Peritubular Capillary Changes and C4d Deposits Are Associated with Transplant Glomerulopathy But Not IgA Nephropathy

Authors


Abstract

We examined our renal transplant population for glomerular diseases demonstrated on biopsy between January 1993 and April 2002, focusing on transplant glomerulopathy (TGP). Of 1156 patients followed in our clinics during this period, glomerular disease was diagnosed in 132 cases (11.4%). Glomerulonephritis was diagnosed in 86 transplants (7.4%), with IgA nephropathy (IgAN) being the commonest diagnosis [32 cases (2.8%)]. Thirty-one cases (2.7%) of biopsy-proven TGP were analyzed for associated factors compared with 27 cases (2.3%) of recurrent IgAN. Transplant glomerulopathy was less frequent with mycophenolate mofetil (MMF) and/or tacrolimus, whereas recurrent IgAN showed no such tendency (P= 0.02). Peritubular capillary (PTC) C4d deposition was observed in six of 24 cases (25%) with TGP but none with recurrent IgAN (P= 0.02). Peritubular capillary basement membrane (BM) multilayering was significantly greater in TGP (4.92 ± 2.94) than in recurrent IgAN (1.86 ± 1.04) (P < 0.001). The graft survival of TGP was worse than recurrent IgAN (P= 0.05). The association of TGP with BM multilayering and C4d deposits in PTC suggests a generalized disorder of the graft microcirculation and its BM, owing to antibody-mediated rejection in at least some cases. Transplant glomerulopathy has a serious prognosis but is less frequent in patients on newer immunosuppression, unlike recurrent IgAN.

Transplant glomerular changes were first observed in the early days of transplantation and soon emerged as a feature of grafts deteriorating from immunologic injury (1–3). Although transplant glomerulopathy (TGP) is uncommon, it is important clinically because many renal allografts with deteriorating function manifest TGP. The prevalence is approximately 5–7% of renal transplants (4–7). Transplant glomerulopathy has now been defined in the Banff schema and is distinct from other entities that have been identified in renal transplants with reduced function, e.g. rejection, allograft nephropathy, specific diseases and accelerating factors (8,9). Zollinger et al. were the first to propose the term ‘transplant glomerulopathy’, which they used to distinguish from glomerulonephritides characterized by deposition of immune complexes and infiltration by inflammatory cells (10). The specific lesions of TGP are reduplication of the glomerular basement membrane owing to widening of the subendothelial space, a moderate increase in mesangial matrix, and mesangial interposition (11). In addition, TGP is associated with peritubular capillary (PTC) basement membrane multilayering (12,13). Both TGP and PTC basement membrane multilayering show basement membrane thickening and reduplication, suggesting common pathogenetic mechanisms, particularly endothelial cell injury (12–15). Recent studies associate TGP with C4d deposition in PTC, suggesting alloantibody-mediated rejection as a pathogenesis (16). However, the pathogenesis of TGP is still unclear.

We performed a retrospective study of patients with biopsy-proven TGP compared with recurrent IgA nephropathy (IgAN) as a control group to assess the frequency and cause of TGP, and the effect of newer immunosuppression [tacrolimus and mycophenolate mofetil (MMF)], as well as to compare clinical features, clinicopathological correlation, and graft survival.

Materials and Methods

During the period from January 1993 to April 2002, 1156 renal transplant patients were followed up at the University of Alberta Hospital (cadaveric 775 and living related/unrelated 381 cases). Total of 618 graft biopsies were performed for unexplained deterioration of renal function or persistent proteinuria >1 g/day. The specimens were examined by light microscopy (LM), immunofluorescence (IF), and electron microscopy (EM). Pertinent clinical data were reviewed from medical records of the University of Alberta Hospital and renal transplant clinic files. The standard immunosuppression during this period was cyclosporine, azathioprine, and prednisone until 1996 when MMF replaced azathioprine. From 1997 to 1999 MMF was changed to azathioprine at 1 year in stable grafts. In 2001, tacrolimus replaced cyclosporine as the standard calcineurin inhibitor (CNI). We examined the prevalence of the glomerular diseases in all kidney transplants biopsied during this period, focusing particularly on TGP. Because TGP was diagnosed as frequently as IgAN in this period, we reasoned that it could serve as a comparator, particularly if TGP is alloimmune and IgAN is not. Cases of biopsy-proven TGP were analyzed for associated factors compared with biopsy-proven recurrent IgAN cases as a control group. Clinically, we recorded recipient gender, recipient age, donor type, donor age,% panel reactive antibody (PRA) pretransplantation, years post transplantation, years to diagnosis, serum creatinine at biopsy, type of immunosuppressive drugs, proteinuria, hematuria, hypertension, and current graft function. Pathologically, we recorded history of T-cell-mediated rejection episodes, rejection in the current biopsy, CNI effect (arteriolar hyalinosis), C4d staining, and PTC basement membrane multilayering.

Transplant glomerulopathy was diagnosed mainly on its typical light microscopic finding according to the Banff 97 classification, characterized by reduplication or ‘double contours’ formation of the glomerular basement membrane on PAS or silver stain (Figure 1A). Severity of TGP was graded as mild (Cg1), moderate (Cg2), and severe (Cg3) by the extent of ‘double contours’ in the most severely affected of nonsclerotic glomeruli (8). Electron microscopy was performed in all cases to support the diagnosis and to exclude de novo or recurrent glomerular disease (Figure 1B). In addition, PTC basement membrane multilayering on EM was reviewed (Figure 1C). The highest number of layers found on the overall PTC circumference was recorded and compared with the control group. The staining of PTC for complement product C4d as an indicator of antibody-mediated rejection was performed if tissue was available (Figure 1D).

Figure 1.

(A) Transplant glomerulopathy: Light microscopy (magnification × 400, PAS stain). There is widespread reduplication (→) of the glomerular basement membrane. (B) Transplant glomerulopathy: Electron microscopy (magnification ×2000). There is widening of the subendothelial space, a moderate increase of mesangial matrix, and mesangial interposition (→). (C) Peritubular capillary basement membrane multilayering (up to eight layers): Electron microscopy (magnification ×4000). (D) Immunofluorescent detection of C4d deposits in peritubular capillaries of renal allograft biopsies.

Indirect staining for C4d

This is performed using 3-μm frozen tissue sections. The tissue is fixed on the slide in acetone for 10 min and then washed with PBS for a minimum of two washes. Murine monoclonal anti-human C4d 100 μL (Quidel Corporation, San Diego, CA, USA) is added to the slides. BSA 1% is added to the negative control slides. Incubate for 20  min at room temperature. Add fluorescent antisera (Cy™2-conjugated AffiniPure Goat Antimouse IgG, Jackson ImmunoResearch Laboratories, Inc, West Grove, PA, USA) on all the sections and incubate for 30 min at room temperature. The slides are then washed in PBS, and GelTol™ (Immunon Thermo Shandon, Pittsburgh, PA, USA) media is used to mount the cover slides.

Statistical analysis

The statistics were analyzed using SPSS 11.0 for Windows. Unpaired t-test and Fisher's exact test were used to compare individual parameters between TGP and recurrent IgAN cases. Graft survival was analyzed by the Kaplan–Meier method and compared using the log-rank test. Results are expressed as mean ± SD. A P-value < 0.05 was considered statistically significant.

Results

Prevalence of glomerular diseases after kidney transplantation

Of 1156 patients followed up in the University of Alberta Hospital clinics during the study time period, glomerular disease was diagnosed by biopsy in 132 cases (11.4%) (Table 1). Glomerulonephritis was diagnosed in 91 transplants (7.9%), with IgAN being the commonest diagnosis [32 cases (2.8%)]. Twenty-seven cases (84.4%) were recurrent, two cases (6.2%) were de novo IgAN, and three cases (9.4%) were undetermined. Thirty-one cases were diagnosed TGP (2.7% of total followed up cases, 5% of total graft biopsies or 23.5% of those with glomerular abnormalities). On the basis of Banff 97 classification, 11 cases were mild TGP (Cg1), 17 were moderate (Cg2), and three cases were severe (Cg3).

Table 1. Prevalence of the glomerular disease after renal transplantation
Glomerular diseaseCases%
  1. GN = glomerulonephritis

All glomerulonephritis 91 7.9
IgA Nephropathy (recurrent, 27; de novo, 2; undetermined, 3) 32 2.8
Membranoproliferative GN 10 0.9
Membranous glomerulonephritis  7 0.6
Focal segmental glomerulosclerosis  6 0.5
Lupus nephritis  3 0.2
Pauci-immune GN  3 0.2
Unclassified GN 30 2.6
Transplant glomerulopathy 31 2.7
Diabetic glomerulosclerosis 10 0.9
Total13211.4

Comparison of the clinical data of TGP vs. recurrent IgA nephropathy

The clinical data of 31 TGP cases were compared with 27  recurrent IgAN cases (Table 2). Recipient gender, recipient age, donor type, donor age, years post-transplantation, years to diagnosis, serum creatinine at biopsy, and%PRA pretransplantation did not differ between TGP and recurrent IgAN. Transplant glomerulopathy was less frequent with tacrolimus and/or MMF, whereas recurrent IgAN showed no such tendency (P= 0.02). Twenty-three cases of TGP (74.2%) had significant proteinuria (>1 g/day), which was significantly greater than recurrent IgAN (P= 0.02). The amount of proteinuria in TGP ranged from 0.7 to 7.1 g/day (mean ± SD: 2.49 ± 1.47 g/day). In addition, nine cases (29%) of TGP had nephrotic range proteinuria (>3 g/day) and three of them (9.6%) had nephrotic syndrome, whereas only one case (3.7%) of recurrent IgAN had nephrotic range proteinuria (significantly different, P= 0.01). The degree of proteinuria was not significantly different (P= 0.28) between mild (Cg1: 1.9 ± 1.1 g/day) vs. moderate to severe TGP (Cg2–Cg3: 2.3 ± 0.9 g/day) as scored on the biopsy. Most cases of recurrent IgAN [25 cases (92.6%)] but few cases of TGP [five cases (16.1%)] had microscopic hematuria (P < 0.001). Hypertension was common in TGP and IgAN, 90.3% and 92.6%, respectively (P= 1.00).

Table 2. Comparison of the clinical data of transplant glomerulopathy vs. recurrent IgA nephropathy
VariableTGPRecurrent IgAN P
  1. 1Defined as treatment with antihypertensive drugs.

  2. TGP = transplant glomerulopathy, IgAN = IgA nephropathy.

Number3127 
Male/Female20/1116/110.79
Cadaveric/Living donor26/518/90.22
Recipient age (years)38.2 ± 12.937.0 ± 13.070.70
Donor age (years)32.9 ± 15.234.0 ± 14.10.80
Years post-transplantation8.3 ± 5.46.9 ± 4.40.32
Time of diagnosis, post KT (years)7.3 ± 4.85.7 ± 3.70.10
Serum creatinine at biopsy (μmol/L)298.4 ± 108.3287.7 ± 112.70.70
% PRA pretransplantation7.3 ± 16.29.7 ± 22.90.69
Hypertension128 (90.3%)25 (92.6%)1.00
Immunosuppressive drugs
 Calcineurin inhibitor27 (87.1%)27 (100%)0.10
 Steroid28 (90.3%)25 (92.6%)1.00
 MMF3 (9.7%)8 (29.6%)0.09
 Tacrolimus3 (9.7%)9 (33.3%)0.05
MMF and/or Tacrolimus5 (16.1%)12 (44.4%)0.02
Proteinuria (>1 g/day)23 (74.2%)11 (40.7%)0.02
Nephrotic range proteinuria9 (29.0%)1 (3.7%)0.01
Microscopic hematuria5 (16.1%)25 (92.6%)<0.001

Comparison of the prognosis of TGP vs. recurrent IgAN

For the renal allograft outcome of TGP, with the mean follow up of 9.3 years, 23 cases (72%) had deterioration of graft function and eventually 17 cases (53%) had graft failure. Seven cases (22.6%) had stable graft function. The graft survival for TGP was worse than for recurrent IgA nephropathy (after transplantation, P= 0.05 and after biopsy, P= 0.01) (Figure 2A,B).

Figure 2.

(A) Graft survival after transplantation of transplant glomerulopathy (TGP) compared with recurrent IgA nephropathy (IgAN) (P= 0.05). (B) Graft survival after biopsy of TGP compared with recurrent IgAN (P= 0.01).

Comparison of the pathological data of TGP vs. recurrent IgAN

The pathological data of TGP patients were compared with recurrent IgAN cases (Table 3). Ten cases of TGP (32.2%) and nine cases of recurrent IgAN (33.3%) had had T-cell-mediated rejection episodes in the past (not significantly, P= 1.0). Rejection in the current biopsy and CNI effect (arteriolar hyalinosis) did not differ between TGP and recurrent IgAN (P= 0.26 and 1.00, respectively). C4d staining of PTC was demonstrated in six of 24 cases (25%) of TGP but of the 27 cases of recurrent IgAN, C4d staining was available in 19 cases and none had C4d staining (significantly different, P= 0.02). In addition, PTC basement membrane multilayering was significantly increased in TGP (4.92 ± 2.94 layers) but not recurrent IgAN cases (1.86 ± 1.04 layers) (P < 0.001). The degree of PTC basement membrane multilayering in mild TGP (Cg1: 4.30 ± 1.63 layers) was not significantly different (P= 0.40) from moderate to severe TGP (Cg2–Cg3: 5.33 ± 3.56 layers). Therefore, the glomerular score in TGP did not correlate with the degree of PTC basement membrane multilayering.

Table 3. Comparison of the pathology of TGP vs. recurrent IgAN
 TGPRecurrent IgAN 
Variable(n = 31)(n = 27) P
  1. 1Peritubular capillary basement membrane multilayering.

  2. TGP = transplant glomerulopathy, IgAN = IgA nephropathy, PTC = peritubular capillary, CNI = calcineurin inhibitor, BM = basement membrane.

C4d deposits in PTC +/−6/18 (25%)0/19 (0%)0.02
PTCBM ML1 (layers)4.92 ± 2.941.86 ± 1.04<0.001
Both C4d (+) and PTCBM ML +/−5/18 (21.7%)0/19 (0%)0.05
Either C4d (+) and PTCBM ML +/−25/1 (96.1%)7/17 (29.2%)<0.001
History of T-cell rejection episodes10 (32.2%)9 (33.3%)1.00
Rejection in the current biopsy6 (19.3%)2 (11.1%)0.26
 tubulitis6 (19.3%)2 (11.1%)0.26
 arteritis2 (6.4%)1 (3.7%)1.00
CNI effect (hyalinosis)18 (58.1%)16 (59.2%)1.00

Discussion

In this study, we compared the clinical and pathological findings in TGP with those for recurrent IgAN, the commonest primary glomerulonephritis seen after kidney transplantation. The prevalence of the diagnosis of TGP is approximately 5–7% from previous reports (4–7), which is very close to our series (5% of total graft biopsies, 2.7% of all total followed-up cases during study period). A major clinical manifestation of TGP is proteinuria, whereas that of recurrent IgAN is microscopic hematuria. Previous studies have shown that patients with TGP often have heavy proteinuria and some cases developed nephrotic syndrome (1,4,6,17). In our series, most cases had significant proteinuria (>1 g/day), with an average protein excretion of approximately 2.5 g/day, and nephrotic range proteinuria in 30% of cases. Nephrotic syndrome was uncommon overall, affecting only 10% of TGP cases. In the current study, TGP developed with a minor degree of proteinuria (<0.5 g/day) in four cases (12.9%). This is in agreement with Habib et al. who demonstrated that TGP may develop in the absence or with a minor degree of proteinuria (11). We also demonstrated that the glomerular score in TGP did not correlate with the degree of proteinuria. We confirmed that microscopic hematuria is uncommon (16%) in TGP, as previously reported (11), unlike recurrent IgAN. Hypertension was observed in TGP as frequently as in recurrent IgAN.

The fact that TGP was less frequent with newer immunosuppression (MMF and/or tacrolimus), whereas recurrent IgAN showed no such tendency, supports our contention that TGP is a manifestation of inadequately controlled alloimmunity, and that it is becoming less frequent as immunosuppression improves. We acknowledge that other changes over time could contribute to this difference. The long mean time to development of TGP may reflect the infrequency of TGP in the recent era, which would increase the mean time to onset in the population as a whole. However, we admit that these explanations remain speculative, and large studies are needed. Previous studies demonstrated that MMF can inhibit in vitro antibody production and in vivo humoral response (18,19). Combination of MMF and tacrolimus can limit B-cell responses in renal allograft recipients with acute antibody-mediated rejection (20). Recently, Theruvath et al. demonstrated that donor-specific antibody production could be reduced in renal allografts with late antibody-mediated rejection by using this immunosuppressive combination (21). Whether MMF or tacrolimus is the more effective agent for inhibiting antidonor antibody synthesis is not clear. The fact that the incidence of recurrent IgAN is not changing in the same period serves as a control. The pathogenesis of IgAN is unknown, but IgAN has generally been poorly responsive to immunosuppressive interventions.

The association of TGP with PTC C4d deposition suggests that the cause of TGP is late alloantibody-mediated rejection in at least 25% of cases. The important of alloantibody in early graft injury has been clear for some time, with effects mediated in the microcirculation. Capillary deposition of C4d has been suggested as evidence for humoral alloreactivity against the graft (22,23). Feucht et al. described C4d, a complement split product of C4 of the classical pathway that can be detected in renal allograft (24,25). Although complement split-products proved valuable in early antibody-mediated rejection, its value in the diagnosis of late antibody-mediated rejection is less well established (22,23). The detection of donor-specific antibody in recipient sera and C4d deposits in PTC implicate antibody-mediated mechanisms in the pathogenesis of late, slowly progressive rejection (23). The%PRA and donor-specific antibody in the patients described here are under study. Some studies in animal models also support a role for antibody-mediated mechanisms (26–28). Shimizu et al., using a miniature swine model, demonstrated that both cell- and antibody-mediated rejection directed at the glomerular injury is an important determinant of acute glomerulopathy and progression to TGP (29). Recently, Joosten et al. demonstrated that Lewis (LEW) recipients of Fisher (F344) rat allografts produce IgG1 antibodies against donor-type perlecan and α1 (VI)/α5 (IV) collagen and develop some features of TGP (28), suggesting a role for autoantibody, but there is no proof of this in the human.

The association of TGP with PTC basement membrane multilayering suggests a generalized disorder of the graft microcirculation and its basement membrane. Both lesions show basement membrane thickening and reduplication, suggesting that they reflect similarities in pathogenesis, i.e. immunologic endothelial cell injury (13–15). Monga et al. first described PTC basement membrane multilayering in renal allografts showing TGP (12,13). This lesion is associated with TGP and the severity of PTC lesions correlated with the severity of TGP in some series (12,13,15,30). Moreover, PTC basement membrane multilayering may also be a reflection of antibody-mediated rejection (14,22,31,32). Recently, Regele et al. demonstrated a link of endothelial C4d deposition with TGP and PTC multilayering (16). In addition, capillary C4d deposition preceded the development of TGP in most patients in whom serial biopsies were available, suggesting a role for local complement activation in the development of TGP.

As a manifestation of alloimmunity, TGP should be regarded as a potentially treatable condition, with a goal of achieving remission or of slowing progression. The prognosis of TGP was worse than the prognosis of recurrent IgAN. Most cases of TGP (72%) had deterioration of graft function. Nevertheless, approximately 20% of cases stabilized. Transplant glomerulopathy is an important cause of late renal allograft dysfunction (7,9,33). Graft survival after biopsy is reduced in patients with TGP compared with control groups whose biopsies show allograft nephropathy without TGP (7). The persistent heavy proteinuria is the main factor correlating with deteriorating renal function, although the presence of nephrotic syndrome in TGP is not always associated with a poor outcome (11). On the other hand, a recent study revealed that the incidence of renal allograft loss at 10 years owing to recurrent IgAN was less than 10% (34). Potential treatments of TGP include reduction of proteinuria with angiotensin-converting enzyme (ACE) inhibitors and/or receptor antagonists, aggressive control of hypertension. In the cases in which antidonor antibody can be demonstrated, therapies to suppress antibodies (i.e. intravenous immunoglobulin, anti-CD 20, and plasmapheresis) should be tried, and a trial of increased or altered baseline immunosuppression (e.g. tacrolimus and MMF) can be considered. We do not know the efficacy of these treatments, and properly powered studies to evaluate these potential treatments are needed. If TGP is a manifestation of alloimmunity and capable of undergoing remission, then we should be optimistic about our chances to develop successful treatments. On the other hand, the opportunities for intervention in recurrent IgAN remain less clear, as is the care for IgAN in the native kidney.

Acknowledgments

We are grateful for grant support from the Canadian Institutes of Health Research, the Roche Organ Transplant Research Foundation, the Kidney Foundation of Canada, Novartis Pharmaceuticals Canada, Inc., The Muttart Foundation, and The Royal Canadian Legion.

Ancillary