SEARCH

SEARCH BY CITATION

Keywords:

  • Antidonor alloantibody;
  • antibody mediated rejection;
  • C4d;
  • pancreas transplantation;
  • humoral rejection

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Case Report
  6. Discussion
  7. References

The role of antibody-mediated rejection (AMR) in pancreas transplantation is poorly understood. Here, we report on a patient who developed AMR of his pancreas allograft after receiving a simultaneous pancreas-kidney transplant. Pre-operative enhanced cytotoxicity and flow cytometry T-cell crossmatches were negative; B-cell crossmatches were not performed as per institutional protocol. The patient's post-operative course was significant for elevated serum amylase levels and development of hyperglycemia approximately 1 month after transplantation. A pancreatic biopsy at this time showed no cellular infiltrate but strong immunofluorescent staining for C4d in the interacinar capillaries. Analysis of the patient's serum identified donor-specific HLA-DR alloantibodies. He received intravenous immunoglobulin (IVIg), rituximab and plasmapheresis, and his pancreatic function normalized. We conclude that clinically significant AMR can develop in a pancreas allograft and recommend that pancreatic biopsies be assessed for C4d deposition if the patient has risk factors for AMR and/or the pathologic evidence for cell-mediated rejection is underwhelming.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Case Report
  6. Discussion
  7. References

Although the importance of AMR in graft dysfunction has been demonstrated in kidney, cardiac and lung transplantation, its role in pancreatic allograft rejection is unknown (1). The distinction between cellular rejection and AMR is important as the treatment for each process differs. Treatment for T-cell-mediated rejection includes corticosteroids and anti-T-cell antibodies such as polyclonal anti-thymocyte globulin and anti-CD3 monoclonal antibody. Treatment for patients with AMR is aimed at depleting circulating donor-specific (DSA) antibodies with IVIg and plasmapheresis or reducing antibody production by eliminating CD20-positive plasma cell precursors with anti-CD20 monoclonal antibody (rituximab) (1).

In renal transplantation, biopsies of the transplanted kidney can help to distinguish between cell-mediated rejection and AMR. Lymphocytic infiltration of kidney tubules is suggestive of a cell-mediated process, while predominantly, neutrophilic infiltrates and peritubular capillary deposition of the complement fragment C4d are consistent with AMR (2,3). C4d deposition within 6 months of transplant has clinical significance in that it correlates with impaired graft function and reduced graft survival (3). In addition to these histological findings, the National Conference to Assess Antibody-Mediated Rejection in Solid Organ Transplantation has concluded that the diagnosis of AMR also requires serological evidence of DSA.

AMR has also been described in lung and cardiac transplantation and found to correlate with reduced graft survival (4). As in kidney transplants, detection of C4d in endomyocardial biopsies is suggestive of AMR (5,6) and portends myocardial ischemia (7) and worse patient outcomes (5). In lung transplant patients with clinical evidence of acute rejection, detection of C4d in bronchoalveolar fluid correlates with the degree of morphologic evidence of humoral rejection and the severity of clinical rejection (8).

While biopsies of pancreatic grafts are frequently performed to diagnose acute cellular rejection, we are not aware of any reports in the literature that document clinically significant AMR of pancreatic allografts. In this report, we describe the course of a patient who was found to have AMR of both his pancreas and kidney after simultaneous pancreas-kidney transplantation.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Case Report
  6. Discussion
  7. References

Histology and immunohistochemistry

The renal and pancreatic biopsies were split with a small piece frozen for immunofluorescence studies. The remainder of each biopsy was submitted in formalin for light microscopic examination. In the case of the renal biopsies, 2-μm sections were cut and stained with hematoxylin and eosin (H&E) and periodic acid-schiff (PAS). For the pancreatic biopsy, 4-μm sections were cut and stained with H&E. An indirect immunofluorescence technique was used for the detection of C4d. Frozen sections of kidney or pancreas were cut at 4 μm, placed on coated slides, and fixed for 20 min in ice-cold acetone. Following rinsing in cold phosphate-buffered saline (PBS), mouse anti-human C4d antiserum (Quidel Corporation, San Diego, CA) was applied at a dilution of 1:100 and incubated at 45 min at room temperature in a humid incubation chamber. Following rinsing in PBS, the secondary antibody, fluorescein isothiocyanate (FITC) rabbit anti-mouse IgG (Dako Corp., Carpinteria, CA) was applied at a dilution of 1:40 and incubated for 25 min as above. The slides were examined in a Zeiss Axioskop 2 plus fluorescence microscope (Carl Zeiss, Inc., Thornwood, NY). Digital images were taken with a Zeiss Axiocam HRc digital camera.

Histocompatibility testing

Pre-transplant crossmatches were performed on donor T-lymphocytes using antiglobulin-enhanced lymphocytotoxicity (AHG-CDC) and flow cytometric techniques. Panel reactive antibody (PRA) was determined using microparticles coated with human leukocyte antigen (HLA) molecules (LABScreen™ PRA Class I and LABScreen™ PRA Class II, from One Lambda, Inc., Canoga Park, CA). The specificities of the HLA IgG antibodies were determined using LABScreen™ Single Antigen Class I and FlowPRA Single Antigen HLA Class II (One Lambda, Inc., Canoga Park, CA). To determine the titer of donor-specific HLA alloantibodies, serial dilutions of serum were prepared, and the titer was determined by evaluating only the antibodies bound to microparticles containing donor-specific antigens.

Case Report

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Case Report
  6. Discussion
  7. References

The patient is a 44-year-old man who has suffered from poorly controlled type 1 diabetes mellitus since 2 years of age. Complications from his diabetes include diabetic retinopathy and end-stage diabetic nephropathy requiring hemodialysis. Twelve years ago, the patient received a deceased donor kidney transplant. The organ donor had four HLA-A, B, DRB1 mismatches (Table 1). This organ was eventually lost to chronic allograft nephropathy, and the patient resumed hemodialysis 7 years ago. During the last 18 months before his kidney-pancreas transplant, his Class I HLA PRA ranged from 56% to 95% and was 73% just prior to his pancreas-kidney transplant.

Table 1.  HLA types of the recipient and his first and second organ donors
 Class I HLAClass II HLA
ABCwDRDQ
  1. DR13 (donor 2) is a subtype of DR6 (donor 1). Antibody specificities of serum collected from the recipient just before the 2004 transplant included DR3 and 13; these HLA types were present in the second donor. Standard serologic and DNA techniques were used for determining HLA types. Bw4 and B46 are public epitopes encoded by the HLA-B molecules.

Recipient3, 3314, 44 (Bw4, Bw6)2, –1, 45, 8
Donor 2 (2004)1, 118, 41 (Bw6)7, –3, 132, 3
Donor 1 (1992)2, 2344, 58 (Bw4)5, –1, 61, –

In 2004, a compatible deceased donor was identified using AHG-CDC and flow cytometric crossmatches using T lymphocytes from the donor, and the patient received a simultaneous kidney-pancreas transplant. This donor had six A, B, DRB1 mismatches with the recipient (Table 1). As per our institutional protocol, no pre-operative B-cell crossmatch was performed. Of note, DRB1*13, which was one of the mismatches in the second donor, is a common subtype of DR6, which was a mismatch in the first donor.

The pancreas was implanted into the right iliac fossa, and the kidney was implanted into the left iliac fossa. The pancreatic arterial supply was reconstructed in standard fashion using a donor iliac artery interposition graft; the venous drainage was systemic via the common iliac vein, and the exocrine secretions were drained enterically. There were no significant intra-operative complications. Polyclonal rabbit anti-human thymocyte globulin (Thymoglobulin, Genzyme Transplant, Cambridge, MA) induction therapy was initiated in the operating room and continued post-operatively to a total dose of 9 mg/kg over 10 days. In addition, the patient received corticosteroids and mycophenolate mofetil according to our standard protocol (9).

Post-operatively, the patient's serum glucose was well controlled throughout his hospital stay without administration of exogenous insulin; however, the patient experienced significant delayed renal graft function and required multiple sessions of hemodialysis between the 2nd and 14th post-operative days (Figure 1). A surveillance renal biopsy performed on the 4th day after surgery showed no evidence of acute cellular or humoral rejection. This biopsy did not stain for C4d. However, a renal biopsy performed on the 10th day after surgery showed patchy tubular dilatation with tubular epithelial cell attenuation. Focal collections of interstitial inflammation were noted including macrophages, lymphocytes and neutrophils within intertubular capillaries in areas with mild interstitial edema (Figure 2).

image

Figure 1. Graph representing the patient's post-operative clinical course, diagnostic studies and therapeutic interventions. The serum levels of glucose, amylase and creatinine are plotted against the post-operative day. The times of diagnostic biopsies and whether the biopsy was C4d+ are indicated by large arrows. The times of therapeutic interventions including intravenous immunoglobulin (IVIg), plasmapheresis (PP) and rituximab (Ritux) are also labeled (small arrows). The inset table lists the titer of donor specific Class II alloantibody at various times before and after transplantation.

Download figure to PowerPoint

image

Figure 2. Light micrograph of transplant kidney showing neutrophils (arrow) within an intertubular capillary. Patchy interstitial edema is seen to the lower right (H&E, ×400).

Download figure to PowerPoint

The C4d staining of the second renal biopsy showed the typical pattern of intertubular capillaries characteristic of acute AMR (Figure 3). Because of these findings, the patient's pre-operative serum was tested more extensively and found to have antibodies against mismatched donor antigens (HLA-DR3 and DR13). The Class II HLA PRA was 100% PRA and single antigen reagents showed strong reactivity with DR13 as well as reactivity with several additional DRB1 specificities (14, 3, 11, 12, 8, 15, 16 and 0103). No antibodies were detected against DRB3. These data are consistent with recognition of a shared epitopes on these DRB1 gene products but not those from the associated DRB3 locus. These data provide strong evidence for HLA alloantibodies against both of the second donor's DRB1 molecules (i.e. DRB1*03 and DRB1*13). The titer for the donor-specific antibodies (i.e. DRB1*03 and DRB1*13) was 1:64 prior to transplant and had increased to 1:1024 by 10 days after the transplant.

image

Figure 3. Immunofluorescence micrograph of transplant kidney showing bright staining of intertubular capillaries. Glomerulus with mesangial staining in the lower right corner serves as a positive control (C4d antiserum, ×250).

Download figure to PowerPoint

Thymoglobulin was discontinued at this point and the patient began treatment with muronomab-CD3 (OKT3, Ortho Biotech, Bridgewater, NJ), IVIg (Gamunex, Talecris Biotherapeutics, Research Triangle Park, NC) and rituximab (Rituxan, Genentech, South San Francisco, CA) (Figure 1). He received three doses of 0.5 g/kg of IVIg, given every other day, and a single dose of 570 mg/m2 of rituximab. The patient's renal function remained poor, and a follow-up biopsy performed on post-operative day 17 revealed persistent severe AMR. The patient received two more doses of IVIg and underwent three sessions of plasmapheresis. Although his urine output improved somewhat after this therapy, his serum creatinine remained markedly elevated (8.9 mg/dL), and he was discharged home on the 22nd day after surgery continuing to require intermittent hemodialysis. In addition, the patient's serum amylase was elevated at the time of his discharge, but this was not pursued further since the levels were relatively stable, his serum lipase was normal and his blood glucose control was good. Immunosuppression at the time of discharge consisted of a prednisone taper, mycophenolate mofetil and tacrolimus.

On post-operative day 25, the patient was readmitted to another hospital with a gastrointestinal hemorrhage from a gastric ulcer. He underwent upper endoscopy that successfully controlled the bleeding with epinephrine injections of the bleeding sites. Despite requiring a blood transfusion, he never developed significant hypotension, and his amylase and creatinine continued their downward trend (Figure 1). When he was transferred to our facility several days later, his serum creatinine had dropped to 3.5 mg/dL, and he was no longer on hemodialysis. His serum amylase, which had dropped to a level of approximately 200 U/L at the time of transfer, began to rise on post-operative day 36 to a peak of 616 U/L (normal: 23–134) by post-operative day 43. This rise was accompanied by elevations in his fasting blood glucose from 116 mg/dL on admission to a peak of 227 mg/dL approximately 10 days later (Figure 1). A pancreas biopsy was performed on post-operative day 39 to evaluate these abnormalities. There was only a scant cellular infiltrate present in the specimen and no venous endotheliitis, septal inflammation, necrosis, pyknotic nuclei or eosinophils were noted on standard H&E staining (Figure 4A). Because of the patient's complex history, the specimen was also processed for immunofluorescent staining with C4d antibody. This assay showed intense staining of the interacinar capillaries but no staining of the pancreatic acini and interstitium, findings consistent with acute AMR (Figure 4B). Another specimen from a patient with evidence of cellular rejection on H&E staining (Figure 4C), showed mild background staining but no capillary staining with C4d antibody (Figure 4D). Evaluation of the patient's serum using LABScreen™ PRA Class II demonstrated persistent donor-specific HLA alloantibodies that had decreased from a titer of 1:1024 on post-operative day 10 to 1:256 on post-operative day 40 (Figure 1).

image

Figure 4. Micrographs of pancreas transplant biopsies. (A) Light micrograph of the transplant pancreas showing normal appearing pancreatic parenchyma without inflammation (H&E, ×250). (B) Immunofluorescence micrograph of the transplant pancreas showing bright staining of interacinar capillaries (C4d antiserum, ×250). (C) Light micrograph of a pancreas transplant from a control patient with mild acute cellular rejection. Note the presence of neutrophils in the duct. Focal interstitial edema is also present at the right edge of the micrograph (H&E, ×400). (D) Immunofluorescence micrograph of a transplant pancreas from same control patient as in (C). Note scattered cells with background staining but no staining of interacinar capillaries (C4 antiserum, ×250).

Download figure to PowerPoint

The patient was treated with two courses of plasmapheresis, three doses of 0.5 g/kg IVIg administered over 5 days and 375 mg/m2 of rituximab. His fasting serum glucose improved rapidly after institution of this therapy and has remained in the normal range since then. His serum amylase at the time of discharge was 483 U/L and decreased to 175 U/L 3 weeks later. His creatinine continued to normalize after discharge and has stabilized at 1.0 mg/dL. Evaluation of the patient's serum on post-operative day 115 showed a substantial decrease in the donor-specific Class II antibody titer to 1:16 (Figure 1). His outpatient immunosuppressive regimen consists of prednisone, mycophenolate mofetil, tacrolimus and sirolimus.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Case Report
  6. Discussion
  7. References

While the correlation between C4d deposition and AMR has been extensively studied in renal transplantation, we are not aware of any published reports of clinically significant AMR of pancreatic allografts confirmed by C4d staining. A recent report presented at the 10th World Congress of the International Pancreas and Islet Transplant Society described increased expression of C4d in pancreatic and kidney allograft biopsies undergoing acute rejection, but the clinical impact of this finding and the role of therapy against AMR in this context was not addressed (10).

In the present report, we describe the course of a patient who received a simultaneous kidney and pancreas transplant and subsequently developed clinical signs of acute kidney and pancreatic dysfunction that were likely secondary to AMR based on criteria set forth in the national consensus conference (1). Biopsies of both the kidney and the pancreas showed capillary staining with C4d. None of the biopsies showed more than a mild lymphocytic infiltrate that would be suggestive of cell-mediated rejection, and no necrosis, pyknotic nuclei or edema suggestive of ischemia were seen. Retrospective evaluation of the patient's pre- and post-transplant serum samples demonstrated the presence of HLA Class II-specific DSA. These antibodies probably developed from the patient's prior exposure to DR6 on his first kidney allograft and placed our patient at an intermediate to high risk for development of AMR (1). These findings underscore the value of screening for Class II HLA alloantibodies and performing a B-cell crossmatch as part of the pre-transplant evaluation in patients with evidence of sensitization since they can facilitate the assessment of rejection risk and the selection of appropriate donors.

It is interesting to note that whereas the biopsy-proven AMR of the kidney overlapped with and probably contributed to the presence of delayed kidney graft function, clinical evidence of pancreas graft rejection did not develop until approximately 35 days after transplant and after the patient had already received some therapy for AMR of his renal allograft. Interpretation of this observation is difficult since no pancreatic biopsies were performed early in the post-operative course and the clinical manifestations of pancreatic graft dysfunction (i.e. hyperglycemia) usually develop late in the course of rejection. Nonetheless, given the persistence of DSA in the patient's serum and continued mild renal graft dysfunction 1 month after transplant, one possibility is that the initial treatment for humoral rejection was not adequate. Indeed, the serum creatinine, amylase and glucose returned to normal only after additional therapy with plasmapheresis, IVIg and rituximab was completed on post-operative day 44 after transplant. The levels of HLA Class II alloantibodies in serum samples obtained after treatment were substantially reduced supporting the effectiveness of this additional treatment course.

Alternative explanations for the pancreatic parenchymal dysfunction include ischemic injury to the pancreas due to hemodynamic instability resulting from the gastrointestinal hemorrhage, and cellular rejection. The former explanation is unlikely for several reasons. (1) The serum amylase and glucose levels continued to trend downward until 11 days after the gastrointestinal bleed (Figure 1). Therefore, there is no strong temporal correlation between the bleed and the pancreatic dysfunction. (2) The pancreas biopsy did not show any edema, necrosis or pyknotic nuclei that would be expected with ischemic injury. (3) The serum amylase decreased and glucose control improved only after treatment for AMR was instituted. Cellular rejection is also unlikely to be the cause of the pancreatic dysfunction since the biopsies showed only minimal cellular infiltration and no venous endotheliitis, septal inflammation or eosinophils. The precise mechanism(s) by which AMR might cause islet dysfunction are not known. One possibility is that antibody deposition and resultant activation of complement and other inflammatory mediators in the microvasculature of the pancreas compromises blood flow to the islets that are supplied by a capillary network several times more dense than the surrounding exocrine tissue (11).

We chose to treat this patient's pancreatic graft dysfunction the same way we treat patients with AMR of their kidney allografts by reducing the amount and production of circulating alloantibodies. Therefore, our patient underwent plasmapheresis and received IVIg and rituximab. The patient had already received a course of polyclonal anti-thymocyte antibody as part of his induction immunosuppressive regimen. Other therapies that have been proposed for treating AMR of kidney grafts, such as immunoadsorption techniques that reduce the circulating IgG levels, may also be applicable to treating pancreas patients with AMR (12). It is likely that the above scenario may have been avoided if the patient's alloantibody levels and specificities had been identified prior to transplantation. To reduce the possibility of AMR in such situations, we are currently developing a protocol for pre-operative screening of Class II antibodies and are also evaluating the use of B-cell crossmatches for high risk patients.

Meanwhile, we recommend that pancreas biopsies from patients with clinical evidence of acute rejection should be studied for C4d deposition using immunofluorescence if there is evidence for donor-specific HLA alloantibodies, the pathologic evidence for cellular-mediated rejection is underwhelming, and/or the patient has risk factors for development of AMR such as prior transplantation or increased PRA levels. In addition, if a recipient of a simultaneous pancreas-kidney transplant has evidence of acute AMR of the kidney, then it would be appropriate to prospectively stain any pancreas biopsy for C4d. Together with other data, such as the presence or absence of donor-specific antibodies, the results of such staining will determine the appropriateness of post-operative monitoring and treatment directed at AMR.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Case Report
  6. Discussion
  7. References
  • 1
    Takemoto SK, Zeevi A, Feng S et al. National conference to assess antibody-mediated rejection in solid organ transplantation. Am J Transplant 2004; 4: 10331041.
  • 2
    Feucht HE. Complement C4d in graft capillaries—the missing link in the recognition of humoral alloreactivity. Am J Transplant 2003; 3: 646652.
  • 3
    Lederer SR, Kluth-Pepper B, Schneeberger H, Albert E, Land W, Feucht HE. Impact of humoral alloreactivity early after transplantation on the long-term survival of renal allografts. Kidney Int 2001; 59: 334341.
  • 4
    Michaels PJ, Fishbein MC, Colvin RB. Humoral rejection of human organ transplants. Springer Semin Immunopathol 2003; 25: 119140.
  • 5
    Behr TM, Feucht HE, Richter K et al. Detection of humoral rejection in human cardiac allografts by assessing the capillary deposition of complement fragment C4d in endomyocardial biopsies. J Heart Lung Transplant 1999; 18: 904912.
  • 6
    Duong Van Huyen JP, Fornes P, Guillemain R et al. Acute vascular humoral rejection in a sensitized cardiac graft recipient: Diagnostic value of C4d immunofluorescence. Hum Pathol 2004; 35: 385388.
  • 7
    Baldwin WM, 3rd, Samaniego-Picota M, Kasper EK et al. Complement deposition in early cardiac transplant biopsies is associated with ischemic injury and subsequent rejection episodes. Transplantation 1999; 68: 894900.
  • 8
    Magro CM, Pope Harman A, Klinger D et al. Use of C4d as a diagnostic adjunct in lung allograft biopsies. Am J Transplant 2003; 3: 11431154.
  • 9
    Freise CE, Kang SM, Feng S, Hirose R, Stock P. Excellent short-term results with steroid-free maintenance immunosuppression in low-risk simultaneous pancreas-kidney transplantation. Arch Surg 2003; 138: 11211125; discussion 1125-1126.
  • 10
    Becker L, Oliveira S, Malheiros D et al. Are pancreas allografts more immunogenic than kidney grafts? 10th World Congress of the International Pancreas and Islet Transplant Association. Geneva , Switzerland ; 2005.
  • 11
    Konstantinova I, Lammert E. Microvascular development: Learning from pancreatic islets. Bioessays 2004; 26: 10691075.
  • 12
    Liu M, Ji SM, Tang Z et al. C4d-positive acute humoral renal allograft rejection: Rescue therapy by immunoadsorption in combination with tacrolimus and mycophenolate mofetil. Transplant Proc 2004; 36: 21012103.