Main

Transplant glomerulopathy is an important cause of proteinuria, rising creatinine, hypertension, and shortened allograft survival. Transplant glomerulopathy is defined as reduplication/multilammination of the glomerular basement membrane as observed by light (Figure 1) and/or electron microscopy (Figure 2) in a kidney allograft in the absence of immune deposits. By this definition, transplant glomerulopathy is a morphologic description of histologic or ultrastructural alterations and not a specific clinicopathologic entity.1 Multiple pathophysiologic mechanisms result in development of this lesion, all related to chronic, repeated endothelial cell injury. Most notable is antibody-mediated rejection. Other reported causes include autoantibodies, cell-mediated rejection, thrombotic microangiopathy, and hepatitis C virus.

Figure 1
figure 1

Transplant glomerulopathy—light microscopy. (a) The glomerulus has patent capillary loops with thickened basement membranes. The refractile appearance of the basement membranes suggests duplication. Epithelial and endothelial cells are swollen. The mesangial regions are mildly expanded by an increase in cells and matrix. Hematoxylin and eosin (H&E), × 40. (b) Silver stain on the same glomerulus shows patent capillary loops with thickened basement membrane with duplication in ≥10% of the patent loops. Jones methenamine silver, × 40. (c) The periodic acid Schiff (PAS) stain on the glomerulus demonstrates capillary basement membrane duplication and widening of the subendothelial compartment. The mesangial regions show an increase in matrix. PAS, × 40.

Full size image
Figure 2
figure 2

Transplant glomerulopathy—electron microscopy. (a) Transplant glomerulopathy with widening of the subendothelial space as a result of edema and amorphous electron-dense material (b) Transplant glomerulopathy widening subendothelial space as a result of cell debris, fragmented red blood cells (consistent with a thrombotic microangiopathy), a macrophage, and amorphous proteinaceous debris. (c) Transplant glomerulopathy with prominent reduplication of the basement membrane. Original magnification, × 4000.

Full size image

The development of glomerular disease in a transplant, manifested by proteinuria and elevated serum creatinine, has been well described for over 50 years.2, 3 The term transplant glomerulopathy was first used in the setting of apparent rejection (called rejection transplant glomerulopathy) by Busch et al4 in 1971. This early report stressed repetitive endothelial cell injury as a prime pathogenic mechanism. Zollinger et al5 in 1973 described in detail the lesion they termed 'transplant glomerulopathy,' and they attempted to differentiate it from recurrent or de novo inflammatory glomerulonephritides. The characteristic changes in the lamina rara interna of the glomerular basement membrane were stressed. In 1985 Maryniak et al6 described the evolution of this lesion over time. The descriptions and pictures presented in these early publications represented the current pathologic criteria for diagnosing transplant glomerulopathy.

In this review, we describe in detail the morphologic features of transplant glomerulopathy found on light, immunofluorescence, and electron microscopies. The pathophysiology of the causes and risk factors are discussed. Clinical manifestations are emphasized and potential therapeutic modalities are examined.

Light microscopy

The hallmark of transplant glomerulopathy is duplication of the glomerular basement membrane. To diagnose transplant glomerulopathy, the Banff ’97 criteria required ≥10% of capillary loops in the most affected glomerulus to evidence double contours.7 Banff 2015 (Table 1) now requires double contours in only a single capillary loop as the minimal light microscopic finding (cg1b).8, 9 An increase in mesangial matrix may be present but is less specific and is not required for diagnosis. To satisfy the Banff 2015 criterion for mesangial matrix increase ('mm'), the expansion must be moderate, defined as mesangial matrix between adjacent capillaries exceeding the width of two mesangial cells in a minimum of two lobules. Mesangiolysis may be present, as well as glomerulosclerosis, the latter mimicking focal segmental glomerulosclerosis. Fibrin deposition can occur.6 Microvascular inflammation, including glomerulitis (g) and/or peritubular capillaritis (ptc) (Figure 3), may be observed, especially in cases mediated by an alloimmune mechanism.

Table 1 Transplant glomerulopathy per 2015 Banff (Adapted from Am J Transplantation, 2017;17: 28–41)
Full size table
Figure 3
figure 3

Peritubular capillaritis (ptc). The peritubular capillaries are dilated and contain lymphocytes. Ptc2. Hematoxylin and eosin (H&E), × 40.

Full size image

Immunoflourescence microscopy

By definition, transplant glomerulopathy is distinguishable from the recurrent and de novo glomerulonephritides. Immunoflourescence for IgG, IgA, IgM, C1q, C3, κ, and λ is negative, with the exception of nonspecific staining for IgM that probably represents physiologic catabolism of IgM-containing immune complexes normally formed in vivo.10

C4d

Critical to evaluating kidney transplant biopsies is determining the endothelial staining of C4d involving peritubular capillaries. Whereas tubular basement membranes and arterioles may occasionally stain positive for C4d, such staining is considered nonspecific and not part of current diagnostic criteria.

The C4 component of complement is activated following C1q activation by antigen–antibody complexes or may be directly activated by polysaccharides through the mannose-binding lectin pathway. In either case, C4d, a C4 split product without known function, can be covalently bound to tissues, and is assumed to represent a 'footprint' of prior antibody activity.11 C4d positivity may remain after initiating factors (such as HLA antibodies) have resolved or are removed.11 It is now standard practice to stain all kidney transplant biopsies for C4d to determine humoral alloreactivity with complement fixing antibodies. C4d evaluation can be accomplished by direct immunofluorescence microscopy on frozen tissue or by immunohistochemistry on fixed tissue (Figure 4). The immunoflourescence technique is reported to be more sensitive.12 There is poor interobserver and interlaboratory reproducibility with immunohistochemistry resulting in an interinstitutional κ score of only 0.17.13

Figure 4
figure 4

Immunohistochemical staining for C4d. The peritubular capillaries show diffuse positivity for C4d deposition by immunohistochemical staining. The staining is moderate to strong with a circumferential pattern. Greater than 50% of the peritubular capillaries were affected (C4d3). Original magnification × 40

Full size image

As not all donor-specific antibodies are pathogenic, their presence alone is insufficient for diagnosis of antibody-mediated rejection. Evidence of repeated endothelial cell–antibody interaction is required by Banff criteria to diagnose either acute antibody- or chronic active antibody-mediated rejections. Peritubular capillary C4d positivity satisfies this requirement. The one major exception to this rule is C4d positivity in ABO-incompatible kidney transplantation, where C4d positivity on the order of 80–90% is reported, usually without histologic evidence of rejection.14, 15 In this setting, C4d staining has been termed 'accommodation' and is not considered pathologic in the absence of HLA donor-specific antibodies. Diffuse peritubular capillary C4d positivity without concurrent rejection was even associated with reduced scarring on subsequent biopsies in one study of 33 ABO-incompatible transplants.16 For ABO-incompatible allografts, other evidence of antibody interaction with the endothelium is required to diagnose antibody-mediated rejection. In all other situations, when transplant glomerulopathy coexists with both donor-specific antibodies against HLA antigens and positive peritubular capillary C4d staining, a definite diagnosis of chronic active antibody-mediated rejection can be made as the cause of the transplant glomerulopathy by Banff criteria.

It is well described that C4d positivity can be evanescent, with successive biopsies changing from C4d-positive to C4d-negative in days to weeks, and vice versa.17, 18 Both C4d-negative acute antibody-mediated rejection and chronic active antibody-mediated rejection are well described. With donor-specific antibody-positive and C4d-negative, chronic active antibody-mediated rejection can be diagnosed by Banff 2015 if either g+ptc≥2 (with requirement for g≥0 when associated borderline or acute cell-mediated rejection exists) or there is increased expression of gene transcripts indicating endothelial cell injury (known as ENDATS). On the other hand, C4d-positive but HLA donor-specific antibody-negative biopsies with appropriate tissue injury, for example, glomerulitis acutely or transplant glomerulopathy chronically, are considered only suspicious for humoral causation by Banff. Others, however, advocate that microvascular inflammation with C4d positivity can rule in antibody-mediated rejection despite HLA donor-specific antibody negativity, at least in the acute situation.19

By current Banff criteria, either focal (≥10% up to 50% peritubular capillary staining, termed C4d2) or diffuse (≥50% peritubular capillary staining, termed C4d3) by immunoflourescence are considered positive for diagnosing antibody-mediated rejection, while staining of 1–10% peritubular capillaries (C4d1) is not. Separating C4d1 (considered negative) from C4d2 (considered positive) may be technically difficult because of peritubular capillary dropout when significant interstitial fibrosis exists. In this circumstance, double immunofluorescence with C4d and CD34 can facilitate more accurate assessment of the true percent positive.20 By less sensitive immunohistochemistry, anything ≥0% peritubular capillary staining is considered positive. These thresholds for positivity with either immunoflourescence or immunohistochemistry represent a reduction compared with prior Banff publications. A recent single center series demonstrated a doubling of cases diagnosed with chronic active antibody-mediated rejection (36% vs 18%) using the newer criteria for C4d positivity along with the g+ptc≥2 criteria in C4d-negative cases.21 The new criteria significantly predicted the combined endpoint of doubling of serum creatinine or graft loss, whereas prior criteria did not. Of note, this positive association with outcome was significantly related only to the lowered C4d threshold and not to the inflammatory component (g+ptc≥2) of the new criteria.

The percentage of cases of transplant glomerulopathy that are C4d positive varies widely in the literature. Some studies used immunoflourescence, others immunohistochemistry, and some both methods. Specific criteria for positivity varied, as many studies were published before the 2013 and 2015 Banff updates. Regele et al18 found 67% of 58 biopsies to be C4d positive by immunofluorescence and Mauiyyedi et al22 found 61% of 38 cases positive by immunohistochemistry. In contrast, Al Aly et al23 used immunoflourescence and found 0 of 20 patients with transplant glomerulopathy had positive staining. Akalin et al24 studied 36 patients by immunohistochemistry and found only 4 with diffuse and 1 with focal (10–50% peritubular capillaries) staining, although the number with ≥0–10% peritubular capillary positivity, now considered positive by current Banff criteria using immunohistochemistry, was not noted. Sis et al25 used immunoflourescence on 50 biopsies with transplant glomerulopathy and found diffuse (≥50% peritubular capillaries) positivity in 13 and focal positivity in 5 others (1–50% peritubular capillaries, not 10–50% as currently specified by Banff for immunoflourescence). Gloor et al26 used immunoflourescence and found 13 of 51 patients positive without specifying the extent. Hayde et al27 used immunohistochemistry and found 7 of 46 transplant glomerulopathy biopsies to be C4d positive. Overall ~20–25% of biopsies with transplant glomerulopathy are C4d positive. If donor-specific antibodies are detectable, chronic active antibody-mediated rejection is the cause of the transplant glomerulopathy.

Biopsies with transplant glomerulopathy are not uncommonly C4d positive in the absence of detectable HLA donor-specific antibodies in ABO-compatible transplants, which is deemed 'suspicious' for chronic active antibody-mediated rejection. The donor-specific antibodies may exist in the serum, but at a lower level than detectable by current methodology or below an arbitrary cutoff. The antibodies may be of high affinity and totally absorbed on the allograft.28, 29 The antibodies may be directed against donor-recipient loci not previously typed, such as DP. The antibodies may be directed against non-HLA antigens, such as MHC class I chain-related genes A and B (MICA and MICB). The alloantibodies may react with epitopes contained within donor antigens, although not donor specific in terms of antigen.30 Finally, the alloantibody response may wane over time, such that particular donor-specific antibodies are no longer being produced. Theoretically, antibodies may not even be involved at all, with activation of C4 via the lectin pathway, although clinical support for this does not exist to our knowledge.

In the case of chronically failing allografts, both donor-specific antibody-positive/C4d-negative and donor-specific antibody-negative/C4d-positive scenarios are not uncommon, and both may indicate an antibody-mediated cause. In a series of 173 patients with late graft dysfunction, 31 were donor-specific antibody positive/C4d negative and 28 were donor-specific antibody negative/C4d positive as compared with 40 donor-specific antibody positive/C4d positive.31 The results are similar if transplant glomerulopathy is specifically considered. For example, in a case series of 71 patients with transplant glomerulopathy, Lesage et al32 noted that 8 were donor-specific antibody positive/C4d negative and 6 were donor-specific antibody negative/C4d positive, and only 12 were donor-specific antibody positive/C4d positive.32 The long-term outcomes for all three groups were similar and significantly worse than those with both Cd and donor-specific antibody negativity. Likewise, Shimizu et al33 found that 12 of 50 biopsies with transplant glomerulopathy were donor-specific antibody positive/C4d negative and 7 were donor-specific antibody negative/C4d positive, although only 17 of 37 patients received ABO-compatible grafts clouding the issue. In a recent case series of 92 patients with transplant glomerulopathy, 35 were considered only suspicious for chronic active antibody-mediated rejection based on either donor-specific antibody positive/C4d negative or donor-specific antibody negative (or not available)/C4d positive, compared with 34 definite chronic active antibody-mediated rejection.34 In C4d-positive cases without donor-specific antibody (or any HLA antibodies), non-HLA antibodies should be considered. If non-donor-specific HLA antibodies are present, epitope analysis searching for antibodies against donor-specific epitopes by HLA Matchmaker (http:HLAMatchmaker.com) may be indicated.35

Normal human glomeruli will often stain positive for C4d by light microscopy, predominantly in the mesangial regions,10 although there may be subendothelial staining as well.10 Immune complex-mediated glomerulonephridites (either in a native kidney or a kidney allograft) stain positive for C4d in a granular or pseudolinear manner, including membranous nephropathy and immune complex-mediated membranoproliferative glomerulonephritis, where there can be intense capillary wall staining.36 Thrombotic microangiopathy in the native kidney also frequently stains positive for C4d in glomeruli.37 The C3 glomerulopathies appear to stain negative for C4d,36 owing to alternate pathway activation of complement.

Glomerular staining for C4d in kidneys allografts is not uncommonly found, although reports of frequency are quite variable (as low as 10% to as high as 49%). 17, 38, 39 Reports of glomerular C4d staining specifically in transplant glomerulopathy have also been variable, ranging from a frequency of 12%18 to as high as 100%.39, 40, 41 Gasim et al41 recently postulated that glomerular C4d staining indicated structural glomerular basement membrane modification, specifically duplication of the glomerular basement membrane by light or electron microscopy, as may occur in transplant glomerulopathy in kidney allografts or thrombotic microangiopathy in native kidneys.41 The concurrent presence of peritubular capillary C4d staining was indicative of a humoral pathophysiology in the allografts and was associated with positive glomerular staining even in the absence of detectable glomerular basement membrane duplication. Glomerular C4d positivity is not part of current Banff criteria.

Electron microscopy

The ultrastructural features of transplant glomerulopathy include endothelial cell swelling and/or vacuolization, loss of endothelial fenestrations, subendothelial widening of the lamina rara interna with electron-lucent or flocculent material, cell debris, and reduplication or multilamination of the lamina densa (see Figure 2). Fibrin may be present. These findings are indistinguishable from those of any thrombotic microangiopathy. Electron-dense deposits are usually not present, and these changes strongly suggest a recurrent or de novo immune complex glomerulonephritis, especially immunoglobulin-mediated (type 1) membranoproliferative glomerulonephritis or a C3 glomerulopathy. Importantly, these latter two conditions would have characteristic immunofluorescence findings separating them from transplant glomerulopathy. Of note, however, sparse deposits may occasionally be found in transplant glomerulopathy, presumably resulting from nonspecific deposition of immune reactants such as IgM and C3.

In a seminal paper, Wavamunno et al42 profiled with protocol biopsies the ultrastructural changes occurring over time in seven patients who eventually developed transplant glomerulopathy by light microscopy as compared with eight controls who did not. Endothelial cell swelling and vacuolization, as well as widening of the lamina rara interna, were detectable within 1 to 3 months and were significantly greater than in controls. Glomerular basement membrane duplication was obvious within the first year, although endothelial fenestration was not significantly reduced until 3 years. Light microscopic changes diagnostic of transplant glomerulopathy appeared only after 2.3 years. Similarly, Haas and Mirocha43 found the combination of endothelial swelling, subendothelial widening, and glomerular basement membrane reduplication together within the first 3 months post-transplantation in 11 of 17 biopsies with C4d-positive acute antibody-mediated rejection and 8 of 16 with C4d-negative acute antibody-mediated rejection compared to 0 of 65 without acute antibody-mediated rejection and/or donor-specific antibody. Individually, however, these changes were found in 5 of 17 with cellular rejection and 4 of 10 with calcineurin inhibitor toxicity. On follow-up biopsies of 18 patients (11 C4d positive), 8 developed overt transplant glomerulopathy by light microscopy, a result that appeared to be inhibited by treatment of the acute rejection.

Based on these findings, transplant glomerulopathy is now diagnosable by the most recent Banff criteria purely by electron microscopy in the absence of double contours by light microscopy. This requires glomerular basement membrane duplication in at least three glomerular capillaries (with associated endothelial changes and/or lamina rara interna electron-lucent widening) and is designated cg1a by Banff 2015.

Similar to the glomerular capillary bed, endothelial cell activation and damage with basement membrane changes occur in the peritubular capillary bed in transplant glomerulopathy. In fact finding such changes is sufficient to satisfy Banff criteria for tissue injury in diagnosing chronic active antibody-mediated rejection, even in the absence of glomerular changes. Such peritubular capillary changes frequently do coexist, however, with transplant glomerulopathy irrespective of the underlying etiology. Monga et al44 initially reported peritubular capillary basement membrane multilayering by electron microscopy (Figure 5) in all 14 patients with transplant glomerulopathy, as well as in all 38 specimens with transplant glomerulopathy in a subsequent study,45 and they suggest that their presence can be used to infer current or future transplant glomerulopathy, when glomeruli are not present in the specimen.45 Although clearly found in association with transplant glomerulopathy in cases of antibody-mediated rejection, numerous authors have emphasized that peritubular capillary basement membrane multilayering is not specific for an antibody cause. Other etiologies include postinfectious glomerulonephritis, lupus nephritis, cyclosporine nephrotoxicity, and obstructive uropathy.46, 47, 48

Figure 5
figure 5

Electron microscopy showing peritubular capillary basement membrane multilayering (PTCBMML). Prominent reduplication of the basement membrane of a peritubular capillary. Original magnification × 4000

Full size image

In properly interpreting peritubular capillary basement membrane multilayering pathologically, several issues require consideration: the minimal circumference of a given capillary that must be involved to consider it positive, for example, ≥60%, 100%; the cutoff for positivity, for example, ≥3 layers, ≥5 layers; how many peritubular capillaries should be studied, for example 10, 25, and so on; and finally, whether only the three most affected capillaries should be graded as opposed to evaluating all that are available. For example, Iványi et al49 considered circumferential changes to involve ≥75% of the peritubular capillary circumference, and they evaluated the mean number of layers in the entire specimen. In contrast, Liapis et al50 scored biopsies based only on the three most affected capillaries. The most severe category required 1 capillary with ≥7 circumferential layers, with 5 or more in the remaining 2. This study served as the basis of current Banff criteria. These severe changes were rare in native kidneys with the exception of advanced thrombotic microangiopathy, and they were not specific for chronic active antibody-mediated rejection in the transplant specimens. Severe changes were also found in acute cell-mediated rejection, chronic active cell-mediated rejection, and calcineurin inhibitor nephrotoxicity. The positive predictive value for the most severe changes indicating antibody causation was only 49%.

Current Banff criteria for using peritubular capillary basement membrane multilayering to diagnose chronic active antibody-mediated rejection require ≥7 layers in 1 capillary plus ≥5 layers in 2 additional capillaries, but it is not specified what percentage of circumference must be involved to be considered positive or how many capillaries should be studied to exclude the diagnosis.

Several recent studies addressed the utility of assessing the mean peritubular capillary basement membrane multilayering in the entire sample or a fraction thereof. 51 de Kort et al52 found that a mean of 2.5 layers/peritubular capillary significantly predicted future development of transplant glomerulopathy. Evaluation of only 10 peritubular capillaries was sufficiently sensitive to calculate means. Additionally, in patients with 2 or more biopsies evaluated by electron microscopy, the presence of at least 1 biopsy with a mean peritubular capillary basement membrane multilayering of ≥2.5 resulted in a significantly greater chance of developing transplant glomerulopathy than if consistently ≤2.5. Others have also found that 10 peritubular capillaries is a sufficient number to evaluate for prognostic purposes.53

Pathogenesis of transplant glomerulopathy

Chronic, repetitive glomerular endothelial cell injury and activation result in the pathologic findings diagnostic of transplant glomerulopathy. When antibodies are involved, complement activation may mediate endothelial damage, although antibodies can activate and injure endothelial cells in the absence of complement. Numerous inflammatory cell types may be involved, including monocytes/macrophages, lymphocytes, NK cells, and neutrophils. Various cytokines may participate as well. Five major, often interrelated, causes of endothelial injury have been described in association with transplant glomerulopathy: alloantibodies, autoantibodies, cell-mediated injury, thrombotic microangiopathy, and hepatitis C. These will be discussed in detail.

Although the endothelium may be initially injured, it has been hypothesized that podocyte depletion may be the final common pathway resulting in proteinuria, reduced glomerular filtration rate, and allograft loss.54 In a recent study, patients with transplant glomerulopathy had 10 to 20 times increased urinary podocin/creatinine ratios, a marker of the urinary podocyte number, as compared with those with stable function.54 A subset of these patients developed transplant glomerulopathy early (within 2 years) with reduced podocyte number per glomerulus without significant change in size, resulting in reduced podocyte density. A second subset developed transplant glomerulopathy much later (~10 years) and had both reduced podocyte number/glomerulus and significant glomerular enlargement, resulting in an even more marked reduction in podocyte density. Podocyte density significantly correlated with proteinuria, glomerular filtration rate, and glomerulosclerosis. Experimental evidence that podocyte damage can follow primary endothelial cell injury comes from a murine model of adriamycin nephrosis.55

Antibody-mediated rejection

By far the most widely recognized cause of transplant glomerulopathy is chronic antibody-mediated injury from donor-specific HLA alloantibodies. Based on current Banff criteria, transplant glomerulopathy represents sufficient histologic evidence of chronic tissue injury to satisfy the morphologic component for a diagnosis of chronic active antibody-mediated rejection by Banff criteria.9 The progression of chronic antibody-mediated damage to transplant glomerulopathy has been demonstrated in Cynomolgus monkeys.56 The clinical evidence in support of antibody-mediated causation is overwhelming. TG can develop in association with either preformed donor-specific antibody or de novo donor-specific antibody. In a study of 1-year protocol biopsies, Gloor et al57 found transplant glomerulopathy in 22% of 37 HLA-incompatible (cross-match positive), 13% of 24 ABO-incompatible, and 8% of 198 conventional live donor transplants. By multivariable analysis, transplant glomerulopathy was significantly associated with prior acute antibody-mediated rejection (odds ratio 17.5, P<0.0001), and 44% of such patients developed transplant glomerulopathy. Similarly, Loupy et al58 found that 43% of patients with pretransplant donor-specific antibody positivity and subclinical acute antibody-mediated rejection on a 3-month protocol biopsy developed transplant glomerulopathy by 1 year. Weibe et al59 studied 315 patients without preformed donor-specific antibody. Using routine surveillance serum screening and protocol biopsies, they determined that de novo donor-specific antibody developed in 47 patients (15%). Transplant glomerulopathy was not significantly detected before antibody development, but was significantly present afterwards in the 29 with clinical renal dysfunction. Other studies demonstrate the association of de novo donor-specific antibody and transplant glomerulopathy.60

Additional evidence supports donor-specific antibody causation. In a study of 598 conventional transplants, 73 developed transplant glomerulopathy over 5 years, with a significant association by multivariable analysis with HLA class II level, HLA class II donor-specific antibody level, and prior acute antibody-mediated rejection.61 Sis et al25 coined the term 'ABCD Tetrad' of late antibody-mediated rejection: 'A' for donor-specific antibody, 'B' for peritubular capillary basement membrane multilayering, 'C' for C4d positivity, and 'D' for reduplication of the glomerular basement membrane (transplant glomerulopathy). In their series of 53 biopsies (41 patients) with transplant glomerulopathy, 60% had donor-specific antibody, 91% had peritubular capillary basement membrane multilayering, and 36% were C4d positive. Many other series showed similar results, with a majority or significant minority of transplant glomerulopathy cases having evidence of HLA antibody involvement, either as donor-specific antibody positive and/or C4d positive.18, 22, 27, 32, 34, 38, 62 The importance of this antibody-mediated mechanism has been stressed and discussed in detail in recent reviews on transplant glomerulopathy. 63, 64

Perhaps, more important than donor-specific antibody directed against donor serologic antigens may be antibodies reacting with donor-specific epitopes not present in the recipient.65, 66 Such epitopes may be shared among HLA antigens and hence not be recognized simply by considering donor serologic antigens. For example, in a study of 17 allograft nephrectomy specimens, ~20% of HLA antibodies were directed against donor-specific alleles, whereas ~80% were directed against donor specific epitopes.30 Similarly, Lachmann et al67 performed epitope analysis on 9 patients with allograft nephrectomies and identified 25 donor–recipient epitope mismatches. Out of a total of 243 class I anti-HLA antibodies in these 9 patients that were not donor-specific antibody, 125 reacted with donor epitopes and were considered donor epitope-specific antibodies. Sapir-Pichhadze et al68 compared the DRB1, 3, 4, and 5 and DQ epitope mismatches of 52 donor–recipient pairs where transplant glomerulopathy developed with that of 104 controls without transplant glomerulopathy. By multivariable analysis, the OR for developing transplant glomerulopathy significantly increased by 25% for each additional 10 DRB1, 3, 4, and 5+DQ eplet mismatches and when comparing either of the higher two tertiles with the lowest tertile. This held for DRB alone but not for DQ alone.

Non-HLA antibodies

Non-HLA antigens may elicit antibody responses that contribute to renal allograft dysfunction and TG, either alone or in conjunction with anti-HLA antibodies. Some are polymorphic and elicit alloresponses. Others are cryptic autoantigens that may be expressed on the cell surface following injury and result in loss of tolerance and autoantibody-mediated injury.69

The most polymorphic non-HLA alloantigens are the MICA and MICB. At least 100 MICA and 40 MICB alleles have been described (accessible at http://hla.alleles.org/nomenclature/stats.html) and to which an alloantibody response may occur.70 Anti-MICA antibodies may be detectable pretransplantation and/or develop post-transplantation. They can be autoreactive, donor derived, or have unrelated third-party specificity. The effect of MICA/B sensitization on transplant outcome remains uncertain.71 Their role specifically in transplant glomerulopathy has not been addressed to our knowledge.

Autoantibodies are now recognized as important mediators of acute and chronic allograft damage in kidney and other solid organ transplants. Naturally occurring IgM autoantibodies have been well described and may react with apoptotic neoantigens facilitating senescent cell clean-up.72, 73 More recently, however, it has been shown that IgG autoantibodies normally number in the thousands and are of uncertain physiologic significance.74 Their numbers increase with age and male sex, although they may be decreased in certain disease states. Numerous autoantibodies in the setting of kidney and other solid organ transplantation have been described,71 resulting in both acute75 and chronic injury.76 Some well-studied antigenic determinants include the antiangiotensin II type 1 receptor,77, 78 glutathione-S-transferase T1,79 and vimentin.80

Some autoantibodies have been described specifically in association with transplant glomerulopathy. Joosten et al81 compared 16 patients with TG to 16 controls with chronic allograft changes but without transplant glomerulopathy. Serum IgG from 11 of the 16 transplant glomerulopathy patients reacted with human glomerular basement membrane, 7 of which also reacted with purified heparin sulfate proteoglycans, predominantly agrin. Only 3/16 controls reacted with glomerular basement membrane (P=0.0044). Interestingly, patients with antiglomerular basement membrane antibodies had significantly more prior acute rejections than the 5 without these antibodies, yet only 6 of the 11 had concurrent anti-HLA antibodies. Of note, agrin is also expressed in the basement membranes of lung, skin, and muscle. Interestingly, there is no clinical evidence of disease in these organs associated with these antibodies. This same group had previously used a murine model of transplant glomerulopathy and found the predominant antibody to be directed against the heparin sulfate proteoglycan perlecan.82 In humans, Yang et al83 showed that pretransplant autoantibodies against LG3, a C-terminal fragment of perlecan, significantly increased the risk for delayed graft function and reduced 1-year allograft function following delayed graft function, although transplant glomerulopathy was not specifically addressed.

Dinavahi et al84 used a protein microarray analysis to study autoantibody repertoires pre- and post-transplantation from patients with and without transplant glomerulopathy. Transplantation per se routinely resulted in changed autoantibody repertoires irrespective of transplant glomerulopathy with unique profiles for individual patients. Pre- and post-transplantation peroxisomal-trans-2-enoyl-CoA-reductase IgG autoantibodies strongly associated with transplant glomerulopathy, a result confirmed by ELISA and in a validation set.

Angaswamy et al85 compared 26 patients with transplant glomerulopathy to 10 stable controls with normal histology. Although only 16 transplant glomerulopathy patients had anti-HLA antibodies (12 donor-specific antibodies), 22 of 26 had autoantibodies (both IgM and IgG) against the kidney-restricted antigens collagen-IV and fibronectin giving an OR of 22 (P=0.001). Six of the 10 HLA-negative patients were autoantibody positive. Four of 18 transplant glomerulopathy patients had these autoantibodies pretransplantation.

Jackson et al75 used a protein array analysis of eluates from 10 endothelial cell cross-match-positive but HLA-negative kidney transplant recipients to identify 4 endothelial cell antigenic targets: endoglin, Fms-like tyrosine kinase-3 ligand, EGF-like repeats and discoidin I-like domains 3, and intercellular adhesion molecule 4. Sera from 150 additional kidney transplant recipients were tested by ELISA for autoantibodies against these 4 antigens, 36 of which tested strongly positive against all 4 antigens. The transplant glomerulopathy score was significantly higher in this group of 36, even restricted to those with no/low HLA donor-specific antibody.

Cell-mediated rejection

Many cases of transplant glomerulopathy lack evidence of antibody involvement. Akalin et al24 found that only 5 of 36 biopsies with transplant glomerulopathy were C4d positive, and only 36% of 28 tested patients were donor-specific antibody positive. In fact, 15 of the 28 patients lacked any HLA antibodies by Luminex Flow Beads. Lesage et al32 found that 45 of 71 patients with transplant glomerulopathy lacked both donor-specific antibody and C4d staining. Of 46 patients with TG in one series, 25 were donor-specific antibody negative and C4d negative.27 In another series of 45 transplant glomerulopathy biopsies, 26% lacked both donor-specific antibody and C4d staining.25 Torres et al62 evaluated the late (≥6 months) allograft biopsies of 59 patients and found 17 with transplant glomerulopathy satisfying Banff 2013 criteria for chronic active antibody-mediated rejection, as compared to 12 that were both donor-specific antibody and C4d negative.

There is a substantial body of evidence supporting cell-mediated rejection in the pathogenesis of transplant glomerulopathy. Transplant glomerulitis, a precursor of transplant glomerulopathy, may be composed of predominantly T cells in C4d-negative biopsies with acute rejection, as opposed to C4d-positive biopsies where monocytes predominate.86 Batal et al87 found that 47% of 32 biopsies with donor-specific antibody-negative, C4d-negative acute cell-mediated rejection had glomerulitis (g≥0), as did 16 of 26 borderline, donor-specific antibody-negative, C4d-negative biopsies. Glomerulitis correlated with subsequent development of transplant glomerulopathy, although data were not provided specifically for donor-specific antibody-negative, C4d-negative, g≥0 cases. Similarly, in a series of 44 patients with acute cell-mediated rejections, 9 had g≥0.88 In a series of late C4d-negative biopsies of donor-specific antibody-negative patients, 5 of 18 with acute cell-mediated rejections had g≥0, as did 1 of 10 with borderline lesions. Furthermore, 3 of 18 with acute cell-mediated rejections also had concurrent transplant glomerulopathy, as did 1 of 10 with borderline lesions.89

Cell-mediated rejection more commonly precedes transplant glomerulopathy than does antibody-mediated rejection in some reports. In a series of 55 patients with transplant glomerulopathy, 6 had prior acute cell-mediated rejection, 9 had prior borderline cell-mediated rejection, but only 3 had antibody-mediated rejection.26 In another series of 37 patients with transplant glomerulopathy who had available rejection history, 20 had prior acute rejections, of which 74% were cell-mediated rejection and 26% were antibody-mediated rejection.25 Of note, both of these studies were published before recognition of C4d-negative antibody-mediated rejection. It is probable that some of the cases considered cell-mediated rejection were mixed cell and antibody mediated. Furthermore, it should be acknowledged that cell-mediated rejection is a defined risk factor for future production of de novo DSA,59 which could mediate the development of TG through antibody-mediated rejection.

Other evidence for T-cell involvement exists. Akalin et al90 compared 16 patients with nonspecific, chronic nephropathy to 5 with transplant glomerulopathy. All transplant glomerulopathy cases stained positive for CD3 and ICOS, present on activated CD4 and CD8 T cells (also on resting B cells, which were not present), as well as for the chemokine receptor CXCR3 (present on activated, effector T cells) and its ligands Mig and IP-10. Biopsies with chronic nephropathy lacking transplant glomerulopathy did not have these findings.

Naïve CD4 T cells can be directed to several lineages, including Th1, Th2, Th17, and T-regulatory cells. Th1 CD4 cells express the transcription factor T-bet and interferon-γ. Homs et al91 found significantly increased expression of T-bet and interferon-γ mRNA in renal biopsies of patients with transplant glomerulopathy as compared to those without transplant glomerulopathy. Sun et al92 also found significantly increased T-bet expression in glomerular and peritubular capillaries of patients with transplant glomerulopathy, which correlated with CD4, CD8, and CD68 infiltration. The Th2 transcription factor GATA-3 was rarely found. In contrast, Sun et al93 found predominant intraglomerular T-bet expression in acute antibody-mediated rejection as compared with acute cell-mediated rejection. Yadav et al94 found significantly greater T-bet-positive mononuclear cell infiltration of transplant glomerulopathy biopsies compared with biopsies with nonspecific, chronic changes without transplant glomerulopathy, although all of the transplant glomerulopathy patients were donor-specific antibody positive and C4d positive. Overall, these results support a role for effector T cells in mediating transplant glomerulopathy, either alone or in conjunction with antibody-mediated damage.

Thrombotic microangiopathies

The thrombotic microangiopathies may produce a morphologic picture by either light or electron microscopy indistinguishable from transplant glomerulopathy (Figure 6). Thrombotic microangiopathy represents a syndrome characterized pathologically by arteriolar and/or capillary thrombosis with associated vessel wall changes, including fibrinoid necrosis acutely and myointimal proliferation ('onion skinning') chronically.95 Red blood cell fragments may be visible within the damaged vessel wall. Typically, a microangiopathic hemolytic anemia is concurrently present, characterized by thrombocytopenia, anemia with schistocytes on peripheral blood smear, reduced or absent haptoglobin, indirect hyperbilirubinemia, and elevated LDH. The thrombotic microangiopathys may be primary, including hereditary or acquired causes, or they may be secondary to various other conditions (Table 2). Disease may be limited clinically to a specific organ, such as native kidneys or a kidney allograft, or may be systemic when associated with a microangiopathic hemolytic anemia or multiple organ involvement. Similar to transplant glomerulopathy, the thrombotic microangiopathys are initiated by endothelial cell injury from any cause as listed in Table 2.

Figure 6
figure 6

Thrombotic microangiopathy (TMA) in a native kidney. Electron micrograph of a glomerulus from a patient with bone marrow-associated TMA. There is prominent widening of the subendothelial space by edema, amorphous proteinaceous deposits, and cell debris. The endothelial cells are swollen and have lost their fenestrations.

Full size image
Table 2 Classification of thrombotic microangiopathy (TMA)
Full size table

Following kidney transplantation, thrombotic microangiopathy may occur de novo or as recurrent disease with approximately equal frequency. Using a registry analysis of kidney transplant recipients, Reynolds et al96 estimated de novo occurrence in 0.8% and a recurrence rate of 29%. Although most cases developed in the first 3 months, new cases continued to develop through the 3 years of follow-up. Single center series found incidence rates averaging about 5%,97, 98, 99 although prevalence rates as high as 12% have been reported.100 Recurrent thrombotic microangiopathy is an indication for genetic testing for complement regulatory protein mutations and possibly for treatment with eculizumab.101

Development of de novo post-transplantation should prompt evaluation for antibody-mediated rejection.98 By Banff 2015, acute thrombotic microangiopathy is one of the defining morphologic features of acute tissue injury sufficient to diagnose acute antibody-mediated rejection in the absence of another cause for the thrombotic microangiopathy. In a single center series of 59 patients with de novo thrombotic microangiopathy, 55% were C4d positive, indicating antibody-mediated rejection was the most common cause.98 In a series of 1073 allograft biopsies, 37 had thrombotic microangiopathy.102 The incidence of C4d positivity in those biopsies with thrombotic microangiopathy was identical compared with the overall population (16%). This result was confounded, however, by time of biopsy: the odds ratio for having a thrombotic microangiopathy if C4d positive was 3.8 for biopsies within 90 days, compared with 0.2 if taken ≥90 days. C4d-positive thrombotic microangiopathies were more likely to have glomerular thrombi and extensive microvascular inflammation, especially neutrophilic.

Calcineurin inhibitor toxicity is a well-described cause of thrombotic microangiopathy both in native kidneys97, 103, 104, 105 and kidney allografts.106, 107, 108 Viral infections are associated with de novo thrombotic microangiopathy, including CMV,109 parvovirus B19,110 hepatitis C (with associated antiphospholipid antibodies),111 and influenza.112 Whether all patients with de novo thrombotic microangiopathy should have screening for genetic mutations in complement regulatory proteins is uncertain at this time,113 but may be a consideration especially with subsequent recurrence.

The pathologic identification of thrombotic microangiopathy as the cause of transplant glomerulopathy is most challenging when disease is allograft limited. By light microscopy, thrombotic microangiopathy is suggested by fibrin thrombi, RBC fragments, and myointimal proliferation with onion skinning involving arterioles. If just endothelial changes are present, the distinction from other causes of transplant glomerulopathy cannot be made. Similarly, by electron microscopy the changes of early transplant glomerulopathy are identical to early thrombotic microangiopathy. The presence of RBC fragments or fibrin thrombi support thrombotic microangiopathy as the cause of the transplant glomerulopathy. Both transplant glomerulopathy and thrombotic microangiopathy stain positive for C4d in glomeruli, although peritubular capillary staining for C4d indicates antibody-mediated transplant glomerulopathy.

Current Banff criteria for transplant glomerulopathy require absence of evidence of chronic thrombotic microangiopathy. Many published series of transplant glomerulopathy exclude cases with other evidence of thrombotic microangiopathy besides transplant glomerulopathy. Typically, this consists of detecting RBC fragments in vessel walls or glomerular capillaries, or onion-skin lesions of arteriole intima. By contrast, Baid-Agarwal et al100 did not make this distinction, and in a series of 25 cases of transplant glomerulopathy, 8 had other evidence diagnostic of thrombotic microangiopathy. Sreedharanunni et al99 evaluated 266 indication biopsies with glomerular pathology, of which 107 satisfied Banff 2009 criteria for transplant glomerulopathy (double contours in ≥10% of capillary loops in most affected glomerulus by light microscopy) and 91 had glomerular thrombotic microangiopathy (fibrin thrombi acutely, and subendothelial fibrin, endothelial swelling, mesangial hypercellularity, mesangiolysis, collapse, and small crescents chronically) with or without vascular thrombotic microangiopathy. Another 23 had only vascular thrombotic microangiopathy. Of the 114 biopsies with any thrombotic microangiopathy, 21 (18%) had coexisting transplant glomerulopathy by their definitions.

Hepatitis C virus infection

The most common glomerulopathy associated with hepatitis C virus infection is type I membranoproliferative glomerulonephritis, usually with low complement, type II cryoglobulinemia, and IgM rheumatoid factor.114, 115, 116 In the kidney allograft, hepatitis C virus can cause either recurrent or de novo type I membranoproliferative glomerulonephritis. Chronic membranoproliferative glomerulonephritis in the allograft may be difficult to distinguish from transplant glomerulopathy by light microscopy, but these lesions should be readily distinguished by detecting immune complexes by immunofluorescence and electron microscopies, along with the associated serological findings such as low complement and serum cryoglobulins, which may117 or may not be present.118

Other renal lesions associated with hepatitis C virus in native kidneys include membranous nephropathy,119 fibrillary glomerulonephritis,120 thrombotic microangiopathy,121 and collapsing glomerulopathy following therapy with interferon122 or pegylated interferon.123 Kidney allografts may be similarly affected, including reports of thrombotic microangiopathy,111 fibrillary glomerulonephritis,124 and collapsing glomerulopathy.125

Gallay et al.126 first noted the association of hepatitis C virus with transplant glomerulopathy in a report of two cases that concurrently had features of membranoproliferative glomerulonephritis. The altered morphological appearance of membranoproliferative glomerulonephritis that may result from immunosuppression following transplantation was noted, highlighting the difficulty in distinguishing this lesion from transplant glomerulopathy. Cosio et al127 found 9 of 27 (33%) patients with transplant glomerulopathy to be hepatitis C virus positive, compared with only 1.5% of 105 transplant patients without transplant glomerulopathy (P=0.0004), although immunofluorescence and electron microscopies were not routinely performed, possibly resulting in misclassification of hepatitis C virus-related membranoproliferative glomerulonephritis as transplant glomerulopathy. Others have found a significant association as well.26, 100 Baid-Agrawal et al100 described a case series of 25 patients with transplant glomerulopathy. Twelve were C4d positive supporting an antibody-mediated cause. Nine (33%) were hepatitis C virus positive (3 of which were C4d positive), a rate significantly higher than the 7% prevalence in 29 controls with calcineurin inhibitor toxicity and no transplant glomerulopathy. Interestingly, five of the nine patients had concurrent thrombotic microangiopathy, as did three other hepatitis C virus-negative patients, highlighting the potential overlapping pathophysiology of hepatitis C virus and thrombotic microangiopathy in causing non-antibody-mediated transplant glomerulopathy. In contrast, Cruzado et al117 noted a similar prevalence of transplant glomerulopathy in patients with (5 of 44, 11.4%) or without hepatitis C virus (6 of 52, 11.5%) using light microscopy and immunofluorescence microscopy in all specimens, with electron microscopy if indicated.

The pathophysiologic link between hepatitis C virus and transplant glomerulopathy remains unclear. Baid-Agrawal et al100 proposed several potential mechanisms. Transplant glomerulopathy may represent an atypical morphologic presentation of type 1 membranoproliferative glomerulonephritis, with altered immune complex deposition resulting from antigen–antibody imbalance secondary to chronic immunosuppression. Hepatitis C virus may upregulate the alloimmune response based on concurrent C4d positive and microvascular inflammation. Hepatitis C virus may cause transplant glomerulopathy as a result of the induction of antiphospholipid antibodies.103

Irrespective of the development of transplant glomerulopathy, hepatitis C virus can be effectively treated post-transplantation with direct acting anti-viral agents.125 Sustained virologic response rates of ~90% have been reported. Given the known association of hepatitis C virus with adverse patient and allograft survival, as well as the recently demonstrated efficacy and safety of direct acting anti-viral therapy of hepatitis C virus in transplant recipients,125, 128, 129, 130 in our opinion treatment should be considered in all viremic transplant patients irrespective of transplant glomerulopathy.

Clinical manifestations

TG may be clinically silent. In one series of 55 patients with transplant glomerulopathy, 27 (49%) were detected by protocol biopsy.26 The prognosis of these patients was no better than those with clinically evident disease. Early clinical manifestations include low-level proteinuria and/or subtle decline in glomerular filtration rate. As the lesion progresses, however, hypertension may develop, glomerular filtration rate declines, and proteinuria increases, not uncommonly into the nephrotic range.131

Transplant glomerulopathy is associated with reduced allograft survival in numerous studies.26, 32, 61, 131, 132 Attempts have been made to refine the prognostic implications of transplant glomerulopathy based on associated pathologic or clinical features. Concurrent peritubular capillary C4d positivity is associated with worse graft survival in some studies of transplant glomerulopathy,32, 61, 132 but not in others.133 Dobi et al134 evaluated 59 patients with transplant glomerulopathy and found that concurrent intimal arteritis resulted in significantly shorter graft survival. When occurring late, microvascular inflammation (g and ptc) was shown by multivariable analysis to predict graft survival independent of transplant glomerulopathy in one study of 251 allograft recipients.88 In another study of 33 transplant glomerulopathy patients, however, the degree of microvascular inflammation did not correlate with survival.135 Concurrent hepatitis C virus may100, 136 or may not34 significantly affect graft survival with transplant glomerulopathy. In an earlier study, Banfi et al131 found that proteinuria ≥2.5 g/day significantly impacted graft survival in 28 patients with transplant glomerulopathy (graft survival 8% vs 67% if lower proteinuria, P<0.005). Patri et al34 studied 92 patients with transplant glomerulopathy and developed a prognostic index for graft survival based on the ci, ct, and ti Banff scores along with serum creatinine and proteinuria that was validated in an external cohort of transplant glomerulopathy patients.

Treatment

There is no specific treatment proven to work for transplant glomerulopathy. All patients should have general supportive measures given to any patient with chronic kidney disease. These include salt and protein restriction, cessation of smoking, weight loss if obese, control of blood pressure, renin–angiotensin system blockade, lipid control, mineral and bone disorder therapy, and control of anemia. Specific therapy may be offered on an individual basis, based on underlying cause and stage of progression. If a cell-mediated alloresponse is involved, optimization of immunosuppression with tacrolimus, mycophenolate, and steroids is recommended.

In the case of HLA alloantibody-mediated transplant glomerulopathy, the best approach to treatment is prevention. That would start by optimal HLA matching and avoiding transplantation if donor-specific antibodies are detectable, both of which are problematic given the shortage of organs. Whether epitope matching would reduce development of de novo donor-specific antibody and subsequently transplant glomerulopathy remains to be proven.66 If donor-specific antibodies are detected, either preformed or de novo, close clinical surveillance as well as serial titration and protocol biopsy may identify the earliest stages and potentially prevent progression. There is some evidence that early, for example, at cg1a, appropriate treatment may forestall progression.43

It may become possible to determine patients destined to develop antibody-mediated transplant glomerulopathy before any microscopic changes appear by using a variety of microarray analyses of biopsy material. Validated ENDATs indicating endothelial cell injury satisfy current Banff criteria for antibody–endothelial interaction in diagnosing acute antibody-mediated rejection and chronic active antibody mediated. This is an active area of research by several groups,137, 138 and was highlighted at the 2015 Banff meeting.9 Algorithms have been proposed to identify antibody-mediated injury as opposed to cell-mediated mechanisms by microarray analyses of transcript expression.139 Whether treatment based on such analyses can prevent development of transplant glomerulopathy and/or improve allograft survival remains to be determined.

Eculizumab initially showed promise in preventing transplant glomerulopathy in sensitized recipients.140 Longer term follow-up showed it was not effective, with the possible exception of those attaining persistently low B-cell flow cross-matches.141 In a retrospective analysis, Orandi et al142 found 4 of 5 HLA-incompatible live donor transplant recipients with acute antibody-mediated rejection given eculizumab plus splenectomy to have functioning grafts free of transplant glomerulopathy on 12-month biopsy, as compared with none of 5 given eculizumab alone and 4 of 14 given splenectomy alone. A pilot randomized controlled trial suggested stabilization of glomerular filtration rate with eculizumab.143 A pilot randomized clinical trial in 18 patients with acute antibody-mediated rejection showed a trend for benefit with a human C1 esterase inhibitor,144 and none of 7 treated patients had transplant glomerulopathy on 6-month biopsy as compared with 3 of 7 given placebo. Choi et al145 reported a single center series of 36 patients with donor specific antibodies and chronic antibody-mediated rejection including transplant glomerulopathy who had failed standard of care were treated with monthly doses of tocilizumab, a monoclonal anti-IL-6 receptor antibody. They found stabilization of renal function without significant adverse events.

Standard treatment of antibody-mediated rejection should be considered for transplant glomerulopathy induced by alloantibodies if the lesion is early (cg1a or cg1b) and/or there is evidence of active microvascular inflammation, that is, g+ptc≥2. In support, a retrospective, single center study of 33 patients with transplant glomerulopathy found that treatment with intravenous immunoglobulin and rituximab stabilized the serum creatinine, but only in the subgroup with significant microvascular inflammation (ptc≥2 or g+ptc≥4).135 In other studies, the Banff microvascular inflammation scores and/or extent of peritubular capillary inflammation (≥50% versus ≤ 50%) were significantly related to allograft survival and development of transplant glomerulopathy,88, 146, 147 further supporting treatment based on microvascular inflammation. Additional potential therapies besides intravenous immunoglobulin and rituximab148, 149, 150, 151 for antibody-mediated transplant glomerulopathy include plasma exchange or immunoadsorption, high-dose steroids, and bortezomib, but data from randomized controlled trials are lacking. The use of such agents for autoantibody-mediated transplant glomerulopathy remains unstudied.

In transplant glomerulopathy patients with active hepatitis C virus, oral anti-viral therapy is required. Other active viruses such as cytomegalovirus should also be treated. In those with recurrent thrombotic microangiopathy or de novo thrombotic microangiopathy with complement regulatory protein mutations, eculizumab is indicated.152 Eculizumab may also be indicated for recurrent thrombotic microangiopathy secondary to antiphospholipid antibodies.153, 154 For thrombotic microangiopathy ascribed to calcineurin inhibitors, as a minimum, there should be dose reduction or possibly discontinuation. Sirolimus is not recommended, as it has been similarly associated with thrombotic microangiopathy. Other calcineurin inhibitor-sparing agents, such as belatacept, require further study.

Conclusion

Transplant glomerulopathy represents a morphologic lesion and not a specific diagnosis. The hallmark pathologic feature is endothelial cell injury with glomerular basement membrane reduplication by light or electron microscopy in the absence of immune complex deposits. C4d may stain peritubular capillaries in cases mediated by alloantibodies, and glomeruli may also stain positive. Clinically, transplant glomerulopathy may be silent, detectable on protocol biopsy. Overt clinical manifestations include up to nephrotic range proteinuria, hypertension, and declining glomerular filtration rate. Either way, transplant glomerulopathy is clearly associated with reduced graft survival. Transplant glomerulopathy results from chronic, recurring endothelial cell injury potentially mediated by HLA alloantibodies, various autoantibodies, cell-mediated injury, thrombotic microangiopathy and/or hepatitis C virus. Multiple causes may exist in a single patient. Other than nonspecific supportive care, the only proactive therapy must be directed at the specific underlying cause.