Allograft biopsies in management of pancreas transplant recipients.P Randhawa
Division of Transplantation Pathology, Department of Pathology, C903.1 Presbyterian University Hospital, University of Pittsburgh, Pittsburgh, PA 15213, USA. , USA
Correspondence Address: Source of Support: None, Conflict of Interest: None PMID: 12082334
Source of Support: None, Conflict of Interest: None
Pancreatic transplantation is becoming increasingly accepted as a treatment modality for Type 1 diabetes mellitus. When allograft dysfunction is noted during follow up of patients, a biopsy is an extremely useful tool to diagnose various forms of rejection, and to rule out non-immunologic causes of graft malfunction, such as donor disease, ischaemic/preservation injury, vascular thrombosis, pancreatitis, post-transplant lymphoproliferative disease, technical complications and recurrence of diabetes mellitus. In addition to its role in establishing the primary diagnosis, a biopsy can grade the severity of pathology present, help determine the most appropriate therapy, and provide information relevant to graft prognosis.
Keywords: Diabetes Mellitus, Insulin-Dependent, pathology,therapy,Graft Rejection, pathology,Graft vs Host Disease, Human, Pancreas, pathology,Pancreas Transplantation, adverse effects,Pancreatitis, etiology,pathology,Tissue Donors, Transplantation Tolerance,
The first pancreas transplant was performed in 1966, and by the end of the 20th century more than 10,000 procedures had been performed.,, The best candidate is a patient with end-stage renal disease due to Type 1 diabetes mellitus. A simultaneous pancreas kidney transplantation in this situation alleviates renal failure, and offers the prospect of controlling diabetes without continual therapy with insulin or hypoglycaemic drugs. Pancreas transplantation alone has also been attempted in Type 1 diabetics without end-stage renal disease, who are on the verge of serious complications such as pre-proliferative retinopathy, and peripheral or autonomic neuropathy. The operation is justified only if the complications are perceived to be more serious than those of surgery, and of the lifelong immunosuppression that the patient must receive thereafter. If these complications occur in a patient who has already received an allograft kidney, a ‘pancreas after kidney’ transplantation can be performed. The results of simultaneous kidney pancreas transplantation have historically been generally superior to ‘pancreas after kidney’ and ‘pancreas alone’ procedures. With increasing experience, and particularly if HLA matching for DR loci is done, comparable results can be reportedly obtained for all three procedures. The University of Pittsburgh has reported 1-, 2-, and 4-year graft survival rates of 98%, 95% and 86% respectively. With modern immunosuppressive regimens it has become possible to achieve acute rejection rates of less than 20%. Type II diabetes mellitus is not a suitable indication for pancreatic transplantation because peripheral insulin resistance is an important element in the pathophysiology of these patients.
Clinical studies of pancreas transplant recipients show that insulin independence is achieved in technically successful cases, although abnormal glucose tolerance tests may persist. After isolated pancreas transplantation, reversal of diabetic lesions in the native kidney has been reported. In diabetic combined kidney-pancreas transplant recipients, renal allograft survival is not demonstrably improved compared to non-diabetic patients.,,,, Clinical improvements in diabetic neuropathy symptoms have been recorded, and these are accompanied by enhanced nerve conduction velocity. Diabetic microangiopathy may stabilize, improve or deteriorate after pancreas transplantation, and it is not clear if its clinical course differs from that seen in non-transplanted controls.
The principal purpose of this review is to demonstrate how examination of pancreatic tissue sampled from the allograft can provide an insight into the functional status of the grafted organ. A systematically conducted pathological examination of such tissue can provide information of great value to the clinical management of these cases.
A knowledge of various operative techniques used in pancreas transplantation is vital to the pathologist when he is called upon to evaluate an allograft in the surgical pathology laboratory or the autopsy suite.,. Briefly, pancreatic allografts may be segmental or total and enteric or bladder drained. Segmental pancreatic grafts, which consist of the body and tail of the pancreas supplied by the splenic vessels, were mostly performed in the past. Currently, the practice is to use total pancreas grafts including either the entire duodenum or a button of duodenum encompassing the papilla of Vater and duct of Santorini. Grafting in each case is done intraperitoneally with vascular anastomoses to the iliac vessels. Lately, portal venous drainage is gaining in popularity as it leads to more physiologic levels of insulin with a lower incidence of dyslipidaemia. In enteric drained grafts, the graft duodenum is anastomosed to the side of a Roux-en-Y limb of jejunum (duodenojejunostomy). In urinary bladder drained grafts, a duodenocystostomy is performed instead. The majority of pancreas transplants performed so far have been bladder drained, but the number of intestine drained grafts in now increasing.
In simultaneous kidney-pancreas transplant recipients, a renal allograft biopsy can be used as a surrogate marker for monitoring pancreatic graft function. Indeed, we have seen acute rejection in renal allograft biopsies performed in the setting of a stable serum creatinine but rising amylase. Nonetheless, it is periodically necessary to specifically monitor pancreatic allograft function by a biopsy obtained during cytoscopy (bladder drained pancreas grafts) or duodenoscopy (enteric drained pancreas grafts).,,, In pancreatico-duodenal grafts, endoscopic biopsy of the allograft duodenum appears to as reliable as a biopsy of the pancreas itself., Ultrasound guided percutaneous biopsies of the pancreas graft are also feasible using an 18 gauge needle, and result yield satisfactory samples in up to 93% of cases., Complications related to the biopsy procedure are infrequent, and include bleeding in 3%, and mild pancreatitis in 7% of cases.
Even though an 18 gauge core biopsy is safe in experienced hands, some clinicians continue to be concerned about the possibility of traumatic enzyme leak and pancreatic tissue necrosis. To minimise this possibility, some centres have used fine needle aspiration cytology for monitoring isolated pancreas grafts. The diagnostic sensitivity of an aspirate, 59.1%, is lower than that achieved by conventional biopsy, 75%. A serious limitation of fine needle aspiration is that it in unsuitable for evaluating arteritis and chronic rejection. Occasionally, aspiration of peritoneal macrophages in the setting of ascites and graft infarction has led to a false positive diagnosis of rejection.
In an effort to avoid trauma to the allograft pancreas altogether, pancreatic juice cytology has also been advocated as a method for monitoring graft function. Samples are obtained via a pancreatic duct catheter passed through the duodenal wall. However, the necessity of requiring a percutaneous catheter drainage limits use of this method to the first few weeks after transplantation, and there is also an increased risk of graft pancreatitis. For bladder drained grafts, urine cytology has also been used instead of pancreatic juice analysis., As with fine needle aspiration, cytologic examination cannot diagnose vascular rejection and chronic rejection in the graft.
The remainder of this review will summarise the clinical and biopsy findings in various graft dysfunction syndromes seen after pancreas transplantation.
This term refers to poor graft function due to use of a suboptimal donor organ. To avoid this, a biopsy of all marginal donor organs should be performed before transplantation, to ascertain the presence of significant ischaemic injury and to evaluate any pre-existing disease. Ischaemic changes are recognised by cytoplasmic swelling, microvesicular steatosis, and frank cytolysis. Fat necrosis related to surgical manipulations is often seen in the donor peri-pancreatic fibroadipose tissue. In elderly donors, arteriosclerosis and vascular hyalinosis may be apparent within the pancreatic arterial tree. Changes of chronic pancreatitis can occur in organs derived from donors with a history of alcohol abuse or stones in the pancreatico-biliary tree.
Antibody mediated rejection is rarely encountered in clinical practice due to the availability of reliable immunologic screening techniques for detecting donor specific pre-formed antibodies. Nonetheless, hyperacute rejection has been reported, even in patients with a negative pre-transplant cross-match. The histopathology of antibody-mediated rejection is characterised by margination of neutrophils, arterial fibrin thrombi, and ischaemic necrosis in advanced cases. Interestingly, low titre anti-donor antibodies detected by flow cytometry are not necessarily injurious to the graft.,
Acute rejection of the pancreas is suspected when pancreatic dysfunction develops, and cannot be explained by an ischaemic insult such as vascular thrombosis, an infectious complication, or drug toxicity.,,,,,, Elevation of serum amylase levels is typically observed, but, in general, correlates poorly with rejection because circulating amylase levels can also rise in the context of intestinal perforation, injury to the native pancreas, and renal failure. In our experience, serum lipase is a more sensitive marker of pancreatic acinar injury than serum amylase. A rise in blood glucose is a late indicator of pancreas allograft rejection, and signifies that more than 90% of the islet cell mass may already have been destroyed. Elevated glucose in the face of normal serum amylase or lipase should raise the possibility of steroid, cyclosporine or tacrolimus induced toxicity to the Islets of Langerhans More Details. In allografts draining into the urinary bladder, a decrease in urinary amylase excretion is perhaps the earliest metabolic marker for rejection. Again, acinar tissue is more sensitive than endocrine tissue to alloimmune injury, hence decreased urinary amylase excretion precedes hyperglycaemia by 1 to 2 days. Of course, hypoamylasuria does not prove that the injury to the pancreas is immunologically mediated, so that a tissue diagnosis is still needed to clinch the diagnosis of rejection.
A histologic diagnosis of acute cellular rejection is made if one finds inflammatory infiltrates in which immunoblasts, activated lymphocytes and plasma cells are a significant component.,, These infiltrates invade and damage the acinar cells, ductal epithelium and the islets of Langerhans. Vascular lesions such as venous endothelialitis and frank arteritis are more specific for the diagnosis of rejection, since acinar and ductal inflammation may also occur in pancreatitis. The islets of Langerhans can show a lymphocytic infiltrate producing a lesion termed isletitis or insulitis. Intra and peri-pancreatic nerves may also become a target for allo-immune injury. Rejection associated infiltrates can spill over into the peri-pancreatic fat. However, fat necrosis and inflammation can also be the result of surgical trauma and pancreatitis.
The sensitivity and specificity of a biopsy diagnosis of acute cellular rejection is high in our experience, but has not been precisely calculated in the literature. Hence, if clinically suspected rejection is not confirmed by histopathologic examination, repeat biopsy should be considered to rule out a sampling problem, provided other potential causes of graft injury have been reasonably excluded. In one study, such a repeat biopsy documented rejection missed on the first specimen in 2/7 cases.
Histologic criteria to assess the severity of rejection in allograft biopsies have been published., The schema of Drachenberg and colleagues, which defines the use of five clinically and statistically validated grades of rejection, is described below.
- Grade 0 or normal refers to the absence of changes attributable to acute rejection.
- Grade I acute cellular rejection, also designated as inflammation of undetermined significance, is charac- terised by sparse, mononuclear infiltrates of mostly small lymphocytes, located within the fibrous septae. The veins, arteries and acini are free of inflammatory change, but mild ductal inflammation is permitted.
- Grade II acute cellular rejection, or minimal acute rejection, is defined by the presence of septal mononuclear infiltrates with superimposed venous endothelialitis. In the absence of venous endothelialitis, grade II rejection can be also defined as the presence of any 3 of the following 4 features: (a) septal inflammation composed of “activated” lymphocytes, (b) presence of eosinophils, (c) focal acinar inflammation with presence of 1 or 2 foci of at least 10 inflammatory cells each, and (d) ductal inflammation with permeation of inflammatory cells through the ductal basement membranes.
- Grade III, or mild acute rejection, is defined by the presence of septal inflammation and 3 or more foci of acinar inflammation, each focus of acinar inflammation being comprised of at least 10 cells. These findings are typically accompanied by acinar cell injury.
- Grade IV, or moderate acute rejection, requires the presence of arterial inflammation in the form of endothelialitis or necrotising arteritis.
- Grade V, or severe acute rejection, is diagnosed in the presence of multifocal or confluent parenchymal-necrosis accompanied by features of grade II, III or IV rejection.
The reproducibility of this grading system is satisfactory as shown by an overall inter-observer kappa statistic of 0.83, and kappa scores for individual grades of rejection varying between 0.66 and 0.9. Correlations between rejection grade and ultimate graft outcome attest to the clinical relevance of this grading schema., Thus, after a mean follow up of 19.3±1.9 months, rates of graft loss associated with rejection grades 0-I, II, III, IV and V were 0%, 11.5%, 17.3%, 37.5%, and 100% respectively. Treatment for acute rejection resulted in clinical biochemical response rates of 0%, 17%, 71%, 75% and 50% respectively in these groups. Tissue documentation of rejection by biopsy is superior to relying on chemical analyses performed on urine: one study on bladder drained grafts showed that only 8 of 17 (47%) biopsies performed during episodes of falling urinary amylase documented histopathologic acute rejection. Relying on urinary amylase monitoring alone led to a false positive diagnosis of rejection 53% of the time.
The pathogenesis of acute cellular rejection in the pancreas follows general immunological principles applicable to all solid organs. A detailed discussion of this topic is beyond the scope of this review. Briefly, class I major histocompatibility antigens are expressed in the normal pancreas by the vascular endothelium, duct epithelium, and weakly by the pancreatic acini and islets of Langerhans. Class II major histocompatibility antigen expression is normally seen only in the ductal epithelium and vascular endothelium. Following transplantation, the expression of both Class I and Class II antigens increases in intensity and generates intra-graft T cells specifically sensitised to donor antigens. The interaction of these sensitised T-lymphocytes with the corresponding antigens triggers the release of a variety of cytokines which mediate allograft injury, and recruit additional immune effector cells into the transplanted pancreas.
The morphological hallmark of chronic rejection is an obliterative vasculopathy involving the major pancreatic arteries and its medium and small-sized branches. The intima is thickened and contains deposits of a mucopolysaccharide-like material. Marked intimal proliferation and luminal occlusion is seen in later cases, and has prompted use of the term concentric fibro-proliferative endarteritis. Variable numbers of mononuclear infiltrates may be seen in the thickened intima. Thrombosis and recanalization can occur. The pancreatic lobules show extensive fibrosis accompanied by relatively sparse mononuclear infiltrates. The islets of Langerhans may be difficult to identify. The pancreatic ducts become thick, fibrous, and infiltrated by lymphocytes. The degree of parenchymal fibrosis can be graded and appears to correlate with clinical outcome.
Bacterial infections can take the form of wound dehiscence, bacterial peritonitis, intra-abdominal abscesses, urinary tract infections, pneumonia, or systemic sepsis. In some transplant programs, urinary tract infections have occurred in virtually all bladder-drained patients., Surgical site infections have been reported in up to 30% of cases, and other miscellaneous infections in 5% of individuals. Reactive histopathologic changes in the pancreas secondary to intra-abdominal infections should not be confused with rejection. In acute infections, biopsies show mixed septal inflammatory infiltrates with proliferation of fibroblasts. Chronic infections can lead to extensive paraseptal fibrosis in the exocrine lobules.
Cytomegalovirus inclusions are reported in up to 13% of pancreatic grafts., Sometimes, viral inclusions are seen in grafts with histologic evidence of concomitant acute rejection. In other cases, CMV infection appears to be the primary cause of graft dysfunction. Extra-pancreatic manifestations of CMV disease include gastro-intestinal haemorrhage, hepatitis, and multisystem viral disease. CMV infected cells are readily recognised at biopsy by cytomegaly and prominent eosinophilic intra-nuclear inclusions surrounded by a clear halo. Biopsy identification of CMV pancreatitis as the primary cause of graft dysfunction spares the patient from unnecessary intensification of immunosuppression and triggers specific anti-viral therapy with ganciclovir. Delays in initiation of therapy can lead to pancreatic abscess formation and graft loss.,
PTLD is a complication of Epstein-Barr virus (EBV) infection, wherein virus infected B-lymphocytes escape normal immune defence mechanisms, and initiate an uncontrolled lymphoproliferative state. The incidence of this complication in pancreas transplant recipients varies from 2.2 to 12%., Discontinuation or significant reduction of immunosuppression and administration of ganciclovir is the usual first line of treatment. Some patients go on to develop fatal disseminated disease despite additional treatment with alpha-interferon, radiation or chemotherapy.
Biopsy distinction between PTLD and severe acute rejection is important, since the appropriate treatment is reduction of immunosuppression for PTLD, but aggressive anti-T cell therapy for severe acute rejection. PTLD typically shows expansile nodular mononuclear infiltrates with irregular foci of geographic necrosis and nuclear atypia exceeding 25% of the total population. Some biopsies present a monotonous appearance indistinguishable from conventional lymphomas. Acinar injury is not striking in PTLD, but venous endothelialitis is common. Arterial endothelialitis or necrotizing vasculitis was not found in one study which examined 4 pancreatic specimens with PTLD. Infiltration of the hilar soft tissues or nerves can be seen in PTLD as well as in severe acute rejection, although this finding has been said to be less prominent in the context of acute rejection. Immunophenotyping, and EBV in-situ hybridisation can be extremely helpful in establishing a correct diagnosis. PTLD lesions are mostly rich in B-cells and are EBV positive, while rejection is associated with a primarily T-cell infiltrate, which is EBV negative. Kappa and lambda light chain staining, immunoglobulin gene rearrangement, and oncogene studies can be performed to determine if the lesion is clonal and to help determine ultimate prognosis.
In this rare complication, better known in bone marrow transplant recipients, graft-derived mononuclear cells migrate into the recipient circulation in large enough numbers to result in host organ damage. Clinical manifestations include skin rash, hepatitis, gastrointestinal symptoms and pancytopenia. Donor lymphocyte-derived antibodies directed against host ABO blood group antigens can precipitate a haemolytic anemia. Biopsy of the affected organs can confirm parenchymal injury and document that the associated mononuclear inflammatory cells bear donor rather than recipient specific HLA antigens.
Graft pancreatitis develops in up to 16% of pancreatic allograft recipients.,, Pathogenic factors include sepsis, ischemia-reperfusion injury, surgical trauma, steroids, calcineurin inhibitors, alcohol, stones and viral infections. The clinical presentation is characterised by fever, local pain, hyperamylasaemia and hyperglycaemia. Histological examination shows neutrophils throughout the parenchyma with focal necrosis of the acinar cells. The diagnosis of pancreatitis has also been made by cytological examination of pancreatic juice.
Recurrent or persistent acute pancreatitis can result in chronic pancreatitis. The histopathology findings include chronic duct inflammation, epithelial proliferation, squamous metaplasia, calcification, and islet cell atrophy or hyperplasia. The pancreatic ducts show concentric periductal fibrosis with compressed, dilated or angulated ducts. An unusual form of chemical pancreatitis is seen in grafts injected with liquid synthetic polymers to reduce the incidence of anastomotic leaks. Unfortunately, these polymers harden inside the ducts forming a solid cast, which leads to obstructive atrophy of the pancreatic acini, although the islets of Langerhans remain functional. Prolonged tissue contact with the polymers initiates foreign-body granulomas which heal with extensive fibrosis.
In urinary bladder drained grafts proteolytic enzymes secreted by the pancreas come in direct contact with the bladder mucosa. These enzymes are secreted in an inactive form (e.g. trypsinogen), and in general, do not get activated since enterokinase is not produced by the urothelial mucosa. Occasionally, however, bacterial infections or pancreatitis generate a milieu conducive to in-situ trypsinisation of urothelial cells. This can lead to distressing complications such as hemorrhagic cystitis, urethritis, and penile necrosis.
Thrombosis of the pancreatic vessels leading to graft infarction is reported in 1-19% of pancreatic allografts.,,, Harvesting/ischaemic injury, surgical trauma, post-operative hypo-perfusion and actual kinking of the pancreatic vessels can all lead to the vascular damage that precedes thrombogenesis. Clinically, patients present with sudden loss of graft function early after transplantation. Before attributing thrombosis to mechanical factors, it is important to exclude cellular as well as antibody-mediated rejection, both of which can cause severe endothelial injury to the pancreatic vasculature. Vascular thrombosis can also be the sequel of acute pancreatitis and chronic rejection.
Type I diabetes mellitus is a disease initiated by autoimmune injury to the islets of Langerhans. Evidence is available to support the participation of both cell mediated,, and humoral,, immunity in the immunopathogenesis of this disease. Recurrence of Type 1 diabetes was initially reported in non-immunosuppressed HLA-identical sibling or monozygous twin pancreas recipients, but can also occur in HLA mismatched individuals. Donor T-cell chimerism and suppression of autoimmune reactions by post-transplant immunosuppressive drugs possibly accounts for the rarity with which this condition is encountered., Clinically, patients present with progressive hyperglycaemia and become insulin dependent 2.5 to 58 months after transplantation. Serum amylase and lipase are normal but there is evidence of circulating antibodies to islet cells, glutamate decarboxylase, and tyrosine phosphatase. A rigorous diagnosis of recurrent diabetes requires exclusion of diabetes in the donor.
Allograft biopsies performed early in the course of recurrent diabetes show mononuclear infiltrates in the islets of Langerhans, (“isletitis”), and degranulation or destruction of the beta cells.,, The infiltrating lymphocytes are primarily CD-8 positive T lymphocytes. Immunofluorescence studies demonstrate strong expression of Class I MHC antigens in the islet cells. Following the onset of insulin dependence, pathologic examination may show complete absence of beta cells in the islets of Langerhans. The inflammatory infiltrate becomes less prominent and can completely disappear. Immunohistochemical methods demonstrate very elegantly that the immune injury to the islets in recurrent Type 1 diabetes is selective for beta cells. By comparison, isletitis occurring in the setting of acute rejection preserves the normal relative proportions of insulin and glucagon producing cells. The ratio of insulin to glucagon producing cells (I/G ratio) falls progressively as the disease progresses. It has been suggested that an I/G ratio below 1.00 can be used to make a diagnosis of recurrent diabetes, even when co-existing rejection makes the actual interpretation of any isletitis present difficult.,
In summary, pancreatic transplantation is becoming increasingly accepted as a treatment modality for Type 1 diabetes mellitus. When allograft dysfunction is noted during follow up of patients, a biopsy is an extremely useful tool to pin point the underlying cause. In patients with simultaneous kidney and pancreas allografts, a renal allograft biopsy can often be used as a surrogate method to monitor pancreatic graft function. In pancreas only transplants, and in cases where the kidney and pancreas are derived from different donors, a pancreatic biopsy becomes essential. The most convenient and safest way of obtaining tissue for study is a percutaneous needle biopsy performed under ultrasound guidance. A common use of the allograft biopsy is to rule out or rule in the possibility of acute cellular rejection. Morphologic criteria used to make a diagnosis of acute rejection include an activated mononuclear inflammatory infiltrate, damage to the acinar or ductal epithelium, venous endothelialitis and intimal arteritis. Grading the severity of acute rejection can aid in selection of the most appropriate anti-rejection therapy, and help ascertain graft prognosis. Chronic rejection is characterised primarily by a fibro-obliterative arteriopathy with accompanying parenchymal fibrosis. As with other solid organ allografts, chronic rejection remains a significant obstacle to the long-term success of pancreatic transplantation.