In Press, 1995, Transplantation Proceedings copyright by Appleton and Lange.

Report of the Third Banff Conference on Allograft Pathology (July 20-24, 1995) on Classification and Lesion Scoring in Renal Allograft Pathology*

Kim Solez (Edmonton), Hallgrimur Benediktsson (Calgary), Tito Cavallo (Cincinati), Byron Croker (Gainesville), Anthony J. Demetris (Pittsburgh), Cinthia Drachenberg (Baltimore), Steven Emancipator (Cleveland), Peter N. Furness (Leicester), Lilian W. Gaber (Memphis), Ian W. Gibson (Edmonton), James Gough (Winnipeg), Rheka Gupta (Halifax), Philip Halloran (Edmonton), Pekka Häyry (Helsinki), Michael Kashgarian (New Haven), Niels Marcussen (Åarhus), Ziad A Massy (Minneapolis), Michael J. Mihatsch (Basel), Kunio Morozumi (Nagoya), Irene Noronha (Sao Paulo), Steen Olsen (Åarhus), John Papadimitriou (Baltimore), Leen C. Paul (Toronto), Maria Picken (Maywood), Lorraine C. Racusen (Baltimore), Eleanor L. Ramos (Wallingford), Parmjeet Randhawa (Pittsburgh), David C. Rayner (Edmonton), David Rush (Winnipeg), Fred Sanfilippo (Baltimore), Eero Taskinen (Helsinki), Kiril Trpkov (Edmonton), Luan Truong (Houston), Yutaka Yamaguchi (Tokyo), and Serdar Yilmaz (Helsinki)

*Footnote: The other Banff Conference papers in this collection available at this site, were written before the meeting and presented in Banff. This paper was written by the Banff meeting participants and is intended to reflect the decisions made there concerning classification and lesion scoring in renal allograft pathology.


The Third Banff Conference on Allograft Pathology was held July 20-24, 1995 in Banff, Alberta, Canada. Participants expressed general satisfaction with the current version of the Banff schema for renal allograft pathology [1] but considered the potential need for changes in several areas. It was agreed that a number of studies of individual lesions in the Banff schema should be undertaken with the option of making alterations in 1997. However, no changes should be made unless supported by well conducted studies from two different investigative groups.


Future Directions

It is recommended that data be accumulated on three of the lesions Racusen et al. suggested designating by an asterisk (i*) in Banff lesion coding [2]: interstitial eosinophils (eo), plasma cells (pl), or polymorphs (pm), and on three lesions suggested as being diagnostically important by Colvin [3]: interstitial edema (ed), lymphocyte activation - blast transformation (bt), and tubular injury (ti). [Tubular injury here refers to tubular cell loss, regenerative changes (including mitoses), or necrosis, and not to tubular changes directly attributable to tubulitis (the "vanishing tubules" in paragraph below)]. The presence or absence of atrophy as defined by Mihatsch et al [4] in tubules with tubulitis should also be documented (definition: tubular basement membrane thickening or a reduction in diameter of 50% compared to surrounding tubules). A hyphen can be placed between the t score and the ct score when the only tubulitis was present in atrophic tubules.

In discussion and correspondence following the Banff meeting, we have agreed to allow the presence of two or more heavily inflamed "vanishing tubules" ("tubules gone underground") [2, 5] with partial loss of basement membrane as a sufficient criterion for a t3 grading and diagnosis of moderate acute rejection (grade 2A), providing that: 1) tubulitis of at least t2 severity is present elsewhere in the biopsy, 2) the general appearances of the tubules involved are consistent with the basic concept embodied in the t3 score that there are more lymphocytes invading a tubule than the number of tubular epithelial cells present within it, and 3) caution is exercised in the interpretation of changes in the immediate subcapsular area. This treatment of the vanishing tubules, a lesion familiar in native kidney pathology [5] and well described by Racusen [6], is consistent with the way many centers are using the Banff classification already and will make assessment of biopsies more rapid and straightforward in many cases. Interstitial infiltrates of i2 or i3 severity would still be required for the diagnosis of rejection on the basis of tubulitis [1].

Both Marcussen et al. [7, 8] and Mihatsch et al. [4] have documented poor reproducibility in assessment of hyaline arteriolar change, and yet this is the best marker for cyclosporine (CsA) nephrotoxicity. A morphometric study of hyaline arteriolar change is needed to determine the number of vessels which must be examined for a meaningful result. Strom, Epper and Mihatsch have suggested that a minimum of ten glomeruli should be present to ensure an adequate sampling of arterioles [9]. The occurrence of specific nodular external arteriolar lesions (which could be coded ah*) as well as nonspecific subendothelial hyaline lesions should be documented [10, 11].

It would be very useful to document the speed of onset and regression of cyclosporine-induced hyaline arteriolar change, as an extension of the work of Morozumi et al. [12]. Baseline biopsies must be emphasized as crucial for interpretation of chronic rejection changes as well as cyclosporine toxicity, since both fibrous intimal thickening (and other chronic rejection changes) and hyaline arteriolar change may preexist in the donor. However hyaline arteriolar change in implantation biopsies is not uncommon, and should not preclude the diagnosis of cyclosporine toxicity in later biopsies.

Further immunohistochemical studies of glomerulitis with monocyte and lymphocyte markers [13-15] are recommended to increase reproducibility and knowledge of this lesion, which may occur both in cyclosporine nephrotoxicity and in so-called antibody-mediated rejection. Further studies of transplant thrombotic microangiopathy (many causes - CsA, FK506, OKT 3, antibody-mediated rejection, HIV infection, and [rarely] perfusion injury) as an entity related to both glomerulitis and hyaline arteriolar change should be encouraged. It would be desirable to establish a registry of cases of transplant thrombotic microangiopathy, possibly under the jurisdiction of the Renal Panel of the AFIP Center for Advanced Pathology. Such a registry would also be useful for post-transplant lymphoproliferative disease (see below)

The need for standardized clinical definitions of rejection and rejection reversal is strongly emphasized. These definitions should include not only the degree of functional change but also its time course and duration. In addition, the important entity of late-appearing acute rejection also needs to be considered [16-18].

There is strong support for making a clear distinction between Grade 2A moderate acute rejection (defined by tubulitis) and Grade 2B moderate acute rejection (defined by arteritis) as it is possible that they have different pathogenesis and prognosis [7, 19]. Severe (Grade 3) rejection may consists of a mix of antibody-mediated and cell mediated rejection and it is recommended that an attempt be made to find means to differentiate these two conditions [20-22], possibly by indicating the presence of transmural inflammation (tm) or frank fibrinoid change (vfc). The severity of change in the most severely affected artery should be given precedence over number of arteries affected in assigning the v score.

Biopsies before treatment of acute rejection should be strongly encouraged, as this makes interpretation much more secure. On the other hand there is a need for investigation of the temporal sequence of lesion regression after treatment. Kolbeck has written "... many (if not most) renal allograft biopsies are now being performed ... following antirejection therapy to evaluate the extent of ongoing cellular rejection" [23]. One hopes that this is not generally true, as such an approach greatly underrates the value of the renal biopsy in therapeutic decision-making. Nevertheless much more study is needed of post-treatment biopsies to provide a knowledge base for those centers in which such biopsies are the norm. Newer clinical trail protocols which require both a pre-treatment and post-treatment biopsy should shed light on this area.

In typical cases of rejection no immunohistochemical studies are needed. However for the difficult, equivocal case it would be useful to have a protocol of additional adjunctive tests including CD45 isotypes [24, 25]. Additional studies of the discriminative value of CD45RO and CD45RA staining are to be encouraged to see if the diagnostic significance of these antigens in kidney (liver and heart) allografts can be confirmed [24-27].

The accurate diagnosis of post-transplant lymphoproliferative disorder [28, 29] is also a very important issue. It is worth emphasizing that this condition may coexist with rejection. Future studies to allow more accurate initial provisional identification of this entity by ordinary light microscopy are in order and a registry of cases of this entity would be a valuable educational resource. A recommended protocol for in-situ hybridization studies as final confirmation of the diagnosis would also be useful.

Chronic rejection connotes a clinico-pathologic entity which usually presents with slowly declining function and increasing proteinuria and whose pathologic findings shows characteristic glomerular and vascular lesions with variable tubulointerstitial involvement in the form of tubular atrophy and interstitial fibrosis [30-34]. Whereas the glomerular and vascular changes are characteristic of chronic rejection, the tubulointerstitial changes which frequently accompany chronic rejection, are not [1]. Also, it is recognized that the vascular and glomerular lesions are not pathognomonic and, owing in part to sampling error, are not always found in clinically straight-forward cases of chronic rejection. In the original Banff classification paper [1] the term chronic allograft nephropathy rather than chronic rejection was used to acknowledge the lack of specificity of the changes for chronic dysfunction. Severity of the process was graded by the severity of interstitial fibrosis and tubular atrophy since these are the variables which can be evaluated in cell biopsy samples. In 1991 we could not take into account the significance of the Chronic Allograft Damage Index (CADI) as the papers on this pathologic scoring system were just being written [35, 36]. Four years later, however, the significance of CADI as a predictor of later chronic renal allograft failure is clear [37-40]. Therefore we have defined the chronic lesions on the Banff scoring system in a way which is harmonized with the CADI scoring system.

Paul and Zaltzman [41] have emphasized that the diagnosis of chronic rejection must incorporate both clinical and pathological data. Neither is sufficient in itself because at present no findings are absolutely specific for this disorder. When only clinical data are available, it may be preferable to use the term "chronic dysfunction" rather than "chronic rejection". On the other hand, for clinical trial endpoints, the use of pathology data alone has great advantages since morphologic differences may emerge years before renal dysfunction could be documented [36-40]. In this respect it is important to distinguish between a diagnosis of presumed chronic rejection on the basis of vascular and/or glomerular lesions, and chronic allograft nephropathy diagnosed on the basis of less specific lesions [1]. Some of the new antirejection agents may be found to have specific effects on the vascular or glomerular lesions of chronic rejection and it is important not to obscure such an effect by relying on chronic allograft nephropathy [1], or on an aggregate scoring system like CADI as the only endpoint [38, 40].

Electron microscopy documenting the characteristic peritubular capillary basement membrane splitting and lamination may be necessary for the definitive diagnosis of chronic rejection [42, 43]. However, pending confirmation of the specificity of these ultrastructural changes, a system of presumptive diagnosis of chronic rejection based on new onset of characteristic light microscopic vascular and glomerular changes unexplained by other mechanisms will be established.

Not all cases of presumed chronic rejection will turn out to be chronic rejection after other causes for the changes seen are explored and immunofluorescence and electron microscopic data are added. Either new onset vascular or glomerular lesions are sufficient for the diagnosis of presumed chronic rejection. Both are not required. Therefore it is not possible at the moment to describe a uniform means of grading severity of presumed chronic rejection, other than describing the severity of the vascular or glomerular change that led to the diagnosis (cv or cg).

For this reason severity of chronic allograft nephropathy should also still be stated on the basis of interstitial fibrosis and tubular atrophy [1]. While less specific, determination of chronic allograft nephropathy provides a uniform basis of grading chronic lesions that can be applied to all specimens, as can the CADI score [38].

Since transplant glomerulopathy may be confused with other lesions by light microscopy, from an operational point of view new onset fibrous intimal thickening should b present for an unequivocal diagnosis of chronic rejection by light microscopy alone.


Click Here to Continue.

Return to Banff Presentations


Last Modified: April 03, 1996 1:44:45 PM