Chronic rejection (CR) is the major cause of long-term graft loss. It usually presents as steady decline in glomerular filtration rate with biopsies showing signs of tubular atrophy, interstitial fibrosis, and intimal arterial wall proliferation among others.
Protocol-biopsy studies revealed that some degree of chronic rejection is already present at 1 year in 80–100% of the patients and moderate to severe forms are seen in 67% of the patients after 5 years (1).
Recently, with the routine use of biopsy staining for the split fraction of the complement C4d, it has become evident the role of antibody-mediated damage in chronic rejection. The more evident expression of it is the transplant glomerulopathy (TG) (2). However, it is not uncommon to see renal biopsies with CR, staining with C4d in peritubular capillaries (PTC), without clear evidence of glomerular lesions or multilayering of PTC basement membrane, at light microscopy.
In the same way, with the use of new methods, the detection of “de novo” antidonor anti-human leukocyte antigen (HLA) antibodies has become more frequent. A multicenter study in thousands of patients revealed that at least 10–15% develop de novo anti-HLA antibodies that may affect transplant outcome (3).
The Banff 2005 meeting report suggests that some forms of the so-called chronic allograft nephropathy (CAN), that present the triad of: 1) transplant glomerulopathy and/or peritubular capillary basement membrane multilayering and/or interstitial fibrosis/tubular atrophy and/or fibrous intimal thickening; 2) with the detection of bright C4d in more than half of PTC; and 3) evidence of donor specific antibodies should be now named as chronic active antibody-mediated rejection (CAMR) (4).
However, the frequency of C4d positive (C4d+) CR as well as its impact on the long-term graft outcome has not been demonstrated in a single study with a reasonable number of cases in a long-term follow-up. Furthermore, there is no evidence whether or not the current immunosuppressive regimens used to manage the former chronic allograft nephropathy modify the course of this kind of CAN.
The purpose of this study is to retrospectively review biopsy-proven cases of CAN, diagnosed between 1997 and 2003 according Banff 97 classification, with a long-term follow-up to define the incidence and impact of C4d+ biopsies on the graft outcome.
This study comprises all patients who received the diagnosis of “chronic allograft nephropathy” between 1997 and 2003 at our institution and fulfilled the inclusion criteria of: 1) biopsy-proven CAN (Banff 1997); 2) submitted to a change in the immunosuppression regimen after the diagnosis; and 3) with paraffin-embedded biopsy, at the time of diagnosis, currently available for C4d staining.
Patients with multiple-organ transplants and the diagnosis of transplant glomerulopathy by light microscopy were excluded.
CAN was suspected when there was renal function deterioration (not protocol biopsies) with no other evident causes.
All transplants were performed after a negative complement-dependent cytotoxicity (CDC) crossmatch using current and historic sera. Panel reactive antibodies (PRA) results were performed prior to transplantation using CDC until 2001 and enzyme-linked immunosorbent assay thereafter. Patients were followed-up until graft loss, defined as return to dialysis or retransplant, death, or December 31, 2006.
Conversion of Immunosuppressive Regimens
All patients were submitted to a conversion of their immunosuppressive regimens to manage CAN, at doctor’s discretion, and were converted to one of the following regimens (all with steroids):
- Low cyclosporine A (CyA)/mycophenolate mofetil (MMF): decreased CyA levels and full dose of MMF
- MMF alone: calcineurin inhibitors withdrawal and full dose of MMF
- Sirolimus (SRL)/low MMF: SRL plus low dose of MMF
- Low tacrolimus (TAC)/MMF: decreased TAC levels and full dose of MMF
Full MMF doses were 2000 mg/day adjusted according to side effects. Low MMF doses were 1000 mg/day or less. SRL levels were maintained between 5–12 ng/mL. Tacrolimus trough levels were 2–5 ng/mL and CyA C2 levels were 200–500 ng/mL.
The cohort of biopsies analyzed in this study had been examined before by two pathologists of our center and received the diagnosis of CAN according to Banff 97 criteria.
Cases with clear evidence of recurrence of native kidney disease as defined by the nature of native kidney disease and immune deposits in the graft, de novo glomerulonephritis, chronic obstruction, chronic lesions of previous bacterial and viral infection, and calcineurin inhibitors nephrotoxicity (arteriolar hyalinosis and/or isometric vacuolization in tubular cells) were differentiated from CR and excluded.
At our center, transplant glomerulopathy is diagnosed by light microscopy and defined as “double contours” of the glomerular basal membrane in at least 10% of the most severely affected tufts (5). Electron microscopy is not done routinely. The patients selected for this study had the biopsies at the time of diagnosis revised by a single pathologist (DRD), blinded to the previous histological diagnosis and to the current C4d staining. The Chronic Allograft Dysfunction Index (CADI) score was used to quantify the interstitial fibrosis and tubular atrophy and fibrous intimal thickening at the time of CR diagnosis (6). Only biopsies with ≥9 glomeruli were considered for evaluation of glomeruli C4d staining.
C4d staining was performed using the immunoperoxidase (IMP) method and classified according to the percentage of sampled peritubular capillaries (PTC) stained as: negative (<10% of staining in PTC) or positive (a strong staining in a range of PTC: 10–24%, 25–50% and >50%). Glomeruli C4d staining was classified from absent to 4+ (Fig. 1). Positive cases were evaluated in four different microscopy fields, counting a minimum of 20 PTC, starting in an area where there was a positive staining. The relative positive percentage staining was then calculated. Zero-hour biopsies of 10 patients, with an excellent 3-month outcome, were used as negative controls. All stained negative for C4d by IMP method.
Briefly, sections with 3 μm were deparaffinized and endogenous peroxidase activity blocked with hydrogen peroxide. Sections were stained with rabbit polyclonal anti-human C4d antibody (Biomedica, Wien, Austria) with dilution 1/50, overnight incubation at 4°C, after antigen retrieval using pressure cooking in 0.01 M citrate buffer at pH 6.0 for 2 min.
Detection was made with secondary antibody polymer peroxidase complex (Novolink Max Polymer, Novocastra, and Newcastle, UK). Diaminobenzidine was used as the chromogen. Counterstaining was carried out with Harris hematoxylin.
Study Design and Statistical Analysis
Patients were grouped according to whether their biopsies were positive or negative for PTC C4d. The primary end-point was death-censored graft survival evaluated up to December 31, 2006. Graft loss was defined as return to dialysis or retransplantation.
The first change in immunosuppressant with the purpose of treating CR was considered (intention-to-treat). The changes were decided at doctor’s discretion and not as random. The presence of proteinuria (>0.3 g/day) was used as a positive categorical variable. Data were analyzed by the General Linear Model adjusted by time after transplantation. Kruskal-Wallis test was used to compare non-parametric variables. Chi-square test was used to test frequency for categorical variables.
Graft survival curves, starting at the time of the diagnosis of CAN, were censored by death with a functioning graft and analyzed by Kaplan-Meier actuarial method. Overall graft survival curves (not censored for death) were also analyzed to exclude the effect of death on outcome comparisons. Survivals were compared by the log-rank analysis.
Cox proportional hazards model was used to analyze risk factors for death-censored graft loss. All variables were tested in a univariate analysis model and those with a P<0.05 participated in the multivariate analysis. The statistics software SPSS 14.0 was used for the statistical analysis. This retrospective analysis was approved by the Committee of Ethics in Clinical Research of this institution.
Between January 1997 and December of 2003, 116 patients with biopsy-proven CAN had been registered at our electronic database as being submitted to a change in the immunosuppressive regimen. One patient with this diagnosis had never been switched from his initial immunosuppressive regimen and was excluded. Of the remaining 115 patients, 82 cases (71%) had paraffin-embedded renal biopsies available for C4d immunoperoxidase staining and were enrolled in this study.
After the renal biopsies revision by light microscopy, two cases of transplant glomerulopathy were further identified by light microscopy and excluded. The remaining 80 cases participated in this analysis. Their biopsies showed different stages of tubular atrophy and/or interstitial fibrosis and/or intimal proliferation.
At the time of CAN diagnosis, patients have been transplanted by a median of 27 months. At the last follow-up (December 31, 2006), 9 patients had died and 25 had lost their grafts.
C4d staining was negative (<10% of PTC) in 30 cases (37.5%) and positive in 50 (62.5%). Positivity of C4d+ was observed in 10–24% of PTC in 18 patients (22.5%), in 25–50% of PTC in 13 patients (16%), and in >50% of PTC in 19 patients (24%). Out of the 80 renal biopsies, 8 had less than 9 glomeruli in the remaining biopsies and were not counted for C4d staining in glomeruli. In the other 72 biopsies, the intensity of C4d staining in glomeruli (0 to 3+) paralleled the extent of staining in PTC.
To verify which extent of PTC C4d staining should be considered as “positive” by immunoperoxidase method, we proceed to evaluate the impact of each range of C4d+ deposit in PTC on death-censored graft survival, after the CAN diagnosis. Figure 2 shows that any extent more than 10% of a strong staining of the PTC impacted on graft survival when compared to C4d-negative staining. Therefore, we have grouped patients into two groups: C4d+ (any positive staining >10%, n=50) and negative (C4d, n=30) for the remaining of this analysis.
Demographics, Transplant Data, and Risk Factors According to C4d Staining in PTC
Demographics, transplant data, and some risk factors of the C4d+ and C4d- groups are described in Table 1. In the C4d+ group, there were more females as recipients and more highly sensitized patients (PRA ≥50%).
Table 2 shows renal allograft parameters (clinical and histological) and the type of immunosuppressive regimen adopted at the time of CAN diagnosis in C4d- and C4d+ groups.
Total CADI score as well as all CADI isolate parameters did not differ between groups. In the same way, graft function was similar. However, among patients with data available for proteinuria, 53% of the C4d+ CR presented proteinuria in comparison with only 18% of C4d- (P=0.005).
In C4d- group, more patients were converted to a regimen of low CyA/MMF/steroids as compared to the C4d+ group, which was more evenly converted to the four different immunosuppressive regimens. Due to the small number of patients in each regimen and a nonrandomized choice of regimens, we decided not to analyze the impact of each type of conversion on graft outcome.
We had analyzed all categories of C4d staining, according to C4d immunoperoxidase staining (negative, 10–24%, 25–50%, and 51–100%), but we did not find any difference. So, we analyzed our data into C4d+ and C4d- group.
Outcomes and Risk Factors Analysis for Death-Censored Graft Survival
Death-censored graft survival was better in C4d- than in C4d+ cases and the difference was highly significant (log rank P=0.002; Fig. 3A). In the same way, overall graft survival (not-censored for death) was also diminished in the C4d+ group (log rank P=0.032; Fig. 3B).
Table 3 shows the Cox proportional hazards for death-censored graft loss by univariate and multivariate analysis of the many study variables.
Among the clinical parameters, a better graft function evaluated by serum creatinine at CR diagnosis was protective. The positivity for C4d in PTC, sensitized patients with anti-HLA antibodies and previous pregnancies, the presence of proteinuria and higher CADI interstitial scores as well as intimal proliferation at the time of CR diagnosis were all identified as hazards for graft loss in univariate analysis.
When all these variables were analyzed together in the multivariate analysis only the C4d+ in PTC, sensitized patients (female patients with previous pregnancies and high PRA) and intimal proliferation popped-up as hazards.
This study reveals a high incidence of C4d+ CR. It also shows that C4d+ CR have a much worse prognosis than C4d- CR. Although retrospective, a new prospective analysis of such cases would take many years to furnish the evidence that can be drawn from this study.
The first challenge of this retrospective analysis was to define the minimal intensity of C4d that should be considered positive by the IMP method. There is no consensus on what extent of PTC C4d staining should be considered positive by IMP. Data from studies using immunofluorescence (IF) defined that only >25% of C4d+ PTC should be considered as positive (7).
Although different from what has been defined by IF, we assumed that any strong staining in ≥10% of PTC should be considered positive by IMP. This definition was fourfold: 1) we tested 10 zero-biopsies of patients who had an excellent 3-month outcome and all stained negative for C4d by IMP (data not shown); 2) in 28 recent cases analyzed concomitantly by IF and IMP, all 3 cases with IF <25% in PTC resulted in IMP negative; 3) the survival curves of the different percentages of stained PTC did not show differences among them but all differed statistically from the C4d- group; and 4) comparison of IMP and IF methods showed IMP as much less sensitive (8, 9) deducing that an extent of staining by IMP is probably larger by IF. The same criteria have been recently used by others (10). Therefore, we believe that the criteria used really differentiated the two groups and shows that any positive (≥10%) staining of PTC should be considered relevant by IMP.
C4d staining in late biopsies is usually reported in transplant glomerulopathy (TG). Regele et al. found an incidence of 34% of CR associated with C4d in PTC, mostly with histology of TG; the cases without glomerular lesions developed them in the follow-up biopsies (11). Similar findings were reported by other authors (12, 13). Nevertheless, TG identified by light microscopy in late biopsies is an uncommon finding, encountered in around 1.6 to 15% of transplants with chronic rejection. When we reanalyzed by light microscopy the 82 cases of CR, only 2 (1.6%) showed features of TG and the other 80 cases had no evidence of glomerular lesions by light microscopy. Considering the high correlation between the staining of C4d in PTC and glomeruli found in this study, we are currently analyzing whether electronic microscopy of such cases is necessary to reveal glomerular lesions and/or multilayering of PTC basement membrane or whether C4d+ is a surrogate marker of such lesions. Nevertheless, none of the 80 analyzed cases had features of TG and still the majority of them were C4d+.
We were surprised by the high incidence of C4d+ in our series (62%), but this seems to be in agreement with others. In the 31 cases of CR reported by Worthington et al., the percentage of C4d+ analyzed by IMP was 42% (10), whereas Bocrie et al. found 75% in 12 CR cases (14). Mauyyedi et al. looked at C4d in 38 cases of chronic rejection and found that 61% of them had PTC C4d+ (2). Herman et al. analyzed late biopsies with features of CR in pediatric recipients and found 50% of C4d+; Ishii et al. encountered 42.9% (15). Other authors report less frequent but still very high frequency of C4d+ in CR. Mroz et al. evaluated 26 patients with CR who underwent biopsies. The frequency of C4d in PTC was 30%, but their cases had features of chronic TG. Our series, as well as all the above-mentioned studies, suggest that the incidence of C4d+ CR is high, ranging from 30% to 75%, depending upon the method and criteria used.
The other aspect covered in this study was the impact of C4d staining on the transplant outcome. We observed that C4d+ CR has a much poorer prognosis than C4d- CR. Similar findings were recently described by others (2, 10). Our data identified that the major risk factors for death-censored graft loss in the multivariate analysis were the presence C4d+ in the indication biopsy and previous HLA sensitization. These risk factors are usually related to antibody mediated injury to the graft. The presence of anti-HLA antibodies is now known as frequent and associated with graft loss (3). Anti-HLA antibodies require activation of the complement cascade through membrane attack complex to produce injury (16).
Although we have not searched for anti-HLA antibodies in this retrospective study (there were no donor cells and/or many recipients sera to evaluate this aspect), the presence of C4d+ in PTC and glomeruli may indicate that antidonor HLA (and possibly other) antibodies are participating in the mechanism of graft destruction because these cases are frequently associated with antidonor antibodies (2, 10, 17). In cynomolgus monkeys submitted to several protocols of renal allograft with mixed chimerism, all animals showing C4d deposits in renal biopsies developed alloantibodies and failed after 3 to 27 months (18).
There is also data to suggest that finding circulating donor specific antibodies (DSA) may not be as easy as required in the 2005 Banff criteria for the diagnosis of chronic active antibody mediated rejection (CAMR), even with the latest technology. Bocrie et al., using Luminex assay, extracted donor-specific HLA antibodies from the cortex and medulla of 12 renal allograft lost due CR. In 75% of the cases, they found anti-HLA antibodies in the tissue, with 58% against the donor HLA antigens. In all these cases, PTC C4d+ deposits were seen. Sera obtained at the time of the transplantectomy revealed only 16% of circulating DSA (14). This confirms a strong relationship between C4d deposits and the presence of antidonor HLA antibodies, and challenges the need to detect circulating antibodies to define CAMR.
In fact, the last Banff meeting defined as CAMR the triad of: 1) interstitial fibrosis/tubular atrophy and/or fibrous intimal thickening in arteries (and/or TG and PTC multilayering of basement membrane); 2) C4d+; and 3) DSA (4). Therefore, in our analysis, due to the lack of DSA, the C4d+CR can only be classified as “suspicious” CAMR, whereas the C4d- CR cases are a range of chronic allograft arteriopathy and/or interstitial fibrosis/tubular atrophy.
In this analysis, fibrosis and intimal proliferation of smooth muscle-cells of the arterial wall were also findings of bad prognosis. The development process of CAMR is characterized by injury and progressive loss of identifiable peritubular capillaries (PTCs) accompanied with the development of interstitial fibrosis (15). Injured PTCs by antibodies and complement activation may initiate the process of angioregression with lamination of the basement membrane and loss of PTCs (11, 15).
Another important aspect of our study is the fact that the current change in immunosuppression to treat CR seems not to modify the course of C4d+ group. However, due to the small number of patients, we avoided drawing conclusions on the course of C4d+ CR within each of the four regimens used to manage C4d+ CR.
In summary, our data shows that C4d+ CR (or “suspicious” CAMR) is a frequent entity that has a poor outcome when managed with the current therapeutic strategies. Our data also suggest that we should, from now on, follow patients with protocol biopsies for early diagnosis of CAMR, as well as frequently search sera for the presence of antidonor antibodies. Larger trials are warranted to search for new interventions to treat this specific condition.
1. Nankivell BJ, Borrows RJ, Fung CL, et al. The natural history of chronic allograft nephropathy
. N Engl J Med
2003; 349: 2326.
2. Mauiyyedi S, Pelle PD, Saidman S, et al. Chronic humoral rejection: identification of antibody-mediated chronic renal allograft rejection by C4d deposits in peritubular capillaries. J Am Soc Nephrol
2001; 12: 574.
3. Terasaki PI, Ozawa M. Predicting kidney graft failure by HLA antibodies: a prospective trial. Am J Transplant
2004; 4: 438.
4. Solez K, Colvin RB, Racusen LC, et al. Banff ’05 Meeting Report: differential diagnosis of chronic allograft injury and elimination of chronic allograft nephropathy
(’CAN′). Am J Transplant
2007; 7: 518.
5. Racusen LC, Solez K, Colvin RB, et al. The Banff 97 working classification of renal allograft pathology. Kidney Int
1999; 55: 713.
6. Isoniemi H, Taskinen E, Hayry P. Histological chronic allograft damage index accurately predicts chronic renal allograft rejection. Transplantation
1994; 58: 1195.
7. Nickeleit V, Zeiler M, Gudat F, et al. Detection of the complement degradation product C4d in renal allografts: Diagnostic and therapeutic implications. J Am Soc Nephrol
2002; 13: 242.
8. Seemayer CA, Gaspert A, Nickeleit V, Mihatsch MJ. C4d staining
of renal allograft biopsies: A comparative analysis of different staining techniques. Nephrol Dial Transplant
2007; 22: 568.
9. Nadasdy GM, Bott C, Cowden D, et al. Comparative study for the detection of peritubular capillary C4d deposition in human renal allografts using different methodologies. Hum Pathol
2005; 36: 1178.
10. Worthington JE, McEwen A, McWilliam LJ, et al. Association between C4d staining
in renal transplant biopsies, production of donor-specific HLA antibodies, and graft outcome. Transplantation
2007; 83: 398.
11. Regele H, Bohmig GA, Habicht A, et al. Capillary deposition of complement split product C4d in renal allografts is associated with basement membrane injury in peritubular and glomerular capillaries: A contribution of humoral immunity to chronic allograft rejection. J Am Soc Nephrol
2002; 13: 2371.
12. Mroz A, Lewandowski Z, Cieciura T, et al. C4d complement split product in diagnosis of immunological activity of chronic allograft nephropathy
. Transplant Proc
2006; 38: 97.
13. Sijpkens YW, Joosten SA, Wong MC, et al. Immunologic risk factors and glomerular C4d deposits in chronic transplant glomerulopathy. Kidney Int
2004; 65: 2409.
14. Bocrie O, Hussein Aly AA, Guignier F, et al. Distribution of donor-specific antibodies in the cortex and the medulla of renal transplants with chronic allograft nephropathy
. Transpl Immunol
2007; 17: 227.
15. Ishii Y, Sawada T, Kubota K, et al. Injury and progressive loss of peritubular capillaries in the development of chronic allograft nephropathy
. Kidney Int
2005; 67: 321.
16. Qian Z, Lee CY, Murata K, et al. Antibody and complement mediated injury in transplants following sensitization by allogeneic blood transfusion. Transplantation
2006; 82: 857.
17. Schwarz C, Regele H, Huttary N, et al. Rescue therapy with tacrolimus and mycophenolate mofetil does not prevent deterioration of graft function in C4d-positive chronic allograft nephropathy
. Wien Klin Wochenschr
2006; 118: 397.
18. Smith RN, Kawai T, Boskovic S, et al. Chronic antibody mediated rejection of renal allografts: Pathological, serological and immunologic features in nonhuman primates. Am J Transplant
2006; 6: 1790.