SEARCH

SEARCH BY CITATION

Keywords:

  • Anti-HLA antibodies;
  • capillaries;
  • C4d;
  • endothelial cells

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Histology of TG
  5. Antibodies and TG
  6. Pathogenesis of TG: A Model
  7. Differential Diagnosis of TG
  8. Management of TG
  9. Summary
  10. Acknowledgments
  11. References

Transplant glomerulopathy (TG) is a histologic entity described more than four decades ago. In the last few years, our understanding of TG has improved significantly. Current evidence supports the postulate that TG is a unique pathologic and pathogenic entity distinct from other forms of chronic allograft injury. Detailed electron microscopic studies have shown basement membrane abnormalities in glomerular and peritubular capillaries, indicating that this is a disease of the entire renal capillary network. Staining biopsies for the complement fragment, C4d, showed positivity in subgroups of TG, suggesting the participation of antidonor antibodies. Consistent with this postulate, the incidence of TG is increased in patients with antidonor HLA antibodies prior to the transplant. The use of surveillance biopsies has demonstrated that TG can develop during the first few months after transplantation, although it may remain clinically quiescent for several years. However, TG is progressive, leading to reduced graft survival. Recent studies demonstrated a close association between TG and anti-HLA class II antibodies. Current therapies for TG are likely of limited value. However, it is also likely that an improved understanding of TG pathogenesis will result in the development of effective therapies for this form of progressive kidney allograft damage.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Histology of TG
  5. Antibodies and TG
  6. Pathogenesis of TG: A Model
  7. Differential Diagnosis of TG
  8. Management of TG
  9. Summary
  10. Acknowledgments
  11. References

Transplant glomerulopathy (TG) is a pathologic condition of renal allografts recognized more than four decades ago (reviewed in (1)). TG includes a constellation of histologic features on light and electron microscopies (2). In this review, we will define TG by the characteristic duplication of glomerular basement membrane (GBM) observed by light microscopy (Figure 1), as recommended by the Banff working group (3).

image

Figure 1. Left: characteristic glomerular features of TG, note extensive duplication of GBM (arrows). Right: TG is often a focal lesion affecting only some glomeruli (top but not the bottom glomerulus in this slide). By Banff criteria (17), the diagnosis of TG is based on the identification of duplicated GBM in more than 10% of glomerular capillary loops in the most affected nonsclerotic glomerulus.

Download figure to PowerPoint

Evidence is accumulating that TG has a unique pathogenesis that distinguishes it from other chronic pathologic conditions of kidney allografts. TG is relatively rare compared to other lesions encountered in protocol biopsies obtained 1 year after transplantation (4). However, at least two circumstances highlight the importance of this disease: first, TG is associated with very poor allograft survival and, second, TG is perhaps the first example of a distinct pathologic entity arising from the histologic morass that was first called ‘chronic rejection’, and then ‘chronic allograft nephropathy’. We will argue here that the distinctiveness of TG from other forms of chronic allograft pathologies is given by the coexistence of three features: (i) histologic pattern, (ii) association with the presence of anti-HLA antibodies and (iii) the absence of other conditions that may cause duplication of GBM.

TG is rarely diagnosed clinically within 1 year posttransplant. However, recent studies showed that the clinical presentation of TG lags behind the initial histologic stages of the disease (5). Advanced TG usually is manifest by proteinuria that can be in the nephrotic range, and by progressive loss of kidney function. In our clinical experience and other's, the clinical manifestations of early TG are nonspecific, consisting of progressive, unexplained loss of kidney function, minor proteinuria and mild hypertension (5–7). Particularly in its early histologic phases, TG is a focal lesion, affecting only a few glomeruli (Figure 1) (3). However, sequential biopsies show progression with increasing percentage of affected capillary loops in an increasing number of glomeruli (5). TG is associated with poor long-term graft survival (Figure 2). In our series, the variables associated with reduced graft survival included: graft function and proteinuria at diagnosis and the severity of GBM duplication (Banff ‘chronic glomerulopathy [cg]’ score).

image

Figure 2. Kaplan-Meier plots of death-censored graft survival after conventional transplantation in patients without (––) and with TG (–––). Data generated from patients who received conventional kidney transplants at Mayo Clinic (N = 582, log-rank, p < 0.0001) (5).

Download figure to PowerPoint

Given the insidious clinical presentation of TG, it is likely that the reported incidence of this disorder, which is based on clinical biopsies, is an underestimate of its true incidence (reviewed in (7)). In agreement with this postulate, recent studies of surveillance and clinical biopsies showed that the incidence of TG is higher than expected, affecting 4% of conventional transplants at 1 year posttransplant. Thereafter, the incidence of TG increases progressively, reaching 20% at 5 years (4,5).

Histology of TG

  1. Top of page
  2. Abstract
  3. Introduction
  4. Histology of TG
  5. Antibodies and TG
  6. Pathogenesis of TG: A Model
  7. Differential Diagnosis of TG
  8. Management of TG
  9. Summary
  10. Acknowledgments
  11. References

The histologic features of TG have been recently reviewed (2). Still, it would be useful to highlight here some of these features as they provide clues to the pathogenesis of TG. Early clinico-pathologic studies suggested that TG represents the final manifestations of capillary injury (8). Thus, prior to the GBM remodeling, evidence of acute endothelial cell (EC) injury could be detected, including capillary EC swelling and glomerular hypercellularity. More recent studies provided evidence of physiologic alterations in glomerular EC in TG that was not present in patients with chronic allograft nephropathy (6). Importantly, electron microscopic (EM) studies demonstrated that the abnormalities of TG are not limited to glomerular capillaries, but also affect peritubular capillaries that show multiple layers of basement membrane (9). Recent studies showed that this multilamination is present in more than 90% of cases of TG (7), indicating that this disease is a pancapillaritis of the renal allograft. EM studies also have shown that the characteristic duplication of the GBM, in fact, often represents multilamination of that structure.

Until recently, histologic classifications of kidney allograft pathology did not separate TG from ‘chronic allograft nephropathy’, a nonspecific term indicating the presence of interstitial fibrosis and tubular atrophy (3). However, recent investigations showed that TG, particularly in its early stages, might develop independently from interstitial fibrosis, tubular atrophy and/or transplant arteriopathy (5,7). In agreement with previous studies (8), these studies also showed that TG is associated with interstitial, peritubular capillary and glomerular inflammation (2,5,7). It is important to note that glomerular inflammation coexists with TG and, in fact, becomes more common and severe as the duplication of the GBM progresses (5), suggesting that TG and its progression is associated with persistent capillaritis.

Earlier studies by Dr. Paul's group, utilizing a mouse model, suggested that TG might be the result of antiglomerular antibodies (10). The relevance of this hypothesis to humans obtained support by the demonstration of C4d deposition in peritubular capillaries of at least some cases of TG and more rarely in glomeruli. Furthermore, antidonor antibodies were also associated with TG (5,7,11,12). Earlier studies showed that during acute antibody-mediated rejection (AMR), antibody deposition in the allograft peritubular capillaries resulted in activation of the classical pathway of complement that generates several protein fragments, one of which, C4d, binds covalently to the allograft's capillary walls (13). Consequently, the demonstration of C4d at the same location in cases of chronic allograft nephropathy provided strong, albeit indirect, support for the hypothesis that antigraft antibodies likely play a role in at least some cases of this disease. This evidence was considered sufficiently compelling for the Banff 2005 group to include TG under the new category of antibody-mediated chronic rejection (14).

Antibodies and TG

  1. Top of page
  2. Abstract
  3. Introduction
  4. Histology of TG
  5. Antibodies and TG
  6. Pathogenesis of TG: A Model
  7. Differential Diagnosis of TG
  8. Management of TG
  9. Summary
  10. Acknowledgments
  11. References

The antibody-mediated hypothesis of allograft injury regained recognition with the description of cases of acute rejection associated with anti-HLA class I antibodies (15). Detailed description of the pathology of AMR (16,17) included glomerular abnormalities similar to those previously shown to herald TG, including glomerular EC swelling and widening of subendothelial cell spaces. Ultrastructural studies of serial biopsies from patients with AMR showed that acute EC injury involves both glomeruli and peritubular capillaries, and that over time, these lesions may evolve into pancapillary multilayering of basement membranes and obliteration of peritubular capillaries that was associated with interstitial fibrosis and tubular atrophy, perhaps secondary to ischemia (18).

Consistent with the antibody-mediated hypothesis of TG, additional studies noted that TG is significantly more common in patients with anti-HLA antibodies (5,7,11,19), particularly when the antibody was donor specific. In an investigation of surveillance biopsies performed 1 year posttransplant, TG was seen in 22% of patients with donor specific anti-HLA antibodies detected by positive crossmatch testing compared to 8% in recipients of conventional transplants, that is, those without a positive crossmatch testing (5). Furthermore, the risk of TG was also increased in patients with a history of AMR. Thus, approximately 45% of patients with AMR develop TG compared to 6% of recipients without rejection (5,19).

The use of newer, more sensitive assays for the detection of anti-HLA antibodies further clarified their association with TG, and showed a particularly strong association between anti-HLA class II antibodies and TG. In several recent studies, it was shown that the presence of anti-HLA antibodies, particularly anti-class II, was associated with poor allograft outcome (20–22). This relationship was noted regardless of whether the antibodies were detectable pretransplant or developed after transplantation. Recent investigations suggest that at least, in part, the association between anti-HLA antibodies and reduced graft survival is due to TG. Thus, among 51 patients with TG, antibodies to anti-HLA class I and/or II were detected in over 70% at the time of diagnosis of TG. Anti-HLA class II antibodies were detected in 64% of patients, and in two-thirds, these antibodies were donor specific (7). Gloor et al. (5) screened final crossmatch sera for anti-HLA antibodies in 582 recipients of conventional transplants. Of interest, 39% of these patients had detectable anti-HLA antibodies, and these antibodies, when directed against HLA class II, were associated with an increased risk of developing TG. Thus, 26% of patients with pretransplant anti-HLA class II antibodies developed TG compared to 8% of patients without these antibodies. Conversely, of the 55 patients who developed TG, 68% had anti-HLA class II antibodies pretransplant, a result very similar to other studies (7). These studies did not show a significant association between anti-HLA class I antibodies and TG. However, we believe that it would be erroneous to dismiss the role of these antibodies in TG pathogenesis, particularly because the highest risk of TG was among patients who had antibodies to both HLA class I and class II (5). These studies also revealed that patients with anti-HLA class II antibodies, even without detectable antidonor reactivity, also had a high risk of TG. However, we should be cautious with the interpretation of these data. Thus, previous studies showed significant crossreactivity between specific anti-HLA antibodies with multiple HLA antigens due to the presence of shared epitopes among these molecules (23). We interpret these results to suggest that the absence of antidonor HLA specificity in one assay does not ensure lack of antibody reactivity to the allograft.

The strong association between TG with anti-HLA class II antibodies appears to contradict observations from several laboratories that, at the time of diagnosis C4d is positive in peritubular capillaries only in a fraction of cases, perhaps fewer than half (2). One possible interpretation of these data is that there are two types of TG: C4d-positive, antibody-mediated and C4d-negative, antibody-independent. We favor an alternative hypothesis, alloantibody-mediated injury is not always associated with detectable C4d deposition. Supporting this hypothesis in one study, 87% of patients with TG had anti-HLA class II antibodies, but only 23% of cases were C4d-positive (5). In another study, only 36% of TG cases were C4d-positive, but 70% had anti-HLA antibodies that in 85% of the cases were donor-specific (7). Several hypotheses could explain the absence of C4d in TG. First, it is possible that after an acute antibody-mediated injury, the progression of capillary lesions does not involve complement activation (12), perhaps due to upregulation of complement regulatory proteins in EC. C4d-negative TG may also indicate that at the time of diagnosis antibodies are no longer pathogenic. However, this latter hypothesis is inconsistent with the presence of persistent capillaritis (5) and anti-HLA antibodies in blood. It is plausible that the interactions between antibodies and the endothelium are episodic and not continuous, making C4d an inconstant marker. Finally, there is evidence that immunoperoxidation is less sensitive than immunofluorescence to detect C4d (2). Thus, studies using the former method may have underestimated the presence of C4d in TG.

Pathogenesis of TG: A Model

  1. Top of page
  2. Abstract
  3. Introduction
  4. Histology of TG
  5. Antibodies and TG
  6. Pathogenesis of TG: A Model
  7. Differential Diagnosis of TG
  8. Management of TG
  9. Summary
  10. Acknowledgments
  11. References

Based on the histologic and the antibody data presented above, a pathogenic scheme for TG is proposed in Figure 3. In the normal kidney, HLA class I and II molecules are expressed in glomerular and peritubular capillaries of EC (24). This localization corresponds neatly to the histologic abnormalities of TG, and suggests that binding of anti-HLA antibodies to EC alters the cell phenotype and causes widespread capillary injury. However, it is unclear how antibodies binding to EC result in basement membrane duplication. Antibodies may have direct stimulatory effects on EC and/or may recruit secondary mediators, both humoral and cellular. We discussed previously the evidence for complement activation in both AMR and TG. However, complement activation may not be necessary or sufficient for the pathogenesis of TG because, as noted, many cases of TG are C4d-negative. Furthermore, in recipients of blood group-incompatible grafts, intense and persistent peritubular capillary C4d is common (2), but the incidence of TG is not significantly increased unless AMR occurs (19). Another group of humoral mediators potentially triggered by antibodies relate to the coagulation cascade. Clearly, antibodies may cause microthrombi in allografts, as observed in AMR (15). Microthrombi are rarely seen in TG. However, activation of the coagulation cascade in the kidney, as it occurs in thrombotic microangiopathy (TMA), results in a glomerulopathy with morphologic similarities to TG. Finally, the histopathology of AMR and TG provide direct evidence for the participation of monocytes, leukocytes and T cells in antibody-mediated allograft injury (2). The interactions between humoral and cellular mediators is suggested in some studies, noting that C4d may determine the type of cell involved in TG with more T cells in C4d-negative cases and more monocytes in C4d-positive cases. It seems likely that the physicochemical characteristics of anti-HLA antibodies (isotype and complement-activating ability, affinity, titer and others) play a role in TG pathogenesis and progression. Preliminary evidence suggests that antibodies with complement-activating activity may be more pathogenic. Otherwise, there is very little evidence for this postulate.

image

Figure 3. Proposed pathogenic mechanism of TG. The main pathway represented in the scheme is that initiated by the binding of antidonor antibodies to allograft endothelial cells. In addition, it should be noted that the pathogenesis of this lesion may be initiated in the absence of antibodies via activation of the coagulation cascade (TMA, thrombotic microangiopathy).

Download figure to PowerPoint

We know little about the mechanisms of TG progression. Serial biopsies of TG showed that ‘early’ or ‘mild’ lesions are frequently progressive, and that the process involves persistent capillaritis and interstitial inflammation (5). Other studies suggest that antibody-mediated injury result in obliteration of peritubular capillaries, leading to tissue underperfusion and fibrosis. Persistent inflammation and ischemia likely lead to progressive interstitial fibrosis and tubular atrophy in TG (4,18). It is also likely that, as in other forms of glomerular disease, there is a hemodynamic component to the progression phase of TG.

Differential Diagnosis of TG

  1. Top of page
  2. Abstract
  3. Introduction
  4. Histology of TG
  5. Antibodies and TG
  6. Pathogenesis of TG: A Model
  7. Differential Diagnosis of TG
  8. Management of TG
  9. Summary
  10. Acknowledgments
  11. References

As we have adopted simple criteria for the diagnosis of TG, that is duplication of the GBMs, it is important to differentiate TG from other conditions that may result in similar histologic changes. These conditions may include: ischemia, recurrent membranoproliferative glomerulonephritis and thrombotic microangiopathies (TMA). Generally, the histology of these conditions is distinct and does not pose a difficult diagnostic dilemma. TMA may be an exception to this latter statement because clotting may be triggered by antiendothelial antibodies. However, although coagulation activation may participate in the pathogenesis of TG, as discussed above, it is very rare to find capillary thrombi in TG. Some studies showed an association among hepatitis C, acute glomerulitis and TG (5).

Recent studies from the Edmonton group identified variable ‘forms’ of TG according to the presence of four components (the so-called ‘ABCD tetrad’) (7): antidonor antibodies (A), capillary basement membrane multilayering (B), C4d (C) and GBM duplication (D). By definition of TG, all of these cases had ‘D’, and in 73% of cases, there was evidence of antidonor antibodies. In contrast, 27% of cases had no detectable antidonor antibodies or C4d, raising the possibility that this is a different disease or at least a different phenotype for antibody-mediated injury. Perhaps related to this observation, previous studies showed that antibodies directed to non-HLA antigens of the donor kidney, such as Agrin, a component of the GBM, can be associated with TG (25).

Management of TG

  1. Top of page
  2. Abstract
  3. Introduction
  4. Histology of TG
  5. Antibodies and TG
  6. Pathogenesis of TG: A Model
  7. Differential Diagnosis of TG
  8. Management of TG
  9. Summary
  10. Acknowledgments
  11. References

Currently, there are no known effective therapies for TG. There is strong evidence that control of blood pressure and angiotensin II inhibition are effective in slowing down the progression of glomerular diseases in native kidneys. We strongly advocate those measures in patients with TG.

The pathogenic model discussed here suggests that reducing the production of antidonor antibodies may be effective in preventing and/or treating TG. However, currently there is no effective therapy to reduce anti-HLA antibody levels over long periods of time. Based on isolated cases, it has been suggested that additional immunosuppression may be helpful in TG, but this evidence is preliminary. Lacking effective therapy, preventing TG seems essential. Based on the evidence presented here, it is postulated that identification of recipients with anti-HLA antibodies, using solid-phase sensitive methods, and avoiding transplantation when those antibodies are anti-class II and have donor specificity may be effective in preventing TG. However, many patients with these antibodies and some with AMR do not develop TG. Thus, we need additional studies to more fully characterize the properties of the particular antibodies that are associated with TG.

Summary

  1. Top of page
  2. Abstract
  3. Introduction
  4. Histology of TG
  5. Antibodies and TG
  6. Pathogenesis of TG: A Model
  7. Differential Diagnosis of TG
  8. Management of TG
  9. Summary
  10. Acknowledgments
  11. References

Accumulating evidence suggests that TG is a unique pathologic entity distinct from other forms of chronic allograft damage. TG is closely associated with the presence of antidonor antibodies, particularly those directed against HLA class II. There is clear evidence that AMR often progresses to TG. However, the role of complement in TG may be less clear than previously believed as more studies show TG in the absence of detectable C4d deposition. Furthermore, it remains unclear whether TG is the result of ongoing antibody-mediated injury or is merely the pathologic sequelae of prior EC injury and activation. Unfortunately, TG even when diagnosed early is associated with poor graft survival. Lacking adequate therapy, preventing TG in patients at risk is essential, although the effectiveness of these maneuvers needs proof. We are confident that the improved understanding of TG pathogenesis will result in the development of effective therapies for this form of progressive kidney allograft damage.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Histology of TG
  5. Antibodies and TG
  6. Pathogenesis of TG: A Model
  7. Differential Diagnosis of TG
  8. Management of TG
  9. Summary
  10. Acknowledgments
  11. References

The authors would like to thank the Mayo Clinic transplant coordinators for their tireless efforts in patient follow-up and data collection. Special thanks to the Mayo Clinic tissue-typing lab for their collaboration in these studies and their efforts in applying new technologies to the measurement of anti-HLA antibodies. These studies were supported in part by grants from the Nephrology and Hypertension Division of the Mayo Clinic, Rochester.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Histology of TG
  5. Antibodies and TG
  6. Pathogenesis of TG: A Model
  7. Differential Diagnosis of TG
  8. Management of TG
  9. Summary
  10. Acknowledgments
  11. References