Banff Categories

In 1991 a group of renal pathologists, nephrologists and transplant surgeons met in Banff, Canada to agree a standardised international classification system of renal allograft biopsies. A Banff conference is held every year and the original Banff classification scheme has evolved significantly since first introduced. The most recent review has 6 categories:

• Normal or Nonspecific Changes
  • Antibody-Mediated Changes
  • Borderline Changes
  • T-cell Mediated Rejection
  • Interstitial Fibrosis and Tubular Atrophy
  • Other Changes Not Considered to be Due to Acute or Chronic Rejection

An adequate biopsy sample for assigning of Banff categories must have a minimum of 7 glomeruli with one artery. Ideally the sample should have 10 or more glomeruli with at least two arteries.

Category 1 shows no change or nonspecific changes.

Category 2, Antibody mediated changes, are classified into three groups, 2a active antibody mediated changes, 2b chronic active antibody-mediated rejection and 2c C4d staining without evidence of rejection. The first two groups 2a and 2b both require 3 things:

• Evidence of tissue injury such as microvascular inflammation
• C4d staining in peritubular capillaries
• Presence of donor specific antibodies (DSA)

Category 2c has C4d without evidence of circulating DSA.

Category 3, borderline changes are biopsies which are suspicious for acute rejection and involves either minor interstitial inflammation with foci of tubulitis or moderate to severe interstitial inflammation (T cell inflammation), both with no arteritis (inflammation of the walls of arteries).

Category 4, T-cell mediated rejection, are classified into two groups, 4a Acute T-cell-mediated rejection and 4b Chronic active T-cell-mediated rejection. Interstitial inflammation >25%, mild to moderate intimal arteritis and tubulitis and presence of arteritis. Chronic active T-cell-mediated rejection additionally requires chronic allograft arteriopathy.

Category 5, Interstitial Fibrosis and Tubular Atrophy, as the name implies, requires the presence of mild, moderate or severe interstitial fibrosis and tubular atrophy.

Category 6, Other Changes Not Considered to be Due to Acute or Chronic Rejection, involves changes seen on biopsy that are not due to acute or chronic rejection. Examples include:

• BK virus nephropathy
• Posttransplant lymphoproliferative disorders
• Calcineurin inhibitor nephrotoxicity
• Acute tubular injury
• Recurrent disease
• De novo glomerulopathy
• Drug induced interstitial nephritis
• Pyelonephritis (inflammation of the kidney tissue and renal pelvis)

This patient has been classified as 2a – Active antibody mediated changes. This category implies the patient has evidence of evidence of tissue injury, is C4d Pos., and the laboratory has demonstrated the presence of DSA. If creatinine levels are also rising, the renal unit may choose to enhance the immunosuppression and/or undertake plasma exchange. The H&I laboratory can support this course of treatment by undertaking more DSA studies. In some circumstances it may also be informative to undertake a post-transplant crossmatch to compare with the crossmatch done at the time of transplant.

Chronic Allograft Nephropathy (CAN)

The definition of Chronic Allograft Nephropathy (CAN) has changed over the years. Initially it was used almost interchangeably with chronic rejection. However ‘chronic rejection’ implies an alloimmune mechanism whereas the histological features of CAN have been shown to be present in post 1 year protocol biopsies even in the absence of alloimmune processes. CAN has come to be defined as an end point of tubular atrophy and interstitial fibrosis in the graft resulting from a number of immune and non immune processes. Typical histological features also include fibrous intimal thickening and the presence of glomerular lesions.

The mechanisms which contribute to the development of chronic allograft nephropathy can be divided into three broad categories, Input causes such as pre-existing chronic donor condition, Immune causes such as acute and chronic rejection and Load causes such as donor – recipient size disparity.

Input mechanisms which contribute to the development of CAN may be donor specific, recipient specific or related to the transplant process itself. Increased donor age has been shown to be associated with increased rates of CAN. Kidneys from older donors may have a reduced ability to withstand stress and may already be carrying some structural damage. They also typically have delayed renal function compared to kidneys from younger donors. This is compounded if the donor has a history of hypertension which, in itself, increases pressure in the glomerulus, causing injury to glomerular cells and leads to glomerulosclerosis – a scarring of the glomerulus. Glomerulosclerosis disturbs the filtering process of the kidney and allows proteinuria – leaking of protein from the blood into urine. In addition, any pre-existing chronic kidney condition in the donor may be carried over to the patient and may contribute to the development of CAN. The type of donor, Donation after Cardiac Death (DCD) versus Donation after Brain Death (DBD), makes differing contributions to CAN.

Patient input factors shown in studies to be associated with the development of CAN include increased patient age, male gender, high body mass index (BMI) and a diabetic status. Like hypertension, diabetes can cause glomerulosclerosis and therefore loss of graft function leading ultimately to CAN. Diabetes pre transplant does carry the additional distinction of being a good predictor of post transplant cardiac events.

The transplant process itself contributes, as an input factor, to the development of CAN. Many of the stages of the transplant process could potentially injure the kidney, triggering pro-inflammatory responses which may further harm the kidney. Injury can occur during donor maintenance prior to retrieval, during the retrieval process and during cold and warm ischaemia. The role of ischaemia in CAN is highlighted by the reduced incidence of CAN in transplantation from live donors, which typically have a much shorter ischemic period, compared to those from deceased donors. Prolonged warm and cold ischaemia leads to the upregulation of HLA antigens and adhesion molecules and to the increased production of cytokines, all contributing to inflammatory responses which cause acute tubular necrosis, one of the symptoms of CAN. Another common mechanism of injury leading to the development of CAN is reperfusion injury. During reperfusion the resupply of blood to an organ that has been deprived of oxygen leads to the production of reactive oxygen species (ROS) which contribute to kidney injury. Taken together, ischaemia and reperfusion injury (IRI) are the biggest non immune cause of delayed graft function. Delayed graft function correlates very strongly with CAN. DGF is contributed to by all the input factors including existing chronic donor kidney condition and IRI as well as a number of immune mechanisms such as acute rejection.

The immune mechanisms which contribute to CAN include acute and chronic alloantibody mediated rejection (AMR) as well as acute and chronic cellular rejection and leads to a distinct set of histological feature, chiefly, Transplant Glomerulopathy (TG), characterised by a doubling of the glomerular basement membrane (GBM). Late acute rejection that is recurrent and resistant to treatment is a key predictor of the development of CAN. The number of acute rejection episodes correlates well with the number of HLA mismatches, even in the CNI era. HLA matching reduces the incidence of both early and late acute rejection. The relationship between pre-existing donor specific anti HLA antibodies (DSA) and the development of hyperacute or accelerate rejection is well documented. The level of panel reactive antibody (PRA) of non donor specific anti HLA antibodies in circulation pre transplantation and the development of donor specific anti HLA antibodies post transplantation are the strongest independent risk factors of acute rejection. Studies show that the presence of class II or of class I and II antibodies together may be more predictive of this antibody mediated rejection (AMR) than the presence of class I antibodies alone. A key surrogate of AMR is the presence of C4d staining in the peritubular capillaries, though not all patients with anti-HLA antibodies will necessarily exhibit C4d deposition (The Banff classification of renal allograft pathology has been changed recently to capture the role of C4d staining.). Acute cellular rejection involves diffuse infiltration of the interstitium with CD4+ and CD8+ T cells and macrophages with an alloreative immune response initiated at the graft endothelium. This leads to cytotoxic T cell activation, NK and monocytes infiltration and antibody formation. Immune responses are mainly directed against HLA though responses involving antibodies against other antigens such as the MHC like class I A and B (MICA and MICB) and antiendothelial cell antibodies have been reported. Alloreactivity involves both the direct and indirect pathways of antigen presentation with the indirect pathway in particular believed to play a dominant role in the development of CAN whilst the direct pathway is believed to be predominantly involved in early acute cellular rejection. As it requires donor antigen presenting cells which will not be long lived post transplant.

The incidence of true immunological chronic cellular rejection is reduced in the CNI era in patients who are compliant with their Immunosuppression regimes. When present however chronic cellular rejection is associated with a significantly increased risk of CAN. Recent data from the 14th international histocompatibility workshop (Terasaki et al) demonstrated that four year deceased donor kidney allograft survival was 20% less in patients with donor specific antibodies compared to donors with no HLA antibodies. The evolution of chronic antibody mediated rejection in believed to begin with the development of DSA, followed by C4d in the peritubular capillaries (PTC), then the development of TG and other graft pathology and finally graft dysfunction.

The use of desensitisation protocols to transplant patients across the ABO and HLA barriers as part of ABO incompatible and/or HLA antibody incompatible transplant programs will present interesting new long term data on the role of immune responses in the development of CAN. These patients are potentially at higher risk of AMR if not in the acute phase then certainly in the chronic phase. In addition, de-novo IgG HLA alloantibodies developed post transplant in addition to functioning as a pathogenic factor in the development of CAN may also be a surrogate marker for ongoing indirect T cell immunity.

The non immune post transplant load factors believed to play a role in the development of CAN include drug toxicity, including CNI nephrotoxicity, donor – recipient size disparity, hypertension, hyperlipidemia, proteinuria, infection, life style and patient disease recurrence.

The CNI’s CsA and Tacrolimus cause damage by constricting the afferent arterioles leading to localised ischaemia and eventually to glomerular collapse.  On the other hand, these CNI’s are beneficial in that they protect against the effects of immune injury. The contribution of these CNI to the development of CAN is therefore a function of the balance between their toxicity and their immunosuppressive capabilities. To reduce toxicity, CNI’s can be replaced with alternatives such as Mycophenolate mofetil (MMF).

Differences in donor – recipient size have been shown to have a small effect on development of CAN with very large recipients showing significantly reduced graft survival in multivariate studies. Female recipients of male kidneys have been shown to have slightly better graft survival than sex matched transplants. This effect is believed to be due to ‘nephron dose’ with low numbers of available nephrons leading to hyperfiltration in the remaining nephrons to meet the excess load, eventually leading to nephron exhaustion. Hypertension, which is very common in kidney patients post transplant, also leads to hyperfiltration causing glomerulosclerosis and is significantly associated with graft failure. The presence of pre transplant hypertension in the patient and/or donor, of delayed graft function at the time of transplant and the use of calcineurin inhibitors (CNI) all contribute to the development of post transplant hypertension. Similarly, hyperlipidemia is very common in post transplant patients and is significantly associated with development of CAN. It is contributed to by the presence of pre transplant hyperlipidemia, by rapid post transplant weight gain and by the presence of diabetes. Development of post transplant diabetes is a significant risk factor for graft failure.

Proteinuria (leaking of protein from the blood into urine) may contribute to CAN through the toxicity of filtered protein to the tubules.

Infection plays a significant role in the development of CAN firstly because it may lead to the altering or otherwise lowering of Immunosuppression therefore increasing the immune response but also because infection can also directly cause CAN. Renal transplant patients who develop opportunistic infections after 6 months usually have higher serum creatinine levels. The polyoma virus, BK, which is highly prevalent in the population but normally asymptomatic in immunocompetent persons, can become reactivated in immunocompromised patients and lead to allograft nephropathy through viral replication in the tubules, leading to apoptosis or necrosis of tubular epithelial cells. Kidneys from CMV positive donors have been shown to be associated with a small but significant reduction in allograft survival though it is not clear this is directly due to the development of CAN. The role of CMV reactivation may well be limited to the effect this has on managing immunosuppression levels and therefore the knock on effect that potentially has on immune reactivity.

Lifestyle choices, particularly non compliance with immunosuppressive regimes and smoking, can contribute significantly to the development of CAN. Studies have shown that smokers are more likely to require allograft biopsy one year post transplant than non smokers. One study found that in histological analysis, the amount of smoking correlated with the severity of vascular fibrosis and intimal thickening observed. The effect of donor smoking on the development of CAN is unclear.

Finally, CAN may be caused by patient disease recurrence. Renal transplantation does not after all cure the underlying disease. Patients whose primary disease was Glomerulonephritis are at risk. The contribution of disease recurrence to CAN looks set to increase, particularly as improved immunosuppression increases the half life of transplanted kidneys.

The current definition of chronic allograft nephropathy (CAN) is that it is the common end point of a series of immune and non immune insults to the kidney resulting ultimately in graft failure. It is contributed to by donor, patient and transplant process input factors, by immune factors involving acute and chronic humoral and cellular processes and by load factors post-transplant including drug toxicity and infection.