SEARCH

SEARCH BY CITATION

Keywords:

  • Abdominal wall transplantation;
  • intestinal transplantation;
  • rejection;
  • vascularized composite allografts

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References

We report our outcomes following combined intestinal and abdominal wall transplantation, focusing on the presentation and treatment of acute rejection of the abdominal wall vascularized composite allograft (VCA). Retrospective analysis of all patients with combined intestinal/VCA transplantation was undertaken. Graft abnormalities were documented photographically and biopsies taken, with histological classification of rejection according to Banff 2007 guidelines. We have performed five combined intestinal and abdominal wall transplants to date. Two patients developed erythematous, maculopapular to papular eruptions confined to the VCA, histologically confirmed as grade II/III rejection, yet with normal bowel on endoscopy. Both patients' rashes resolved within 72 h of increasing immunosuppressive treatment. One patient later developed a recurrence of the rash, confirmed as skin rejection, but did not immediately seek medical attention. Treatment was therefore delayed, and mild intestinal rejection developed. We describe the rash associated with VCA rejection, and propose that while the skin of an abdominal wall VCA may reject independently of the intestinal allograft, delay in treatment of rejection episodes may result in rejection of the intestinal graft.


Abbreviations
CT

computed tomography

G-CSF

granulocyte colony stimulating factor

GvHD

graft–versus–host disease

HPN

home parenteral nutrition

LIMA

left internal mammary artery

PUO

pyrexia of unknown origin

TPN

total parenteral nutrition

VCA

vascularized composite allograft.

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References

Closure of the abdominal wall at the time of intestinal transplantation has historically presented a significant clinical conundrum. Low quality recipient abdominal wall, which is often secondary to multiple operations, scarring of the skin and the presence of multiple enterocutaneous fistulae, combined with a loss of abdominal domain and discrepancy in size between donor and recipient, often leads to insufficient or inadequate tissue to cover the edematous transplanted organs [1]. The advent of abdominal wall transplantation early in the last decade offered a potential solution to this complex issue [2-4]. After upper limb transplantation, combined intestinal and abdominal wall transplantation is now the second most common vascularized composite allograft (VCA) performed [5].

Despite increasing usage, there is little in the literature describing the outcomes from abdominal wall transplantation, particularly the presenting signs and symptoms of rejection. There is also little evidence correlating rejection of the abdominal wall with function of the underlying intestinal graft. Experience from upper limb transplantation suggests that rejection of the skin presents with a maculopapular rash and that this occurs while underlying components remain healthy. The presence of skin within VCAs presents a unique immunological challenge as it has long been known that skin is more susceptible to rejection than any other tissue [6]. Indeed, skin is virtually always the first and usually the only tissue to reject in VCA transplants [5]. Recent studies have suggested that this property may facilitate the use of skin as a herald to impending rejection in other transplanted viscera [7].

We report our experience of abdominal wall allograft rejection after combined intestinal and abdominal wall transplantation with regard to diagnosis, management and outcome.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References

Patients were considered for abdominal wall transplantation in addition to intestinal transplantation if they demonstrated loss of abdominal domain on CT volumetric assessment, along with clinical evidence of a scaphoid abdomen and poor quality abdominal skin with or without enterocutaneous fistulae. Demographic, etiological, surgical and immunosuppressive details were analyzed in the context of any significant postoperative events.

The abdominal wall allograft was obtained from the same donor as the bowel and perfused on vessels from the recipient's nondominant forearm while the intestinal graft was implanted intra-abdominally in order to reduce cold ischaemia time. In one patient (patient 3) a portion of colon was transplanted along with the small bowel, while all others received small bowel alone.

Directly following implantation of the bowel, three patients had the abdominal wall allograft transferred from the forearm to the inferior epigastric or deep circumflex iliac vessels. In the remaining two, this was not possible due to a clinical need to prevent a prolonged anesthetic. However, in both these cases the abdominal allograft was inset onto the recipient abdomen. The final transfer of the abdominal wall from the forearm vessels to the recipient abdominal vessels was planned as a secondary delayed procedure. At 30 days postoperatively, one of these patients had the abdominal wall graft detached from the ulnar artery and re-vascularized from the left internal mammary artery (LIMA) via a saphenous vein graft. Venous drainage was not required as the abdominal wall had developed sufficient venous drainage through the flap delay phenomenon.

Induction immunosuppression was with alemtuzumab (CAMPATH-1H) 6 h postreperfusion of the bowel and again 24 h later. Tacrolimus (therapeutic range 8–12 ng/mL) and prednisolone were used for maintenance immunosuppression.

Bowel function was assessed using validated biochemical markers including citrulline levels [8, 9] and endoscopic assessment. Endoscopic assessment was based on that described by Kato et al. [10] and consisted of weekly zoom endoscopy for the first month posttransplantation, followed by fortnightly endoscopy for the following 2 months. Abdominal wall skin abnormalities were documented photographically and 4 mm punch biopsies were taken. Rejection was classified histologically according to the Banff 2007 classification [11].

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References

From January 2012 to January 2013, five patients underwent combined intestinal and abdominal wall transplantation. Their details are shown in Table 1. The mean cold ischaemia time for the abdominal wall transplant was 345 min (range 320–360 min), primary warm ischaemia time was 48 min (range 45–55) and the secondary warm ischaemia time resulting from transfer of the wall from the ulnar vessels to the abdominal vessels was 75 min (range 62–86 min). Mean cold ischaemia time for the intestinal grafts was 374 min (range 315–414 min).

Table 1. Clinical details of patients receiving an abdominal wall transplant between January 2012 and January 2013
PatientAgeGenderEtiology of intestinal failureReason for abdominal wall transplantationHLA mismatch (A-B-DR)Abdominal wall perfusionStatus
168MRadiation enteritisFriability of abdominal wall1-1-1Ulnar arteryDeceased
240MMesenteric infarctionVentral defect in the abdominal wall2-2-1Inferior epigastricAlive
333MNECRectus atrophy, scarring and loss of domain from multiple transverse incisions2-2-1Ulnar arteryAlive
LIMA
427FCrohn'sMultiple enterocutaneous fistulae2-2-1Inferior epigastric/deep circumflex iliacAlive
559MCrohn'sMultiple enterocutaneous fistulae1-2-1Inferior epigastricAlive

Posttransplantation, all five patients developed primary graft function with stoma outputs of less than 20 mL/kg/24 h, macroscopically and histologically healthy bowel on routine surveillance endoscopy, and citrulline levels of >13 μmol/mL.

Sixty-two days postoperatively, the patient that had not had his abdominal wall allograft transferred from the forearm to the abdomen died from intra-abdominal sepsis secondary to a bowel anastomotic leak. All four surviving patients have attained independence from total parenteral nutrition.

Patient 2 developed a culture-negative pyrexia of unknown origin (PUO) approximately 21 days postoperatively, and after detection of 4% donor-derived chimerism in the peripheral blood was treated with a single intravenous dose of methylprednisolone for presumed graft–versus–host disease (GvHD). Sixteen days postoperatively, patient 3 also developed a PUO and was readmitted to intensive care and treated for suspected GvHD. The remaining inpatient stay for these patients was uneventful, with both being discharged on full enteral diet and with normally functioning bowel.

Fifty-five days postoperatively, patient 3 developed a pale, erythematous, maculopapular/papular eruption confined to, and covering 75% of, the abdominal wall allograft together with a neutropenia of 1.76 × 109/L. Some papules were follicular-based, and there was enlargement of macular lesions at the periphery of the graft to form coalescing patches of erythema (Figure 1). Stoma output remained at approximately 20 mL/kg/24 h and endoscopy revealed a macroscopically normal bowel. Biopsies were not possible due to poor bowel preparation. Abdominal wall biopsies revealed Grade III acute rejection.

image

Figure 1. Cutaneous manifestations of rejection in the abdominal wall vascularized composite allograft. A pale erythematous, maculopapular to papular eruption confined to the graft. Both follicular and nonfollicular based papules are seen, with macular lesions coalescing at the peripheries to from patches of erythema.

Download figure to PowerPoint

At day 67 postoperatively, patient 2 developed an absolute neutropenia (neutrophils < 1 × 109/L). This was treated with Granulocyte Colony Stimulating Factor (G-CSF) and broad-spectrum intravenous antibiotics, after which he developed a pale erythematous macropapular, follicular, and nonfollicular eruption confined to the transplanted cadaveric abdominal wall, similar to that of patient 3. Biopsies revealed Banff grades II–III rejection (Figure 2). There were no systemic signs of disease, and bowel function remained stable. Endoscopic assessment revealed healthy mucosa, and ileal biopsies displayed no evidence of rejection. Citrulline levels were 26.8 μmol/L.

image

Figure 2. Histological appearances of rejection in the skin of the abdominal wall vascularized composite allograft. Abdominal wall biopsies showing an interface reaction with apoptotic keratinocytes limited to the basal layer. There is mild acanthosis and focal parakeratosis. The dermis shows an inflammatory lymphohystiocytic infiltrate including occasional eosinophils. There are atypical hysticytoid cells and several mitoses in an area of florid interface change.

Download figure to PowerPoint

Both episodes of rejection were managed with a bolus of Basiliximab, intravenous methylprednisolone pulsed over 3 days and topical tacrolimus ointment, as well as the addition of azathioprine to the maintenance regimen. Both patients demonstrated resolution of the rash within 72 h of treatment.

Eighty-six days postoperatively, patient 3 developed a pale erythematous rash. On this occasion, however, the rash had been present for 8 days prior to presentation to hospital. A skin biopsy revealed Banff grade II rejection, which was treated with IV methylprednisolone. Biopsies from the distal ascending colon revealed a mild lymphocytic colitis. Over the subsequent week there was improvement but not complete resolution of the rash, and repeat skin biopsies revealed ongoing Banff grade II rejection. At the point of skin biopsy, colonic biopsies from routine endoscopic surveillance revealed evidence of mild rejection, as evidenced by an increase in basal crypt apoptotic bodies to 8 per 10 crypts, with preservation of normal glandular architecture and no evidence of acute or chronic inflammation. Immunohistochemistry for CMV, HSV, and adenovirus were negative. Full histological reports are documented in Table 2. Citrulline levels remained between 25 μmol/L and 30 μmol/L, and there was no change in stoma output. Tacrolimus levels were noted to be persistently under 3 ng/mL, attributed to poor compliance. Intravenous tacrolimus was therefore administered to return levels to within normal range, resulting in a resolution of the skin rash.

Table 2. Detailed histological findings from patients 2 and 3, and their relation to rejection of the intestine and abdominal wall
Postoperative dayHistological findingsConclusions
Patient 3
98Large bowel mucosaMild acute intestinal rejection
 Intact surface with no erosion or ulceration 
 Glandular architecture preserved 
 No acute inflammation seen 
 No appreciable excess chronic inflammatory cells or atypical lymphoid infiltrate 
 Increase in basal crypt apoptotic bodies (“hot spots 8 per 10 crypts”) 
 No viral cytopathic effect or surface debris seen 
 Immunohistochemistry for CMV, HSV, and adenovirus negative 
94Abdominal wallBanff grade I/II rejection
 Mild superficial and deep perivascular inflammatory cell infiltrate composed predominantly of lymphocytes with some eosinophils 
 Overlying epidermis shows spongiosis with associated exocytosis of lymphocytes 
 Features suggestive, but not diagnostic, of interface change 
 Normal underlying sweat glands and no vasculitis 
89Distal ascending colonMild lymphocytic colitis, no acute rejection
 Marginal increase in intra-epithelial lymphocyte population 
 No increased apoptosis or glandular destruction 
88Abdominal wallBanff grade I/II rejection
 Very mild superficial and deep perivascular infiltrate of small lymphocytes including within the subcutaneous tissue 
 A few scattered interstitial eosinophils 
 Small foci of mild epidermal spongiosis and tiny foci raising the possibility of very mild interface change with mild basal vascular change. No convincing apoptotic bodies 
 A few scattered interstitial eosinophils 
 No evidence of vasculitis and no changes in the eccrine glands 
80Large bowel mucosaNormal bowel
 Large bowel mucosa within normal limits. No evidence of rejection or atypical infiltrate 
55Abdominal wallBanff grade III rejection
 Interface reaction with apoptotic keratinocytes limited to the basal layer 
 Mild acanthosis and focal surface parakeratosis 
 Inflammatory lymphohystiocytic infiltrate including occasional eosinophils in the dermis 
 Atypical hystiocytoid cells and several mitoses, with the latter located in an area of florid interface change. Most likely reactive changes in histiocytes 
Patient 2
68Ileal mucosaIleal mucosa within normal limits
 No architectural distortion, surface or crypt epithelial injury. 
 1 apoptotic body/10 crypts with crypt mitoses 
 No other abnormal findings 
68Abdominal wallBanff grade II rejection
 Perifollicular chronic inflammatory cell infiltrate and overlying spongiosis 
 There is a single focus of possible interface change 
28Ileal mucosaIleal mucosa within normal limits
 No architectural distortion, surface or crypt epithelial injury 
 2 apoptotic bodies/10 crypts, and occasional mitoses in the crypts 
 Occasional intra-epithelial lymphocytes 
 No other abnormal findings 

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References

There is limited experience of abdominal wall transplants and the presentation of rejection. Here, we report a peri-follicular, maculopapular/papular rash that we suggest is the hallmark of rejection of an abdominal wall graft. A similar rash in upper limb transplants has been reported to be indicative of rejection [12]. Interestingly, rejection occurred in both patients in combination with a neutropenia. While this is most likely secondary to immunosuppression, and in particular alemtuzumab induction [13], it is conceivable that a granulocyte-driven rejection episode leading to infiltration with and sequestration of neutrophils in the skin may result in neutropaenia.

Experience of cutaneous manifestations of rejection in human VCAs is modest. In animal models, however, a hierarchy of tissue susceptibility to rejection has been demonstrated, with skin being shown to be the most susceptible [6, 14]. Kanitakis et al. have proposed that skin may act as a precursor to rejection of other components of a VCA. In two of our patients, early recognition of skin rejection and prompt treatment resulted in a resolution of skin rejection with no progression to intestinal rejection. However, when there was a combination of delayed presentation and sub-therapeutic immunosuppression, evidence of early rejection of the intestinal allograft was also detected. Levi et al. [2] and Cipriani et al. [4] have presented the only other documented series of abdominal walls with isolated bowel or multivisceral transplants. In both of these studies, cutaneous evidence of rejection of the abdominal VCA has also been documented independent of intestinal involvement. The distinctive sensitivity of skin to rejection that we and others have noted clinically may therefore assist in the identification of early alloresponses, allowing prompt treatment and therefore prevention of the eventual rejection of less sensitive organs such as the intestine or pancreas [2, 4, 5, 12].

In summary, we report on a series of combined abdominal wall and intestinal transplants, demonstrating that the skin component may display the earliest and only signs of rejection.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References

The authors would like to thank the patients who took part in this study and Oxford University Hospitals NHS Trust where the transplants were performed. F.I. is a Duke of Kent Research Fellow at Restore, the Burn, and Wound Research Trust.

Disclosure

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References

The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References