A 2018 Reference Guide to the Banff Classification of Renal Allograft Pathology

Abstract

The Banff Classification of Allograft Pathology is an international consensus classification for the reporting of biopsies from solid organ transplants. Since its initial conception in 1991 for renal transplants, it has undergone review every 2 years, with attendant updated publications. The rapid expansion of knowledge in the field has led to numerous revisions of the classification. The resultant dispersal of relevant content makes it difficult for novices and experienced pathologists to faithfully apply the classification in routine diagnostic work and in clinical trials. This review shall provide a complete and simple illustrated reference guide of the Banff Classification of Kidney Allograft Pathology based on all publications including the 2017 update. It is intended as a concise desktop reference for pathologists and clinicians, providing definitions, Banff Lesion Scores and Banff Diagnostic Categories. An online website reference guide hosted by the Banff Foundation for Allograft Pathology (www.banfffoundation.org) is being developed, which will be updated with future refinement of the Banff Classification from 2019 onward.

FIGURE 1

The content of the Banff Classification of Kidney Allograft Pathology can be inventoried as Banff Lesion Scores and Additional Diagnostic Parameters required by the algorithms behind the Banff Diagnostic Categories to reach a diagnosis. Moreover, overarching definitions are important and inform, for example, how one or even several Banff Lesion Scores are applied. TMA, thrombotic microangiopathy.

FIGURE 2

Banff Lesion Score i (interstitial Inflammation in nonscarred areas of the cortex). A, Interstitial inflammation in nonscarred areas of the cortex. This Banff Lesion Score, often a marker of TCMR, ranges from 0 to 3, based on the percentage of nonscarred cortex involved, and is usually dominated by mononuclear cells in the case of Acute TCMR. Note the contrast between the noninfiltrated interstitium in the right half of the micrograph and the infiltrate in the edema between the tubules on the left (long arrow). PAS, original magnification, ×400. B, An example of plasma cell rich interstitial inflammation. If the infiltrate comprises more than 5% to 10% of either eosinophils, neutrophils or plasma cells an asterisk is added to the Banff Lesion Score i (eg, i1*). H&E, hematoxylin and eosin, original magnification, ×400.

FIGURE 3

Banff Lesion Score t (tubulitis) in nonatrophic or mildly atrophic tubules. These images display various degrees of tubulitis which is characterized by the presence of mononuclear cells on the basolateral aspect of the tubular epithelial cells, within the confines of the basement membrane. Mononuclear cells (long and short arrows) are noticeable by their characteristic halo and smaller nucleus and more condensed chromatin compared to tubular epithelial cells. A, Banff Lesion Score t0—Cortical tubules without tubulitis which would be scored as t0. H&E, original magnification, ×200. B, Banff Lesion Score t1—defined as foci of 1-4 mononuclear cells (arrows) per tubular cross section or per 10 tubular epithelial cells. PAS, original magnification, ×400. C, Banff Lesion Score t2—defined as 5 to 10 mononuclear cells per tubular cross section or per 10 epithelial cells (long arrows). Note that the tubule to the left displays mild tubulitis (short arrows), but the most severely affected tubule dictates the score. PAS, original magnification, ×400. D, Banff Lesion Score t3—defined as foci with >10 mononuclear cells/tubular cross section. Note that for this particular tubule the denominator is per 10 tubular epithelial cells as this tubule is sectioned longitudinally. PAS, original magnification, ×400.

FIGURE 4

Banff Lesion Score t (tubulitis) in moderately atrophic tubules. In biopsies with Banff Lesion Scores i, ti and i-IFTA sufficient for a diagnosis of acute TCMR grade IA, IB or chronic active TCMR grade IA and IB, Banff Lesion Score t must also be scored in moderately atrophic cortical tubules. Moderately atrophic tubules are defined as having less than 50% down to 25% of the diameter of the surrounding “unaffected or minimally affected [cortical] tubules in the biopsy”. This example shows such unaffected or minimally affected tubules with their diameter marked in black. Their mean diameter in this image would be around 48 μm. The tubule with the diameter marked in gray has a diameter of 27 μm which is more than 50% of 48 μm. Thus, this tubule would still qualify as mildly atrophic. It is heavily infiltrated with mononuclear cells (gray arrows). In contrast, the tubule with the diameter of 20 μm marked in red is moderately atrophic. The mononuclear tubulitis in this particular tubule must be scored toward Banff Lesion Score t in this biopsy which was diagnosed as Acute TCMR Grade IB. PAS, original magnification, ×400.

FIGURE 5

Banff Lesion Score v (intimal arteritis). These photomicrographs demonstrate intimal arteritis, characterized by the presence of inflammatory cells beneath the lining endothelial cells. A, Banff Lesion Score v1—mild to moderate arteritis with mononuclear cells (long arrows) immediately beneath lifted endothelial cells (short arrow). H&E, original magnification, ×200. B, Banff Lesion Score v2—severe intimal arteritis involving over 25% of the arterial lumen with mononuclear cells (long arrows) immediately beneath lifted endothelial cells (short arrow). H&E, original magnification, ×200; C, Banff Lesion Score v3 -Transmural arteritis with fibrinoid necrosis in the media (long arrow) and mononuclear infiltrate in the arterial wall (short arrows). Intimal arteritis can be seen in both Acute TCMR Grade II and III and Active AMR. The most severely affected artery determines the score. Masson trichrome, original magnification, ×100. D, This image demonstrates an area of interstitial hemorrhage characterized by extravasation of red blood cells into the surrounding interstitium (arrow). Although there is not a specific Banff Lesion Score for this feature, it can be recorded by attaching an asterisk to the v score (eg, v*). Note that this asterisk attached to Banff Lesion Score v is not specific for interstitial hemorrhage as an area of cortical infarct (not shown) would also be coded like this. H&E, original magnification, ×400.

FIGURE 6

Banff Lesion Score g (glomerulitis). Glomerulitis is a form of MVI and a feature of AMR activity. A, Segmental glomerulitis; PAS, original magnification, ×400. B, Global glomerulitis. Note the characteristic complete or partial occlusion of capillary loops by leukocytes (short arrows) and endothelial cell swelling (long arrows). The score of g0 to g3 is determined by the percentage of glomeruli involved with either segmental or global glomerulitis. Complete or partial occlusion of a single capillary loop suffices to mark the respective glomerulus as involved by the glomerulitis. PAS, original magnification, ×400.

FIGURE 7

Banff Lesion Score ptc (peritubular capillaritis). Peritubular capillaritis is a form of MVI and a feature of AMR activity. Each image demonstrates the various ptc scores which are in themselves determined by the number of inflammatory cells present within capillary lumina. A, Banff Lesion Score ptc1—Mild peritubular capillaritis defined as at least 1 cell in ≥10% of cortical PTCs (short arrows) with 3 to 4 in the most severely involved PTC (long arrow). Please note the slightly distended, open appearance of the capillary which can be a helpful feature; PAS, original magnification, ×400. B, Banff Lesion Score ptc2—Moderate peritubular capillaritis defined as at least 1 cell in ≥10% of cortical PTCs (short arrows) with 5-10 in most severely involved PTC (long arrow); PAS, original magnification, ×400. C, Banff Lesion Score ptc3—severe peritubular capillaritis defined as at least 1 cell in ≥10% of cortical PTCs (short arrows) with >10 in most severely involved PTC (long arrow). PAS, original magnification, ×400. D, This peritubular capillary is cut longitudinally (short arrow) and although containing 4 mononuclear cells is to be disregarded for scoring. However, the neighboring peritubular capillary (long arrow) is cut orthogonally and would qualify for Banff Lesion Score ptc1 provided that at least 10% of all PTCs contain at least 1 leukocyte. PAS, original magnification, ×400.

FIGURE 8

Banff Lesion Score C4d. A, IHC staining with peroxidase yielding a brown reaction product for C4d. An example of C4d3, this image demonstrates linear and circumferential staining of endothelial cells in virtually all peritubular capillaries. The staining was similar in all areas of the cortex and the medulla. The proportion of stained peritubular capillaries and medullary vasa recta informs the score. B, IF staining for C4d. This image shows an example of a Banff Lesion Score of C4d3; using IF, a minimum score of C4d ≥ 2 is considered positive. In addition to this, the staining intensity for an individual capillary or medullary vas rectum must be at least 1+ on the usual scale from negative, trace, 1+, 2+ to 3+. Indirect IF, mouse antihuman C4d followed by fluorescein isothiocyanate-conjugated antimouse IgG, original magnification, ×100.

FIGURE 9

Banff Lesion Scores for ct (tubular atrophy) and ci (interstitial fibrosis). The ci and ct scores are both based on calculating the total percentage of cortex involved and require a diligent assessment of all foci of ct and ci as this process is often multifocal; ct and ci scores may not always be equally advanced. A, This image demonstrates an area of nonatrophic tubules (long arrow), compared to an area of tubular atrophy (short arrow) without an obvious increase in interstitial fibrosis. PAS, original magnification, ×200. There are different morphological types of tubular atrophy with differing histological appearances, including conventional, thyroidization, and endocrine-like types. B, Tubular atrophy of conventional type with interstitial fibrosis. Tubular areas are separated by areas of interstitial fibrosis and tubules show thickened basement membranes and > 50% reduction in tubular diameter (long arrows). PAS, original magnification ×200. C, Thyroidization type atrophy. Here, tubules appear dilated, have flattened epithelial cells, and contain eosinophilic and brightly periodic-acid-Schiff-positive uromodulin casts (long arrow). PAS, original magnification, ×200. D, endocrine-like type, characterized by shrunken tubules with cuboidal epithelium and “tubular simplification” (long arrow). Compared with the other types of tubular atrophy, endocrine-like type does not have thickened basement membranes but still counts toward the ct score. PAS, original magnification, ×400.

FIGURE 10

Visual analogue scales provided by the Banff Working Group on Fibrosis. This working group developed schematic diagrams to facilitate and standardize scoring of Banff Lesion Scores ci and ct. A, Scale for the assessment of interstitial fibrosis without tubular atrophy. B, Scale for the assessment of diffuse tubular atrophy with “replacement fibrosis.” Reproduced with kind permission from American Journal of Transplantation.

FIGURE 11

More visual analogue scales provided by the Banff Working Group on Fibrosis. A, Scale for the assessment of patchy (left) and confluent (right) interstitial fibrosis without glomeruli. B, Scale for patchy (left) and confluent (right) fibrosis with glomeruli. Reproduced with kind permission from American Journal of Transplantation.

FIGURE 12

Banff Lesion Score cv (vascular fibrous intimal thickening). A, Banff Lesion Score cv1—very mild purely fibrous thickening of the arterial intima (arrow). PAS, original magnification, ×200. B, Purely fibrous intimal thickening is depicted here in between the arrows in a trichrome stain. Note that this type of fibrous intimal thickening can also represent chronic damage in AMR. Masson trichrome, original magnification ×400. C, Arterial fibrous intimal thickening in between the arrows. Note the multiplication of the internal elastic lamina. Trichrome-elastica, original magnification, ×400. D, Severe fibrointimal thickening cv3, with mononuclear infiltrates (long arrow) and foam cells (short arrow) in the fibrotic intima which can be a feature of both Chronic Active TCMR and Chronic Active AMR. Both types of lesion qualify for Banff Lesion Score cv, the score is determined by the loss of luminal area as shown in Figure 13 below. H&E, original magnification, ×100.

FIGURE 13

Visual analogue scale for the determination of Banff Lesion Score cv (arterial fibrous intimal thickening). The remaining luminal area is related to the square of the remaining luminal radius. Thus, relatively modest decreases in luminal radius of 13% or 29% translate into relatively large reductions in luminal area of 25% or 50%, reflecting the thresholds for Banff Lesion Score cv.

FIGURE 14

Banff Lesion Score cg (GBM double contours). This score represents the presence and extent of GBM double contours, a criterion for Chronic Active AMR. The score ranges from 0 to 3 and is based on the percentage of capillary loops with double contours as evident on EM (Banff Lesion Score cg1a) or LM (cg1b to cg3) in the most severely affected glomerulus. A, cg1a—GBM with double contours (short black arrows point to areas of original basement membrane and red arrows point to areas of new basement membrane formation), visible by EM only. Double contours, such as those noted in this image must be accompanied by endothelial cell swelling (long black arrow) and/or subendothelial rarefaction, and must involve at least 3 glomerular capillaries by EM for a score of cg1a. Scores of greater than cg1a are based on light microscopic appearance which can best be examined by silver stains. Transmission EM, original magnification ×8000. B, Banff Lesion Score cg1b—double contours (arrow) identified on LM which involve up to 25% of the capillary loops of this most affected glomerulus. Jones silver stain, original magnification, ×400. C, Banff Lesion Score cg2—double contours (arrows) present in 26-50% of this most affected glomerulus; Jones silver stain, original magnification, ×400. D, Banff Lesion Score cg3—double contours (arrows) present in >50% of this most affected glomerulus. Jones silver stain, original magnification, ×400.

FIGURE 15

Banff Lesion Score mm (mesangial matrix expansion). This glomerulus fulfils the criteria for moderate mesangial matrix expansion with more than 2 mesangial cells in these 2 adjacent glomerular lobules (arrows). The proportion of glomeruli with such mesangial matrix expansion among all nonsclerosed glomeruli informs the score. The underlying reason for the mesangial matrix expansion in this biopsy was recurrent IgA glomerulonephritis revealed by IHC and EM. PAS, original magnification, ×400.

FIGURE 16

Banff Lesion Score ah (arteriolar hyalinosis). A, Banff Lesion Score ah1—mild focal arteriolar hyalinosis (arrow). PAS, original magnification, ×630. B, ah2—Moderate arteriolar hyalinosis (arrow). PAS, original magnification, ×630. C, Banff Lesion Score ah2—Note in this image there is both linear (short arrow) and nodular hyalinosis (long arrow). For a score of ah2, more than 1 arteriole displaying moderate to severe is required. Jones silver stain, original magnification ×630. D, Banff Lesion Score ah3—severe circumferential arteriolar hyalinosis with luminal occlusion. For Banff Lesion Score ah3, hyalinosis of this severity (arrow) must be present in many arterioles as depicted here. PAS, original magnification, ×630.

FIGURE 17

Banff Lesions Score i-IFTA (Inflammation in areas of IFTA). Image A shows Inflammation in areas of IFTA (arrow). This lesion score ranges from 0 to 3, based on the percentage of scarred areas of the cortex (ie, areas qualifying for ci and ct) involved by inflammation. It is one of the criteria necessary for a diagnosis of chronic active TCMR Grade IA or IB. Masson trichrome, original magnification ×200. B, In contrast shows interstitial fibrosis without significant infiltrate (arrow). H&E, original magnification, ×400.

FIGURE 18

Acute TMA. A, An acute TMA affecting a glomerulus with fibrin thrombi (long arrows) and fragmented red blood cells (short arrow) in capillary loops. Trichrome, original magnification ×400. B, An acute TMA affecting a small arteriole (arrow). Acute TMA is one of the histological features used as histological evidence of acute tissue injury in Active AMR. However, TMA is not specific for AMR and can be seen in, for example, recurrent disease or calcineurin inhibitor toxicity. Trichrome, original magnification, ×400.

FIGURE 19

Severe Peritubular Capillary Basement Membrane Multilayering (PTCML) as demonstrated by EM. A, This Additional Diagnostic Parameter is a criterion for AMR chronicity. It is defined as 7 or more layers of basement membrane in at least a single cortical peritubular capillary and 5 or more in at least 2 additional capillaries. This particular capillary shows 8 layers (arrow). Transmission EM, original magnification ×14 000. B, This image demonstrates a peritubular capillary with 5 layers of basement membrane (arrow). Transmission EM, original magnification, ×10 000.

FIGURE 20

Banff Classification Diagnostic Category 6 (other). These images illustrate some of the more common examples of key lesions specified under category 6. A, Pyelonephritis with neutrophilic casts (arrow) and neutrophilic infiltrates with tubulitis. H&E, original magnification, ×200. B, BK virus nephropathy with typical ground glass intranuclear inclusions as seen on hematoxylin and eosin stain (arrows). H&E, original magnification ×400. C, Acute tubular injury with widespread isometric vacuolization of tubular epithelial cells (arrow) associated with acute Calcineurin Inhibitor Toxicity and other forms of injury. H&E, original magnification, ×200. D, Recurrent glomerulonephritis (membranoproliferative immune complex glomerulonephritis type I in this case) with split GBMs (arrow). The diagnosis was confirmed and TG excluded by positive IF for immunoglobulin heavy-, light-chains and complement slit products as well as abundant subendothelial electron dense immune complex deposits on EM. PAS, original magnification, ×400.