nishida S, Levi D, Kato T, et al. Ninety-five cases of intestinal transplantation at the University of Miami. J Gastrointest Surg 2002; 6: 233–39. 2 Kato T, Ruiz P, Thompson JF, et al. Intestinal and multivisceral transplantation. World J Surg 2002; 26: 226–37. 3 Alexandrides IJ, Liu P, Marshall DM, et al. Abdominal wall closure after intestinal transplantation. Plast Reconstr Surg 2000; 106: 805–12. 4 Starzl TE, Miller C, Broznick B, et al. An improved technique for multiple organ harvesting. Surg Gynecol Obstet 1987; 165: 343–48. 5 Abu-Elmagd K, Fung J, Bueno J, et al. Logistics and technique for procurement of intestinal, pancreatic, and hepatic grafts from the same donor. Ann Surg 2000; 232: 680–87. 6 Tzakis AG, Kato T, Nishida, et al. Campath-1H in intestinal and multivisceral transplantation: preliminary data. Transplant Proc 2002; 34: 937. 7 Nishida S, Levi D, Kato, et al. Induction therapy for adult small bowel transplant with Campath-1H. Transplant Proc 2002; 34: 1889. 8 Jones JW, Gruber SA, Barker JH, et al. Successful hand transplantation. One-year follow-up. Louisville Hand Transplant Team. N Engl J Med 2000; 343: 468–73. 9 Hettiaratchy S, Butler PEM, Lee WPA. Lessons from hand transplantations. Lancet 2001; 357: 494–95. 10 Fontes P, Rao AS, Demetris AJ, et al. Bone marrow augmentation of donor-cell chimerism in kidney, liver, heart, and pancreas islet transplantation. Lancet 1994; 344: 151–55.
THE LANCET • Vol 361 • June 28, 2003 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet.