Preserved Endocrine Function in a Pancreas Transplant Recipient with Pancreatic Panniculitis and Antibody-Mediated Rejection

M. Prikis; D. Norman; S. Rayhill; A. Olyaei; M. Troxell; A. Mittalhenkle


American Journal of Transplantation. 2010;10(12):2717-2722. 

In This Article


Histology and Immunohistochemistry

The percutaneous pancreas allograft needle biopsies were performed with ultrasound guidance with an 18 gauge Bard Monopty biopsy instrument and immediately fixed in formalin. They were routinely processed and embedded in paraffin. Three-micrometer sections were prepared on glass slides and stained with hematoxylin & eosin (H&E). Immunohistochemical staining was performed with antibodies specific for complement 4, split product d (C4d) on formalin fixed, paraffin embedded sections (polyclonal antibody, Biomedica Gruppe, distributed by ALPCO diagnostics, catalog number 004-BI-RC4D, Windham, NH, USA).

Histocompatibility Testing

HLA Class I and II Flow panel reactive antibodies (PRA) analysis was performed using the FLOWPRA Screening microparticles (One Lambda, Inc., Canoga Park, CA, USA) according to the supplier recommendations. Analysis was done with a FACSCan (Becton Dickinson) on a 256 Channel linear scale. HLA Class I and II Single Antigen Bead testing was performed using the LabScreen Single Antigen Class I and Class II products (One Lambda, Inc.) according to the supplier's recommendations. The Flow HLA Class I and II Single Antigen bead analysis and donor specific antibody determination was performed using Luminex x Map technology on the LaBScan 100 flow analyzer for data acquisition and analysis).

The Case

A 44 year-old female patient who had received a pancreas after a kidney transplant (PAK) for diabetic nephropathy was admitted to our transplant unit after she suddenly developed several erythematous, painful nodules on her lower extremities (Figure 1).

Figure 1.

Cutaneous findings. A: Multiple erythematous subcutaneous nodules on both feet and legs with left predominance. B: Skin biopsy showing prominent subcutaneous fat necrosis; some of the adipocytes had faint cellular membranes (ghost cells, arrowheads) with neutrophils, leukocytoclastic granular basophilic debris and calcium deposition (saponification, arrows). This was consistent with the diagnosis of pancreatic panniculitis.

Her past medical history included type I diabetes mellitus, diagnosed at 8 years of age. She suffered renal dysfunction at the age of 30, with nephrotic range proteinuria. She had proliferative retinopathy, autonomic neuropathy, hypertension, hyperlipidemia, hypothyroidism, depression and gout. In addition to her kidney and pancreas transplant, her past surgical history was significant for a cesarean section, a tubal ligation and a hysterectomy performed for the management of menorrhagia.

Possible sensitizing events prior to transplantation included two pregnancies, leading to one live birth. There was no history of blood transfusions. At the age of 40, she received a zero antigen mismatch deceased donor kidney transplant. Her PRA at the time of the kidney transplantation was 0%. She was cytomegalovirus (CMV) negative at that time while the donor was CMV positive, putting her at high risk for CMV infection. She was Epstein-Barr virus (EBV) positive. She received prophylactic intravenous ganciclovir during her hospital stay and oral valganciclovir subsequently for 3 months per our protocol. Following basiliximab induction, her maintenance immunosuppressive regimen included prednisone, mycophenolate mofetil and tacrolimus. She had immediate renal allograft function and her serum creatinine level ranged between 1.3 mg/dL and 1.5 mg/dL.

The patient did well after her kidney transplant except for CMV viremia at 5 months, treated with intravenous ganciclovir and one episode of cellular rejection at 15 months after she was tapered off prednisone. She was treated with pulse steroids for the acute rejection and then re-started on low dose prednisone. A follow-up biopsy showed no acute rejection.

Eighteen months after the kidney transplant, she underwent a five-antigen mismatch pancreas transplant. The flow cross match prior to the transplantation was negative for both class I and II HLA antibodies. She received induction treatment with anti-T lymphocyte globulin (Thymoglobulin, Genzyme, Cambridge, MA, USA) per PAK protocol. Maintenance immunosuppression consisted of prednisone, mycophenolate mofetil and tacrolimus. She received oral valganciclovir for a total of 1 month for CMV prophylaxis given her history of CMV infection. Her immediate postoperative course was uneventful, with immediate normalization of serum glucose level and rapid normalization of her pancreatic enzymes.

Subsequently, she did well except for persistent chronic diarrhea, until she developed CMV disease at 1 year posttransplantation, which was diagnosed by colonic biopsies. She was treated with a full course of ganciclovir per protocol. She then had clostridium difficile infection and was successfully treated with oral metronidazole. The colonic disease resolved and she did well until 21 months after the pancreas transplant when she was noted, on routine laboratory testing, to have markedly elevated pancreatic enzymes (amylase 1063 U/L [normal range 25–115 U/L] and lipase 1331 U/L [normal range 114–286 U/L]). Both were substantially increased from her baseline values of 86 U/L and 39 U/L, respectively (Figure 2). Her blood glucose levels had been stable below 150 mg/dL until then, and, her C-peptide was 2.8 ng/mL. A pancreas allograft biopsy was performed (POD 628) which demonstrated moderate acute cellular rejection (Grade II per Banff 2008, Figure 3). The biopsy was remarkable for venulitis and a brisk infiltrate comprised of abundant eosinophils and plasma cells, with fewer numbers of lymphocytes and neutrophils. In addition, there was one large focus of acinar parenchymal effacement (Figure 3). The remainder of the biopsy showed a less intense inflammatory infiltrate, still rich in eosinophils, with individual acinar cell dropout. Of note, several enlarged nuclei were seen within the of the infiltrating lymphocytes, raising the possibility of posttransplant lymphoproliferative disorder; however, EBV in situ hybridization studies were negative and light chain studies showed polytypic plasma cells (not shown). She was admitted to the transplant service and treated for pancreas allograft rejection with anti-T lymphocyte globulin (Thymoglobulin, Genzyme) 1.5 mg/kg for a total of seven doses (4 as inpatient and 3 as outpatient). Her treatment course was uneventful; however, despite thymoglobulin treatment, her pancreatic enzymes did not improve significantly (Figure 2).

Figure 2.

Laboratory studies over time: Serum amylase, lipase and glucose.

Figure 3.

Microscopic images and immunostaining of pancreas transplant biopsies. A: Area of intense eosinophilic and plasmacytic inflammation in initial allograft biopsy (POD 628); residual pancreatic acinar parenchyma with damage, right. B: Second graft biopsy (POD 652) with endothelialitis in two vessels, and septal inflammation. C: Acinar parenchyma with lymphocytic inflammation and damage, most severe lower right. D: Immunohistochemical staining for CD3 shows a predominance of T cells in second biopsy. Arrows and inset shows T cells in association with islets. E: Serial section stained for CD20, demonstrating a minority of B-cells; arrows demonstrate islets. F-G: Second biopsy immunohistochemically stained for C4d, which focally highlights capillaries within acinar parenchyma, as well as the endothelium of arteries and veins (the staining highlights arterial elastica, likely non-specific). Inset in F shows interacinar CD68 positive macrophages, likely in capillaries. Original magnifications: A: 400x, B-F: 200x, insets 400x; A-B: Hematoxylin & eosin staining, C: CD3 staining, D: CD20 staining, E-F: C4d staining.

Her blood glucose control continued to be good despite high dose of steroids given as premedication for anti-T lymphocyte globulin (Figure 2). She was discharged home in good condition (despite elevated enzymes), with a plan for a repeat biopsy in 1 month, per protocol.

Two days after discharge, however, she noticed the rapid development of multiple, tender, erythematous nodules on both her legs and feet (Figure 1A). On admission, both upper and lower extremities were warm with several tender, erythematous subcutaneous nodules on both feet (left greater than right). She was afebrile, with normal vital signs and had no abdominal pain or pedal edema. However, computed tomography scan of her abdomen and pelvis showed stranding around the pancreas consistent with acute pancreatitis, but no other acute changes were seen. A biopsy of a left medial ankle nodule was taken demonstrating prominent fat necrosis, neutrophilic infiltration, the presence of leukocytoclastic granular basophilic debris and numerous calcium deposits (saponification) consistent with pancreatic panniculitis (Figure 1B). Based on a previous study suggesting that octreotide may have a role in the treatment of pancreatitis, octreotide 100 mcg three times daily was given for seven days.[3] Even though mycophenolate mofetil has been associated with pancreatitis, the dose was not modified due to the concern for late rejection. Clinical improvement of the nodules was observed, but her pancreatic enzymes started to rise again after a brief decline (Figure 2). Blood and urine cultures as well as bacterial and viral cultures of the skin biopsy tissue were negative.

Blood glucose control remained normal throughout. Serum creatinine was fluctuating between 1.7–2.2 mg/dL (baseline 1.5–1.9 mg/dL). On hospital day 13 (POD 652), another pancreas allograft biopsy was performed and the patient's serum was checked for donor specific antibodies (DSA) against both the kidney and the pancreas donor.

The second pancreas biopsy showed severe acute cellular rejection (Grade III per Banff 2008, Figure 3).[1] Biopsy sections demonstrated septal edema with a diffuse, moderate intensity lymphocytic infiltrate involving acinar parenchyma, with accompanying tissue damage and prominent multifocal endothelialitis (Figure 3). The plasmacytic and eosinophilic components of the infiltrate were markedly diminished compared to the prior biopsy. Overall the biopsy reflected increased injury, with worsened rejection, progressively greater parenchymal atrophy and increased fibrosis (Figure 3). High power examination demonstrated mild dilation of capillaries scattered throughout the biopsy, with 1–2 mononuclear cells per luminal cross section, but without neutrophilic capillaritis. Immunohistochemical studies showed large numbers of CD3-positive T-cells, including rare T-cells within islets, many CD68-positive monocytes/histiocytes and very few CD20-positive B-cells (Figure 3). The brisk T-cell infiltrate, acinar damage and arteritis provided evidence of ongoing cellular rejection, while the presence of weak to moderate C4d immunohistochemistry staining of a subset of parenchymal capillaries (20–30%, Figure 3) suggested that there might also be a component of antibody-mediated (humoral) rejection.[1] Interestingly, a retrospective C4d staining of the initial pancreatic biopsy showed only weak staining of septal veins and few interacinar capillaries (on the order of 5% or less).

The patient was treated with high dose steroid pulse (5 mg/kg) for 5 days with a taper while waiting for the flow PRA and DSA results. She was found to have flow PRA class I 7% and class II 79% with DSA DQ5 and DR51 to the pancreas donor (Table 1). Based on the presence of DSA, allograft dysfunction and C4d staining on tissue biopsy, the clinicopathologic picture was consistent with humoral and concomitant cellular rejection.

With the diagnosis of antibody-mediated rejection, a course of plasmapheresis and intravenous immunoglobulin (IVIG) infusion was instituted. The humoral rejection treatment protocol included 4 cycles of plasmapheresis and 4 doses of IVIG (500 mg/kg). In addition, given the severity of the rejection, she also received one dose of rituximab 375 mg/m2 body surface area. She tolerated the treatments well. Her serum glucose remained normal or slightly elevated throughout the treatment and her pancreatic enzymes decreased slowly to physiologic levels over a period of 2 months. Six months after treatment, she was euglycemic, with HgA1c of 5.8 and a C-peptide of 2.6. Two years from this episode, she continues to do well, with good allograft function, without elevation of her pancreatic enzymes and her HgA1c consistently below 6.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.