SEARCH

SEARCH BY CITATION

Keywords:

  • Antibody-mediated rejection;
  • C4d;
  • HLA;
  • kidney;
  • renal allograft rejection;
  • transplantation

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of Interest Statement
  9. References
  10. Supporting Information

We studied the phenotype of late kidney graft failure in a prospective study of unselected kidney transplant biopsies taken for clinical indications. We analyzed histopathology, HLA antibodies and death-censored graft survival in 234 consecutive biopsies from 173 patients, taken 6 days to 31 years posttransplant. Patients with late biopsies (>1 year) frequently displayed donor-specific HLA antibody (particularly class II) and microcirculation changes, including glomerulitis, glomerulopathy, capillaritis, capillary multilayering and C4d staining. Grafts biopsied early rarely failed (1/68), whereas grafts biopsied late often progressed to failure (27/105) within 3 years. T-cell-mediated rejection and its lesions were not associated with an increased risk of failure after biopsy. In multivariable analysis, graft failure correlated with microcirculation inflammation and scarring, but C4d staining was not significant. When microcirculation changes and HLA antibody were used to define antibody-mediated rejection, 17/27 (63%) of late kidney failures after biopsy were attributable to antibody-mediated rejection, but many were C4d negative and missed by current diagnostic criteria. Glomerulonephritis accounted for 6/27 late losses, whereas T-cell-mediated rejection, drug toxicity and unexplained scarring were uncommon. The major cause of late kidney transplant failure is antibody-mediated microcirculation injury, but detection of this phenotype requires new diagnostic criteria.


Abbreviations: 
ABMR

antibody-mediated rejection

DSA

donor-specific HLA antibodies

IFTA

interstitial fibrosis and tubular atrophy

PRA

panel-reactive HLA antibodies

TCMR

T-cell-mediated rejection

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of Interest Statement
  9. References
  10. Supporting Information

Kidney transplants often progress to late failure, contributing to the burden of end-stage renal disease (1). The causes and phenotypes of late graft failure are not understood. Many kidney transplants presenting with slow deterioration of function never undergo biopsies, and those that are biopsied are classified by imprecise terms such as ‘chronic rejection’ or ‘chronic allograft nephropathy’ (CAN) (2,3). CAN was replaced in the consensus Banff classification by ‘interstitial fibrosis and tubular atrophy’ (IFTA), acknowledging that scarring is not a disease but a feature of all progressive kidney diseases, and that specific diseases causing scarring should be sought (4). Moreover, calcineurin inhibitor toxicity contributes to injury and scarring in transplants (5) and native kidneys (6), but its actual contribution to late kidney graft failure is unknown. Recurrent diseases such as glomerulonephritis account for a minority of graft losses (7,8).

The recent demonstration that circulating donor-specific antibodies (DSA) against HLA antigens increase the risk of late graft loss (9) argues for a major role of antibody-mediated rejection (ABMR). ABMR was originally recognized by the presence of DSA and microcirculation injury (10), including electron microscopic changes (11), but is currently defined by deposition of complement factor C4d in peritubular capillaries (12–16). However, C4d may not be sensitive: many cases of transplant glomerulopathy with anti-HLA are C4d negative (17–19). Therefore, the recent update of the Banff classification introduced the diagnostic category ‘suspicious for ABMR’ if C4d (in the presence of antibody) or alloantibody (in the presence of C4d) cannot be demonstrated but morphologic evidence of antibody-mediated tissue injury is present (20). Moreover, in microarray studies, expression of endothelial transcripts was increased in biopsies with DSA and associated with graft loss even in C4d negative biopsies (21).

In the present study, we hypothesized that ABMR was the main cause of late graft failure, but that much of it is missed by current criteria. We reasoned that, in addition to C4d positive kidneys with ABMR features (C4d+ABMR), kidneys with anti-HLA plus microcirculation changes but lacking C4d (C4dABMR) are also at risk for failure. We prospectively and completely analyzed 234 kidney transplant biopsies for clinical indications, including histopathology, electron microscopy, anti-HLA and outcomes. This unselected cross section of kidney transplants with clinical problems at various times posttransplant allows us to describe the phenotypes of kidney transplants that progress to failure and their relative frequency.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of Interest Statement
  9. References
  10. Supporting Information

Patients and sample collection

The study was approved by the institutional review board of the University of Alberta (issue 5299). All patients undergoing clinically indicated biopsies between September, 2004 and October, 2007 were included after written informed consent. Biopsies were obtained under ultrasound guidance by spring-loaded needles (ASAP Automatic Biopsy, Microvasive, Watertown, MA). Most patients were recruited at the University of Alberta, which performs approximately 70–90 kidney transplants per year, with 1200 kidney transplant patients followed in the outpatient clinics, and performs approximately 200 renal transplant biopsies for cause per year. All patients are followed by transplant nephrologists in the outpatient transplant clinic of the University of Alberta Hospital at regular intervals, at least every 3 months. At the University of Illinois, 120 transplants are performed each year. The rejection rate at both centers is 10–15%.

Histopathology and diagnostic classifications

Paraffin sections were graded by Banff criteria by a renal pathologist (BS) (4,20,22,23). All samples were adequate (24) except seven biopsies lacking arteries. Interstitial inflammation was scored both in nonscarred and scarred cortical parenchyma (20). C4d staining was performed on frozen sections (available in 223 biopsies) by indirect immunofluorescent staining using a monoclonal anti-C4d antibody (100 μL, 30-min incubation; Quidel Corporation, San Diego, CA), followed by fluorescent antisera (30 min, Cy2-conjugated affinity purified goat anti-mouse IgG, Jackson ImmunoResearch Laboratories, Inc. West grove, PA) (17). Staining was scored according to Banff 2007 (20). Diffuse linear C4d staining (>50% of biopsy area) was interpreted as positive (20). Peritubular capillary basement membrane multilayering was examined by electron microscopy (available for biopsies > 3 months posttransplant). Cases showing more than five basement membrane layers were considered positive (25).

Banff diagnoses (4,22,23) included T-cell- mediated rejection (TCMR), borderline TCMR, antibody-mediated rejection (C4d+ABMR), transplant glomerulopathy, calcineurin inhibitor toxicity, BK nephritis, acute tubular necrosis and interstitial fibrosis and tubular atrophy not otherwise specified (IFTA). Diffuse C4d+ biopsies meeting the criteria for ABMR as well as TCMR were called mixed C4d+ABMR plus TCMR (n = 4). Transplant glomerulopathy was diagnosed by double contours of glomerular basement membrane. Biopsies showing BK nephritis (confirmed by in situ hybridization and/or electron microscopy) were designated as BK (n = 6), regardless of histological signs of TCMR (n = 1).

HLA antibodies

Screening was performed using FlowPRA® beads representing HLA-A, -B, -Cw -DR -DQ and -DP antigens. Further testing for specificities was only done if screening was positive (≥5% PRA or clear pattern of reactivity with screening beads). Antibody specificities of patient sera were determined by FlowPRA® Specific class I and or II and/or FlowPRA® single antigen I and II beads (One Lambda, Canoga Park, CA). Manufacturer's instructions for staining and acquiring were followed. Beads were analyzed on a BD FACSCalibur™ cytometer (Becton Dickinson Biosciences, Mississauga, Ontario, Canada). A result was considered positive when a bead population showed a clear reactivity with a positive shift in fluorescence when compared to negative control sample. Single antigen beads were used to test for antibodies against HLA-A, B, DRB1, DRB3, 4 and 5, DQB1 and DP. We did not test for specificities to Cw. Donor typing for DP was not performed and therefore DSA were not attributed to DP. DSA was defined as by single antigen bead technology or a donor specific flow crossmatch that was negative pretransplant and positive at biopsy. Flow T- and B-cell crossmatches were previously described (26).

Data analysis

We used SPSS 14.0 statistical software package (SPSS Inc., Chicago, IL) and Bioconductor version 1.9, R version 2.4 (http://www.bioconductor.org) (27). For comparisons between groups (early vs. late biopsies, failed vs. functioning grafts) t-tests were used for means, while chi-square or Fisher's exact tests were used for count data.

Graft failure was defined as return to dialysis. Death-censored graft survival analysis was performed using the Kaplan–Meier method. Patients were censored for the end of study (October 21, 2008), death with functioning graft (n = 5), or lost to follow-up (n = 1). The influence of clinical variables and Banff lesions on graft survival was analyzed by univariable and multivariable Cox regression. HLA antibody status, C4d staining, glomerulopathy, intimal arteritis, arteriolar hyalinosis and fibrous intimal thickening were analyzed as binary variables (score = 0 vs. score >0; in the case of C4d staining C4d grade 3 vs. <3). Proteinuria was defined as positive if the dipstick result was greater than ‘trace’ and analyzed as binary variable (positive vs. negative). Glomerular filtration rate (GFR) and time posttransplant were analyzed as continuous log-transformed data, and all other variables were analyzed as ordinal variables (scores 0, 1, 2, 3). Significance was assessed by the log rank test. We included all variables with a p-value < 0.05 in the univariable analysis.

All biopsies (n = 234) were included in the analysis of HLA antibody status, histologic lesions and diagnosis to provide an overview of the distribution of lesion scores as it is observed in a biopsy population in the clinic. To exclude that the frequency especially of chronic lesions might be influenced by repeat biopsies, we repeated the analyses using only one biopsy (the last) per patient. Analyses assessing the effects on graft survival and comparisons between failed versus functioning grafts were limited to the last biopsy for each patient (n = 173), excluding repeat biopsies so that the same event would not be analyzed more than once per patient. The number of biopsies included in each analysis is specified in the figures and tables.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of Interest Statement
  9. References
  10. Supporting Information

Patient demographics

We prospectively studied all consenting patients undergoing kidney transplant biopsies for clinical indication (n = 234 biopsies in 173 patients) during the study period, with no selection (Table 1). Patients were biopsied from 6 days to 31 years posttransplant (median 17.0 months; 2.6 months for early, 20.2 months for late biopsies), and were followed up to 49 months post biopsy (median 27.8 months; 27.4 for early biopsies; 28.2 for late biopsies). Indications for biopsy included loss of function (59%), impaired function (12%) and proteinuria (11%) and were similar in biopsies taken before (n = 106) or after 1 year (n = 128).

Table 1.  Patient demographics and clinical characteristics at the time of biopsy
Patient demographics (n = 173)All patients (n = 173)Patients biopsied <12 months (n = 68)Patients biopsied >12 months (n = 105)p-Value (early vs. late)
Recipient gender (% male) (n = 173)107 (62%)42 (62%)65 (62%)0.99
Race (n = 173)
 Caucasian99 (57%)35 (51%)64 (61%)0.22
 Black18 (10%)8 (12%)10 (10%)0.64
 Other26 (15%)10 (15%)15 (14%)0.94
Primary disease (n = 173)
 Diabetic nephropathy25 (14%)12 (18%)13 (12%)0.63
 Hypertension/large vessel disease17 (10%)11 (16%)6 (6%)0.02
 Glomerulonephritis/vasculitis76 (44%)25 (37%)51 (49%)0.13
 Interstitial nephritis/pyelonephritis14 (8%)6 (9%)8 (8%)0.78
 Polycystic kidney disease25 (14%)11 (16%)14 (13%)0.60
 Others8 (5%)1 (1%)7 (7%)0.15
 Unknown etiology8 (5%)2 (3%)6 (6%)0.48
Previous transplant (n = 173)22 (13%)10 (15%)9 (9%)0.21
Donor gender (% male) (n = 144)67 (47%)26 (38%)41 (39%)0.91
Donor type (% deceased donor transplants) (n = 173)95 (55%)38 (56%)57 (54%)0.84
Clinical characteristics at the time of biopsy (n = 234)All biopsies (n = 234)Biopsies taken <12 months (n = 106)Biopsies taken >12 months (n = 128)p-Value (early vs. late)
Maintenance immunosuppressive regimens at biopsy
 MMF, tacrolimus, steroid102 (44%)54 (51%)48 (38%)0.04
 MMF, tacrolimus7 (4%)6 (6%)1 (1%)0.05
 MMF, cycbsporine, steroid56 (32%)22 (21%)34 (27%)0.30
 MMF, steroids5 (3%)2 (2%)3 (2%)1.0
 Azathioprine, cyclosporine, steroids21 (12%)0 (0%)21 (16%)<0.001
 Others43 (25%)22 (21%)21 (16%)0.39
Indication for biopsy
 Primary nonfunction8 (3%)8 (8%)0 (0%)0.002
 Rapid deterioration of graft function48 (21%)17 (16%)31 (24%)0.12
 Slow deterioration of graft function88 (38%)36 (34%)52 (41%)0.29
 Stable impaired graft function28 (12%)20 (19%)8 (6%)0.003
 Investigate proteinuria25 (11%)8 (8%)17 (13%)0.16
 Follow-up from previous biopsy13 (6%)7 (7%)6 (5%)0.52
 Others10 (14%)3 (3%)7 (5%)0.52
 Indication unknown14 (6%)7 (7%)7 (5%)0.72

Death-censored graft survival after biopsy

Graft failure with return to dialysis occurred in 28 of 173 biopsied patients. (Five patients died with functioning grafts.) Failure after biopsy was strikingly related to the time of the biopsy posttransplant, i.e. the time at which clinical indications for biopsy appeared (Figure 1A). Twenty-seven failures occurred in 105 patients biopsied after 1 year, compared to one failure in the 68 patients biopsied less than 1-year posttransplant. Of the 28 patients with graft failure, only one had an early biopsy (at day 360) and then a late biopsy within the study period. Thus the low incidence of progression to failure in kidneys after an early biopsy is likely due to the fact that different diseases operate in grafts biopsied early compared to those biopsied late: diseases operating in grafts biopsied early are more amenable to treatment. Surprisingly, all graft losses after a biopsy for cause occurred within 3 years after biopsy (although more are likely with continued follow-up).

imageimageimageimage

Figure 1. Antibody status, histologic lesions, histopathologic diagnosis and graft loss in early and late biopsies. (A) Death-censored graft survival was analyzed in all patients (n = 173), comparing patients biopsied early (<12 months posttransplant, n = 68) and patients biopsied late (>12 months posttransplant, n = 105). In the case of multiple biopsies from one patient, only the last biopsy was used. (B) HLA antibody status at the time of biopsy was available for n = 206 biopsies and is shown separately for early (n = 94) and late (n = 112) biopsies (p < 0.05). (C) Histologic lesion scores were summarized as the average lesion score in all biopsies (n = 234), early biopsies (n = 106) and late biopsies (n = 128). Differences between early and late biopsies were assessed by t-test (*p < 0.001). C4d staining was positive in 24 biopsies, mostly in late (21/128) rather than early biopsies (3/106) (p = 0.002). (D) Lesion scores were interpreted as histopathologic diagnoses based on current Banff classification criteria. The frequency of diagnoses is shown for all biopsies, early biopsies and late biopsies. Comparisons between early and late biopsies were done by chi-square or Fisher's exact test. P-values are indicated where significant (p < 0.05).

Clinical and histopathologic features at the time of biopsy

HLA antibody status:  Circulating PRA and DSA were more frequent in patients biopsied >1-year posttransplant (Figure 1B). Compared to patients biopsied early, those biopsied late more frequently displayed PRA (late 64%, early 24%, p < 0.001) and DSA (late 48%, early 7%, p < 0.001). Differences were particularly striking for class II PRA (late 49%, early 9%, p < 0.001) and DSA (late 47%, early 3%, p < 0.001). Anti class I was less frequent than anti class II but was more common in patients undergoing late biopsies, in both PRA (late 47%, early 19%, n.s.) and DSA (late 21%, early 5%, n.s.). Thus patients having late biopsies were more likely to have anti-HLA, particularly class II.

Biopsy lesions and diagnoses:  We focused first on histologic lesions rather than diagnoses (17,20,22,23). Many histologic lesions were impacted by time posttransplant (Figure 1C). Microcirculation changes (glomerulitis, capillaritis, transplant glomerulopathy and capillary multilayering), scarring, arterial fibrous intimal thickening (i.e. fibrous intimal thickening without signs of immunological activity) and arteriolar hyalinosis were more frequent in late biopsies, whereas tubulitis, interstitial inflammation and intimal arteritis were not. No biopsies displayed arterial changes suggesting previous arteritis. C4d staining was diffusely positive in 24 biopsies, mostly late chronic-active ABMR (21/128) rather than early acute ABMR (3/106) (p = 0.002). Twelve cases showed focal C4d staining (Banff C4d grade 2); no minimal focal cases (Banff C4d grade 1) were observed. According to the Banff classification, we only considered diffuse C4d staining (Banff grade 3) positive.

The late biopsies were more frequently assigned diagnoses of C4d+ABMR (acute or chronic-active), transplant glomerulopathy, glomerulonephritis and calcineurin inhibitor toxicity than early biopsies (Figure 1D).

Since the frequency of chronic lesions or the percentage of patients with alloantibody might be influenced by including repeat biopsies from the same patients we repeated the analyses presented in Figure 1B–D using only one biopsy (the last) per patient. The results were almost identical to the results when including all biopsies.

Biopsy features that predicted progression to graft loss

We assessed the association of histologic Banff lesions, HLA antibody status and clinical variables with graft survival by univariable and multivariable analysis (Table 2). Given the differences between early and late biopsies and the rarity of graft loss after early biopsies, we limited this analysis to late biopsies, using only the last biopsy for each patient (n = 105).

Table 2.  Univariable and multivariable analysis of features associated with graft loss
Biopsies >1 year (only last biopsy for each patient) (n = 105)Biopsies >1 year, GN excluded (only last biopsy for each patient) (n = 80)
Univariable analysis* FeatureHazard ratio (95% CI)p-ValueFeatureHazard ratio (95% CI)p-Value
Proteinuria10.9 (2.6–46.5)<0.001Proteinuria10.7 (2.50–46.5)<0.001
GFR0.95 (0.93–0.98)<0.001GFR95 (0.92–0.98)<0.001
ci2.20 (1.34–3.59)0.001ci2.71 (1.48–4.95)<0.001
g2.04 (1.37–3.04)0.002g2.14 (1.33–3.43)0.004
mm1.72 (1.15–2.58)0.008ct2.09 (1.17–3.74)0.01
ptcml1.98 (1.16–3.38)0.010ptcml1.97 (1.09–3.55)0.02
ct1.86 (1.14–3.02)0.01mm1.65 (1.06–2.58)0.03
ptc1.53 (1.10–2.13)0.02ptc1.47 (1.02–2.13)0.04
DSA II+2.62 (1.14–5.99)0.02DSA II+2.49 (0.99–6.24)0.05
ah0.32 (0.15–0.72)0.01C4d2.64 (1.05–6.65)0.05
i1.63 (1.03–2.57)0.04i1.55 (0.94–2.57)0.09
C4d2.36 (0.99–5.64)0.07cg2.01 (0.85–4.77)0.11
PRA II+1.86 (0.80–4.35)0.14PRA II+1.78 (0.68–4.62)0.23
cg1.7 (0.79–3.63)0.17v1.74 (0.51–5.90)0.41
v1.7 (0.51–5.67)0.42ah0.48 (0.18–1.30)0.18
cv0.75 (0.75–0.26)0.61DSA I+1.47 (0.56–3.83)0.45
DSA I+1.30 (0.51–3.27)0.59Time post Tx1.40 (0.37–5.30)0.62
t1.12 (0.72–1.73)0.63t1.13 (0.71–1.80)0.63
PRA I+0.92 (0.41–2.06)0.84cv1.31 (0.30–5.60)0.71
Time post Tx1.09 (0.35–3.36)0.89PRA I+0.89 (0.37–2.15)0.80
Multivariable analysis FeatureHazard ratiop-ValueFeatureHazard ratiop-Value
  1. *C4d staining, cg v, ah and cv scores, proteinuria, PRA and DSA were analyzed as binary variables (0 vs. >0); all other variables were analyzed as ordinal variables (0, 1, 2, 3).

g3.6 (1.8–6.9)0.002GFR0.93 (0.88–0.98)0.004
Proteinuria12.8 (2.5–64)0.002ci13.7 (1.8–104)0.006
GFR0.96 (0.93–0.99)0.003g4.0 (1.2–13.6)0.02
ah0.22 (0.08- 0.62)0.004Proteinuria12.5 (1.1–144)0.04
i2.4 (1.1–5.2)0.02   
ci2.0 (1.1–3.7)0.03   

In univariable analysis, class II DSA and lesions related to microcirculation changes, as well as inflammation, scarring, arteriolar hyalinosis, intimal fibrosis, GFR and proteinuria were associated with graft loss. In multivariable analysis, glomerulitis, interstitial inflammation and scarring were independent predictors of graft loss in addition to GFR and proteinuria. Arteriolar hyalinosis was associated inversely with graft loss.

To assess whether these relationships were influenced by including glomerulonephritis cases, we repeated the analysis after excluding these 25 cases. In univariable analysis, graft loss was associated with anti class II DSA, scarring, and microcirculation changes including C4d. The association of hyalinosis and inflammation with outcome disappeared. The relationship of C4d staining with graft loss was borderline significant (p = 0.052) in univariable analysis and not significant in multivariable analysis. Focal C4d did not increase the probability of graft loss: only one case of 12 with focal C4d staining progressed to failure; this case was diagnosed as BK nephropathy.

Repeating the multivariate analysis without GFR and proteinuria, we obtained very similar results. When including all last late biopsies, the significant features are g-score (hazard ratio 2.11 [1.39–3.18], p < 0.001), CI-score (hazard ratio 2.40 [1.44–3.99], p < 0.001), and ah-score (hazard ratio 0.21 [0.09–0.50] p < 0.001). After removing GN cases, the significant features were ci-score (2.39 [1.28–4.44], p = 0.006) and g-score (hazard ratio 1.73 [1.09–2.75], p = 0.02).

Thus the associations with graft loss were microcirculation changes particularly glomerulitis, as well as scarring. The lesions indicating microcirculation changes are strongly correlated with one another as well as with alloantibody which may explain why only the g-score but not other microcirculation changes were significant in multivariate analysis. The lesions of TCMR (tubulitis, interstitial infiltration, intimal arteritis) and calcineurin inhibitor toxicity (hyalinosis) were not associated with increased risk of graft loss.

Characteristics of failed grafts in the late biopsy group

In the late biopsy group, kidneys that failed (n = 27) were compared to those that did not fail (n = 78) (Figure 2). We also studied a subgroup of grafts that did not fail and had reached >1000 days post biopsy (n = 30 patients, antibody status available in 29). Demographics for these patients are shown in Table S1.

imageimageimage

Figure 2. Antibody status, histologic lesions, histopathologic diagnosis in failed vs. nonfailed grafts. We compared the features at the time of biopsy between grafts that subsequently failed and grafts still functioning at the censoring date. Biopsies still functioning at 1000 days post biopsy are shown as an additional group. Since we observed only one graft failure in the grafts that were biopsied within 1-year posttransplant, we limited this analysis to biopsies taken >12 months posttransplant. We compared (A) the frequency of HLA antibodies (PRA and DSA), (B) the average histologic lesion scores, (C) the frequency of histopathologic diagnoses between grafts that failed and grafts still functioning at end of follow-up. Differences between the groups were assessed by chi-square or Fisher's exact test (frequency of antibodies or diagnoses) or by t-test (average lesion scores).

Surprisingly, given the high frequency of anti-HLA in late biopsies noted above, neither PRA nor DSA was strongly associated with failure, confirming a previous report (28), although DSA was more frequent in failed grafts (failed 67%, functioning 38%), particularly class II (failed 56%, functioning 26%) (Figure 2A).

The striking difference in biopsies of kidneys that failed was in microcirculation changes, as well as the expected increase in scarring (Figure 2B). C4d staining was present in biopsies of kidneys that failed (7/27 [25%]) but was present in some that did not (8/78 [10%]) (not significant, p = 0.16). Diagnoses of C4d+ABMR and C4d transplant glomerulopathy, as well as glomerulonephritis, were more common in kidneys that subsequently failed (not significant) (Figure 2C).

Banff diagnoses in 27 late biopsies from kidneys that failed (Figure 3A) consisted of 7 C4d+ABMR (all chronic-active), 6 glomerulonephritis, 2 TCMR, 4 borderline TCMR, 1 IFTA not otherwise specified and 7 ‘other’. Survival analysis of late biopsies by Banff diagnostic categories showed no significant differences, with many biopsies progressing to loss not explained (Figure 3B).

image

Figure 3. Phenotypes of failing grafts and death-censored graft survival in relationship to Banff diagnoses or antibody-associated microcirculation changes. Panel A shows the distribution of histopathologic diagnoses by Banff classification in grafts that subsequently failed. Panel B shows death-censored graft survival in patients biopsied late (>12 months posttransplantation) in relationship to the histopathologic diagnoses by Banff classification. Biopsies diagnosed as recurrent or de novo glomerulonephritis were analyzed as a separate category. Panels C and D show survival curves before and after reassigning the patients without C4d+ ABMR into diagnostic categories based on the presence or absence of panel-reactive antibodies (PRA) and the presence or absence of microcirculation changes (inflammation and or deterioration). Patients with glomerulonephritis were kept as a separate category. Microcirculation change was defined as glomerulitis score ≥1 and/or peritubular capillaritis score ≥1 and/or glomerulopathy score ≥1 and/or peritubular basement membrane multilayering score ≥2 (≥ 5 layers).

Download figure to PowerPoint

The association of microcirculation changes with graft loss, along with the marginal significance of C4d staining, suggested that the current definition of ABMR emphasizing C4d staining was not sufficiently sensitive. We reassigned cases based on a modified definition of ABMR, which included cases suspicious for ABMR based on the presence of HLA antibody (PRA) and microcirculation changes without C4d. PRA was used because the DSA test did not include anti DP specificities. The results for DSA were similar. Microcirculation change was defined as glomerulitis score ≥1 and/or peritubular capillaritis score ≥1 and/or glomerulopathy score ≥1 and/or peritubular basement membrane multilayering score ≥2 (≥5 layers). This classification created five groups: C4d+ABMR (acute and/or chronic active ABMR by current Banff criteria); C4dABMR (i.e. PRA+ with microcirculation changes); PRA+ cases without microcirculation changes; PRA cases; and glomerulonephritis.

Most graft losses (63%) occurred in kidneys with ABMR, either C4d+ or C4d (Figure 3C) (sensitivity: 0.65, specificity: 0.65, area under curve 0.68). This compares with C4d+ABMR alone (sensitivity: 0.27, specificity: 0.88, area under curve 0.58). Excluding cases with glomerulonephritis, graft survival was significantly impaired in late biopsies with either C4d+ or C4d ABMR, compared to those without ABMR (p = 0.01) (Figure 3D). Graft survival in PRA+ patients whose biopsies lacked microcirculation changes was similar to PRA negative patients.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of Interest Statement
  9. References
  10. Supporting Information

We analyzed all consenting patients presenting for kidney transplant biopsy (primarily for dysfunction or proteinuria) between 2004 and 2007, with the goal of understanding the phenotype of failing transplants. Progression to graft loss after biopsy was frequent in patients presenting for biopsy late, i.e. >1-year posttransplant, and occurred within 3 years of the biopsy. Compared to early biopsies, late biopsies were more often associated with anti-HLA and displayed more microcirculation changes (i.e. glomerulitis, capillaritis, peritubular capillary multilayering, glomerulopathy), scarring, arterial fibrous intimal thickening and hyalinosis; tubulointerstitial inflammation and arteritis were similar. In multivariable analysis, features robustly associated with graft loss were microcirculation changes and interstitial fibrosis. The risk for graft loss was not captured well by the current Banff diagnoses because many cases with ABMR features (anti-HLA and microcirculation changes) were C4d negative and thus given other diagnoses. However, when ABMR was redefined as the presence of HLA antibody and microcirculation changes, regardless of C4d status, it was the most frequent phenotype associated with subsequent graft loss. In contrast, nonspecific scarring (IFTA), calcineurin inhibitor toxicity and TCMR were rare diagnoses in grafts that subsequently failed. Thus antibody-mediated microcirculation injury accounts for the majority of kidneys presenting with indications for biopsy and subsequently failing, but many cases are C4d negative.

The strength of this study lies in its ability to represent the prevalent kidney transplants that deteriorate at various times posttransplant. This unselected study population represents the full spectrum of disease phenotypes and times posttransplant, with complete characterization including electron microscopy and HLA antibody testing at biopsy. The time of biopsy posttransplant predicts risk: kidney transplants beyond 1-year posttransplant presenting for renal biopsy are at high risk for progression to renal failure within 3 years, reflecting the disease entities that trigger such late biopsies. Many relatively benign or treatable conditions present in the first year, and more serious diseases (e.g. ABMR) present later provided the crossmatch is negative. This study could not have been performed with protocol biopsies, which exclude patients with clinical indications. Moreover, protocol biopsies are synchronized for time, which plays a dynamic role in the pathology findings, and have few subsequent failures. Furthermore, C4d positivity and microcirculation changes are rare in protocol biopsies (prevalence ∼2% of biopsies) (29).

The paucity of early ABMR and graft loss in our study reflects the effectiveness of current cross-matching, since all early biopsies were in transplants after 2004, when new anti-HLA screening methods were operating. We first described early ABMR due to HLA class I in an era of inadequate crossmatch technologies (10), but this is now rare. A different picture emerges in centers transplanting kidneys across positive crossmatches after desensitization (30–32): early ABMR and graft loss are more frequent in patients who had undergone desensitization for positive crossmatches. The higher prevalence of anti class II versus class I in late biopsies for cause has been described previously (33), yet remains puzzling, as dealt with in detail in a separate manuscript (Hidalgo et al., manuscript submitted).

Two findings of interest are the failure of C4d staining to define some kidneys undergoing ABMR, and the weak predictive value of PRA or DSA in late biopsies. C4d staining was a diagnostic advance because of its relative specificity (34), but it is neither completely sensitive nor specific. Kidneys transplanted across ABO blood group incompatibilities often have diffuse C4d staining with no features of renal dysfunction and microcirculation changes (35) and many cases with late ABMR lack C4d staining (17,19,36). C4d staining is useful but is the ‘tip of the iceberg’ for ABMR, as previously acknowledged when Banff designated C4d negative cases with anti-HLA and microcirculation changes as ‘suspicious for ABMR’. We suggest that they be designated ABMR in the Banff classification. However, anti-HLA at the time of biopsy for cause has little predictive value when there are no microcirculation changes.

The conventional belief that late failure of kidney transplants is progressive nonspecific scarring and/or calcineurin inhibitor toxicity must be reassessed, since undiagnosed antibody-mediated injury accounts for many of these cases. In the present study, only one patient who presented for biopsy and was diagnosed as calcineurin inhibitor toxicity progressed to renal failure, confirming other observations (7). Both scarring (formerly CAN, now IFTA not otherwise specified) and calcineurin inhibitor toxicity are diagnoses of exclusion, based on nonspecific lesions that are common in all late kidney transplants. Scarring develops over time with the cumulative burden of injury and disease, including calcineurin inhibitors (5), but when grafts are lost the fibrosis is almost always secondary to an identifiable disease. Similarly, hyalinosis is a common finding in chronic renal disease, and while characteristic of calcineurin inhibitor toxicity is actually nonspecific. The relationship between arteriolar hyalinosis and outcome in this data set is complex. The reverse association of ah with outcome was driven by poor survival in patients with ah0, suggesting that patients on calcineurin inhibitors with no hyalinosis despite years of exposure may be not adequately immunosuppressed. However, this observation must be confirmed and further investigated.

One implication of this study is that the well-known association between rejection and poor outcome (37) is primarily due to ABMR (either C4d+ and C4d). In our data, TCMR lacking anti-HLA and microcirculation lesions has little effect on outcome after a biopsy for clinical indications compared to other conditions found in such biopsies, probably reflecting effective treatment after the biopsy. Practice in kidney transplant follow-up clinics will probably change, with a more aggressive attempt to define ABMR, since this would be the best leverage for improving kidney survival, but the real impetus for change will be effective strategies for prevention and treatment of late ABMR.

While efficacy of various immunosuppressive strategies before and after biopsy on outcomes cannot be answered in this study, the new definition of ABMR sets the stage for such studies. The study population represents the prevalent renal transplant population transplanted over decades, and thus includes many immunosuppressive regimens. Like most centers, we do not switch stable patients to new immunosuppressive regimens without cause, resulting in persistence of a variety of regimens. Protective effects of maintenance immunosuppression on the incidence of late ABMR cannot be estimated in this study because we lack the comparator, i.e. the patients who do not develop biopsy indications. Furthermore, we cannot estimate the impact of treatment after the biopsy: center policy with ABMR is to optimize maintenance immunosuppression, switch to tacrolimus and mycophenolate if not in use or contraindicated, and to consider plasmapheresis, IVIG and rituximab, tailored to the individual patient circumstances. Moreover, in this study many ABMR diagnoses were made retrospectively. The emergence of a new definition of ABMR and realization of its importance set the stage for studies of the relative effect of various immunosuppressive options at preventing and treating late ABMR.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of Interest Statement
  9. References
  10. Supporting Information

The authors thank Dr. Zija Jacaj for help with collection of the clinical data; and Vido Ramassar and Anna Hutton for technical support.

Funding: This research has been supported by funding and/or resources from Genome Canada, Genome Alberta, the University of Alberta, the University of Alberta Hospital Foundation, Alberta Advanced Education and Technology, Roche Molecular Systems, Hoffmann-La Roche Canada Ltd., the Alberta Ministry of Advanced Education and Technology, the Roche Organ Transplant Research Foundation, the Kidney Foundation of Canada and Astellas Canada. Dr. Halloran also holds a Canada Research Chair in Transplant Immunology and the Muttart Chair in Clinical Immunology. Dr. Sis’ research has been supported by funding from Roche Organ Transplantation Research Foundation, University of Alberta Hospital Foundation and Kidney and Urology Foundation of America—Renal Pathology Society.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of Interest Statement
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of Interest Statement
  9. References
  10. Supporting Information

Table S1: Demographics in failed versus nonfailed grafts biopsied late.

Please note: Wiley-Blackwell are responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
AJT_2799_sm_TableS1.doc173KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.