World Journal of Surgery

, Volume 36, Issue 12, pp 2923–2930

Impact of Posttransplantation Glomerulonephritis on Long-term Outcome of Kidney Transplants: Single-Center 20-Year Experience

Authors

    • Division of Nephrology, Department of Internal MedicineTaichung Veterans General Hospital
    • Graduate Institute of Clinical Medical ScienceChina Medical University
  • Mei-Chin Wen
    • Department of PathologyTaichung Veteran General Hospital
  • Ming-Ju Wu
    • Division of Nephrology, Department of Internal MedicineTaichung Veterans General Hospital
  • Cheng-Hsu Chen
    • Division of Nephrology, Department of Internal MedicineTaichung Veterans General Hospital
  • Chi-Hung Cheng
    • Division of Nephrology, Department of Internal MedicineTaichung Veterans General Hospital
  • Chi-Yuan Li
    • Graduate Institute of Clinical Medical ScienceChina Medical University
  • Kuo-Hsiung Shu
    • Division of Nephrology, Department of Internal MedicineTaichung Veterans General Hospital
    • Chung-Shan Medical University
Article

DOI: 10.1007/s00268-012-1759-x

Cite this article as:
Yu, T., Wen, M., Wu, M. et al. World J Surg (2012) 36: 2923. doi:10.1007/s00268-012-1759-x

Abstract

Background

Successful renal transplantation has been performed in patients with end-stage renal disease and has been routine in patients with end-stage renal failure for more than two decades. Despite advances in the use of immunosuppressants, there has been only modest improvement in long-term allograft survival. Accumulating data have demonstrated that chronic rejection and recurrent glomerulonephritis are major causes of long-term allograft loss. However, data regarding the long-term impact of posttransplantation glomerulonephritis (PTGN) on ethnic Chinese populations are still unavailable.

Methods

From 1984 to 2010, a total of 268 patients who underwent renal allograft biopsies were reviewed retrospectively. Renal outcomes were compared by Kaplan–Meier analysis, and risk factors for renal survival and all-cause mortality were analyzed using the Cox proportional hazards model.

Results

In all, 85 patients (31.7 %) had PTGN, and the mean time of disease onset was 5.32 ± 5.18 years after transplantation. Among the 85 PTGN cases, 33 (39 %) were immunoglobulin A (IgA) nephropathy, 24 (28 %) were focal segmental glomerulosclerosis, and 8 (9.4 %) were membranous GN. Significant risk was associated with posttransplant IgA GN in hepatitis B virus carriers (odds ratio 5.371, 95 % confidence interval 1.68, 17.19; p = 0.0064). A total of 45 PTGN patients had allograft loss, of whom 49 % had IgA nephropathy. Patients with PTGN had inferior allograft survival rates compared to those with other pathologic findings (p < 0.0003).

Conclusions

Taken together, our results indicate that PTGN had a strong negative impact on long-term kidney graft survival. Posttransplant IgA nephropathy is a leading cause of allograft loss in Chinese kidney transplant patients with PTGN.

Introduction

Successful renal transplantation has been performed in patients with end-stage renal disease (ESRD) since 1955 [13] and has been routine in patients with end-stage renal failure for more than two decades. Following the development of potent immunosuppressive agents, such as calcineurin inhibitor (CNI) and mycophenolate mofetil (MMF) [4, 5], acute graft rejection is significantly less common, and 1-year patient and graft survival rates are now more than 90 % [6]. There has been only modest improvement in the actuarial 20-year graft survival rate, however [7, 8], and late graft loss due to chronic rejection and recurrent glomerulonephritis (GN) are major causes of ESRD [915].

Despite advances in the use of immunosuppressants, the occurrence of recurrent GN after transplantation has not declined and remains a problem for allograft survival [16]. Previous studies have estimated that recurrent GN develops in 10–20 % of allografts [11, 13, 17, 18]. Briganti et al. [11] reported that in addition to chronic rejection and death with a functional allograft, recurrent GN plays an important role in allograft loss after transplantation, with a 10-year incidence of 8.4 %.

Previous research documented that many immunologic and nonimmunologic factors increase the risk of poor outcome from renal transplantation, including recurrent GN after transplantation [19, 20]. However, the effect of posttransplantation GN (PTGN) on long-term outcomes of Chinese individuals with a transplant has not been determined. In the present retrospective cohort study, we investigated the effect of PTGN on the long-term outcomes of kidney transplants in Chinese patients.

Patients and methods

From 1984 to 2010, a total of 1,050 patients with ESRD underwent renal transplantation and regular follow-up at our hospital. All donors and recipients were ABO-compatible, and none of the patients had positive donor crossmatch of T and B cells. Computerized databases and hospital and clinical charts were reviewed retrospectively, and all renal biopsy reports and slides were reviewed. Demographic information was documented including age, sex, race, donor source (cadaveric or living), human leukocyte antigen (HLA) matching, histopathologic diagnosis of the native kidney, and type of renal replacement therapy (hemodialysis or peritoneal dialysis). Immunosuppressive regimens and laboratory data were recorded, including HLA matching, panel reactive antibody (PRA), virus serologic markers [including hepatitis virus B and C (HBV, HBC)], urinary protein, and presence of urinary sediment.

The following time intervals were recorded: time to ESRD (date of diagnosis of native kidney GN to onset of ESRD); dialysis duration (date of ESRD onset to date of transplantation); duration of follow-up (date of transplantation to date of death, of allograft loss, or of last follow-up visit); and time to event (date of transplantation to date of posttransplantation biopsy).

All renal transplant patients were monitored monthly at our outpatient department, where they underwent regular renal function examinations that included measurement of glomerular filtration rate (GFR) by Cr-EDTA, clearance of creatinine, 24-h urinary protein, presence of urinary sediment, and antinuclear antibody (ANA) assay. In total, 268 of the 1,050 recipients underwent 322 renal graft biopsies based on clinical indications. Although renal biopsies were not routinely performed in all transplant patients, a renal biopsy was performed if daily urinary protein was > 500 mg, serum creatinine was > 0.5 mg/dl above baseline, or if there was significant urinary sediment (> 5 red blood cells per high-power field).

Renal graft biopsy examinations were by light, immunofluorescence, and electron microscopy. Characteristic pathologic findings of renal allografts included acute T cell rejection, acute antibody-mediated rejection, CNI toxicity, tubular atrophy/interstitial fibrosis, and chronic transplant GN, based on the 1997 Banff classification [21, 22]. Immunohistochemical (IHC) staining for C4d, BK virus, and cytomegalovirus (CMV) was performed routinely.

Acute rejection was diagnosed by graft biopsy and treated with intravenous pulse methylprednisolone and by anti-thymocyte globulin for steroid-resistant episodes. A patient was enrolled in the PTGN group if there was a recurrence of glomerular disease as classified by Golgert et al. [23]: (1) true recurrence of primary GN—IgA nephropathy (IgAN), recurrent focal segmental glomerulosclerosis (FSGS), membranous GN (MGN), or membranoproliferative GN (MPGN)—or secondary GN—systemic lupus erythematosus (SLE), hemolytic-uremic syndrome/thrombotic thrombocytopenic purpura (HUS/TTP), crescentic GN, diabetic nephropathy, cryoglobulinemia, among others; (2) de novo disease; or (3) new onset of chronic GN after transplantation, indicating that the recipient was without native kidney pathologic reports. Patients with findings characteristic of acute cellular rejection, acute humoral rejection, tubular atrophy and interstitial fibrosis (TA/IF), chronic transplant GN, acute tubular necrosis (ATN), or CNI toxicity were included in the nonposttransplantation glomerulonephritis (NPTGN) group.

From 1984 to 1999, the standard immunosuppressive therapy after renal transplantation at our institution was prednisolone, starting at 30 mg/day and tapering to 5 mg/day by the third month. Cyclosporine was started at 10 mg/kg/day and was tapered to the lowest possible dose that maintained a stable serum trough level of 50–150 ng/ml. During this period, four patients also were given azathioprine (50–100 mg/day). From 2000 to 2010, tacrolimus (0.2–0.3 mg/kg/day) and mycophenolate mofetil (MMF; 500–750 mg twice per day) were substituted for cyclosporine and azathioprine, respectively. Tacrolimus was administered at the lowest possible dose that maintained a stable serum trough level of 6–8 ng/ml.

Statistical analysis

Descriptive statistics were expressed as the mean and standard deviation. Discrete variables were presented as frequency and percentage. The characteristics of both groups were compared with an independent two-sample t test or Wilcoxon rank sum test for continuous variables and with Fisher’s exact test for categorical variables. Categorical and nominal data were compared using the Pearson χ2 test. The analysis of patient survival came from the date of renal transplantation to censorship at death or the end of follow-up period by Kaplan–Meier analysis and using the log-rank test. Analysis of the renal outcome between both groups was derived from Kaplan–Meier analysis by the end-point of serum creatinine level up to 6 mg/dl during the follow-up period after kidney transplantation. The Cox proportional hazards model was applied to identify the presumed risk factors associated with the renal survival including seven recipient- and donor-related characteristics in a univariable–multivariable stepwise model: HBV virus, HCV markers, donor type (cadaveric or living), status of HLA mismatch, peak PRA level, use of immunosuppressants (tacrolimus or cyclosporine), and pretransplant dialysis duration. A 95 % confidence interval (CI) of a hazard ratio (HR) that did not include unity was considered to be statistically significant. In these analyses, all covariates were treated as time-dependent variables. A value of p ≤ 0.05 was considered significant. All analyses were performed using SPSS version 12 software (SPSS, Chicago, IL, USA).

Results

We enrolled 268 renal transplant patients in this retrospective cohort study, including 85 patients (32.7 %) who were diagnosed with PTGN. Overall, disease onset occurred at a mean duration of 5.3 ± 5.2 years after the kidney transplant, and the mean follow-up duration was 8.1 ± 5.9 years.

Among the 85 patients, 30 in the PTGN group (35 %) had true recurrence, 7 (8.2 %) had de novo GN, and 48 (56 %) had presumed new-onset GN (Table 1). The subtypes of PTGN were IgAN (33 cases, 39 %), FSGS (24 cases, 28 %), MGN (eight cases, 9.4 %), MPGN (six cases, 7.1 %), and other types (three cases of diabetic nephropathy, six cases of recurrent lupus nephritis, three cases of thrombotic microangiopathy, two cases of cryoglobulinemic GN).
Table 1

Classification of PTGN in 85 cases

Disorder

True recurrent GN

De novo GN

New-onset GN after transplantation

Total

30 (35.3 %)

7 (8.2 %)

48 (56.5 %)

IgA nephropathy (IgAN)

13

0

20

Focal segmental glomerulosclerosis (FSGS)

5

3

16

Membranous GN (MGN)

1

0

7

Membranoproliferative GN (MPGN)

3

3

0

Other systemic diseases

8

1

5

 Hemolytic uremic syndrome/TTP

0

0

3

 Lupus nephritis

6

0

0

 Cryoglobulinemic GN

0

0

2

 Diabetic nephropathy

2

1

0

PTGN posttransplantation glomerulonephritis; GN glomerulonephritis; TTP thrombotic thrombocytopenic purpura

Table 2 summarizes the demographic data of the PTGN and NPTGN groups. The only statistically significant differences were the status of HBV hepatitis (p = 0.001), duration of disease onset after transplantation (p = 0.003), and daily urine protein (p < 0.0001). In particular, 17 of the PTGN cases (19 %) had HBV virus, and 12 of these patients (71 %) were diagnosed with posttransplantation IgAN. Compared to the other subtypes of PTGN, a statistically significant risk was associated with posttransplantation IgA GN in HBV carriers [odds ratio (OR) 5.371, 95 % CI 1.68, 17.19; p = 0.0064]. Multivariate analysis using the Cox regression model indicated that none of these variables was independently associated with the incidence of PTGN, as shown in Table 3.
Table 2

Demographic data in both groups

Characteristic

PTGN (n = 85)

NPTGN (n = 183)

p

Sex

 

 Male

49.0 (57.6 %)

100.0 (54.6 %)

0.743

 Female

36.0 (42.4 %)

83.0 (45.4 %)

 

Blood type

0.456

 A

48.0 (56.5 %)

106.0 (58.6 %)

 

 B

15.0 (17.6 %)

21.0 (11.6 %)

 

 O

20.0 (23.5 %)

45.0 (24.9 %)

 

 AB

2.0 (2.4 %)

9.0 (5.0 %)

 

Donor type

0.053

 Cadaveric

75.0 (91.5 %)

176.0 (97.2 %)

 

 Living

7.0 (8.5 %)

5.0 (2.8 %)

 

Hepatitis B virus

0.001

 Negative

68.0 (81.0 %)

172.0 (95.0 %)

 

 Positive

17.0 (19.0 %)

9.0 (5.0 %)

 

Hepatitis C virus

0.580

 Negative

73.0 (86.9 %)

163.0 (90.1 %)

 

 Positive

11.0 (13.1 %)

18.0 (9.9 %)

 

Donor sex

0.555

 Male

68.0 (82.9 %)

155.0 (86.6 %)

 

 Female

14.0 (17.1 %)

24.0 (13.4 %)

 

Donor age

30.8 ± 9.0

33.6 ± 12.2

0.072

HLA mismatch

2.4 ± 1.3

3.0 ± 1.0

0.000

Peak level of PRA (%)

15.5 ± 16.7

18.1 ± 14.3

0.168

Recipient’s age at transplantation (years)

39.8 ± 11.2

40.7 ± 13.4

0.415

Immunosuppression

0.115

 Tacro-MMF-PD

24 (28.2 %)

69 (38.1 %)

 

 Cyclo-PD

61 (71.8 %)

112 (61.9 %)

 

Anti-CD 25 induction therapy

0.375

 Yes

6 (7.4 %)

8 (4.4 %)

 

 No

75 (92.6 %)

173 (95.6 %)

 

Pretransplantation dialysis time (years)

2.40 ± 2.64

2.41 ± 2.99

0.556

Time of graft abnormalities (years)a

5.32 ± 5.18

3.55 ± 4.61

0.003

Daily proteinurina (g) at the time of biopsy

0.30 ± 0.46

3.29 ± 3.51

0.000

Time of graft biopsy (months)

68.49 ± 64.28

49.50 ± 59.63

0.012

PRA panel reactive antibody; Tacro tacrolimus; MMF mycophenolate mofetil; PD prednisolone; cyclo cyclosporine

aPatients with clinical abnormalities including an elevation of serum creatinine, persistent microhematuria, and significant proteinuria

Table 3

Cox proportional hazards model for the effect of baseline variables on risk of posttransplantation GN

Variable

Univariate analyses

Multivariate analyses

Adjusted HR

95 % CI

p

Adjusted HR

95 % CI

p

Donor type (cadaveric)

2.38

0.228–24.87

0.469

1.16

0.443–3.027

0.766

HBV infection

2.98

0.728–12.238

0.129

0.89

0.494–1.598

0.692

HCV infection

2.11

0.456–9.781

0.339

0.94

0.461–1.907

0.860

Immunosuppressant (tacrolimus-based)

0.37

0.082–1.646

0.191

0.85

0.353–2.041

0.714

HLA typing mismatch (increment)

1.39

1.053–1.828

0.020

0.95

0.767–1.179

0.644

Panel-reactive antibody (per 10 % increment)

0.99

0.963–1.014

0.367

0.99

0.975–1.004

0.154

Duration of pretransplant dialysis

1.01

0.863–1.176

0.929

1.01

0.918–1.110

0.848

HR hazard ratio; CI confidence interval; HBV hepatitis B virus; HCV hepatitis C virus

We then estimated the cumulative incidence of PTGN and NPTGN (Fig. 1a). Of 268 patients, 85 were diagnosed with PTGN at a mean of 68.5 ± 64.3 months after transplantation. The cumulative incidences in the PTGN group were 36.90, 72.61, and 82.14 % at 5, 10, and 15 years, respectively. The cumulative incidences differed in the various PTGN subgroups (Fig. 1b), and most patients had IgA GN. We also compared the cumulative incidence rates of PTGN before and after the year 2000. In all, 61 cases were diagnosed with PTGN before 2000 and 24 were diagnosed after 2000; the difference between them was not significant.
https://static-content.springer.com/image/art%3A10.1007%2Fs00268-012-1759-x/MediaObjects/268_2012_1759_Fig1_HTML.gif
Fig. 1

a Cumulative incidence of posttransplantation glomerulonephritis (PTGN) and nonposttransplantation glomerulonephritis (NPTGN). The incidence was comparable for the two groups (p = NS). b Cumulative incidence of PTGN in subgroups. GN glomerulonephritis; IgA immunoglobulin A nephropathy; MGN membranous GN; MPGN membranoproliferative GN; FSGS focal segmental glomerulosclerosis. Comparison among all subgroups showed increasing numbers of posttransplantation IgA nephropathy and focal segmental glomerulosclerosis, which reached statistical significance (p < 0.001)

The overall allograft survival rates for patients with PTGN were 92, 76, 50, and 38 % at 1, 5, 10, and 15 years, respectively. Comparison of the PTGN and NPTGN groups indicated worse allograft survival in the PTGN group (p = 0.014) (Fig. 2a), but there was no significant difference in patient survival (p = 0.392) (Fig. 2b).
https://static-content.springer.com/image/art%3A10.1007%2Fs00268-012-1759-x/MediaObjects/268_2012_1759_Fig2_HTML.gif
Fig. 2

a Kaplan–Meier analysis of graft survival in the PTGN and NPTGN groups. Graft survival in the PTGN group was inferior to that in the NPTGN group, reaching statistical significance (p = 0.014). b Kaplan–Meier analysis of patient survival in the PTGN and NPTGN groups. The difference between the groups was insignificant (p = 0.392). c Kaplan–Meier analysis of various causes of renal graft loss including calcineurin inhibitor (CNI) nephrotoxicity, acute rejection, chronic allograft nephropathy, and PTGN. Graft survival was significantly inferior in the PTGN group, which was comparable to graft survival in patients with chronic allograft nephropathy (p = 0.567). d Kaplan–Meier analysis of patient survival in the same subgroups as in (c). e Graft survival of PTGN subgroups with IgA nephropathy, MGN, MPGN, and FSGS

Comparison of the allograft survival rates for patients with CNI nephrotoxicity (n = 23), acute rejection (n = 68), chronic allograft rejection (n = 92), and PTGN (n = 85) indicated significantly worse survival rate for the PTGN group (p < 0.0003) (Fig. 2c). Notably, comparison of patients with chronic allograft rejection and PTGN revealed worse allograft survival rates between two groups, but without a significant difference (p = 0.5683). Comparable patient survivals were noted in the subgroups (Fig. 2d) (p = 0.0686).

Further comparison of the four subgroups of PTGN indicated no significant difference in graft survival (Fig. 2e) (p = 0.458). Overall, 45 (53 %) patients with PTGN experienced allograft loss, 49 % of whom had IgA GN, 24 % had FSGS, 8.8 % had MGN, and 6.6 % had MPGN.

Discussion

Accumulating evidence has shown that long-term outcomes of patients undergoing kidney transplantation are affected by several co-morbidities, including cardiovascular diseases, malignancy, and infections [7, 8]. In addition, previous studies have identified PTGN as an important cause of end-stage renal failure and poor long-term outcome in patients undergoing renal transplantation [11]; however, the impact on Chinese renal transplant populations was unknown. The current study of Chinese transplant patients indicated that PTGN was associated with inferior long-term patient and graft survival and that posttransplantation IgA nephropathy was the leading cause of long-term allograft failure in PTGN patients.

Previous comparisons of graft complications including acute rejection, calcineurin nephrotoxicity, and chronic allograft nephropathy (CAN) indicate that PTGN significantly contributes to graft loss and poor long-term outcome [9, 11]. Following the development of new potent immunosuppressants (e.g., CNI, MMF, anti-CD25 agents), acute complications of kidney allografts were reduced and short-term survival rates have improved [6]. However, our results indicate that the long-term outcomes of kidney allografts were similar in patients with PTGN and NPTGN over a period of 20 years. Our findings highlight the role of PTGN and chronic allograft nephropathy in the long-term outcome of kidney transplantation. We examined the effects of modern immunosuppressants to see whether PTGN could be reduced in our cohorts. We found that the occurrence of PTGN steadily increased at the same rate before and after the year 2000. Thus, reducing PTGN by modifying the immunosuppressant regimen may be ineffective, perhaps because PTGN itself is a complex of diseases.

Comparable to other reports [9, 11], posttransplantation IgA was the most common cause of PTGN and the main cause of long-term graft loss in our cohort. Interestingly, we found a significantly greater percentage of HBV carriers in the PTGN group, and these patients had a significantly increased risk of posttransplantation IgAN. This could be due to the high prevalence of HBV infection in our area [2426]. A previous study by Lai et al. [27, 28] clearly demonstrated a relation between HBV and primary IgA GN in chronic hepatitis B surface antigen (HBsAg) carriers with native kidneys, presumably because HBsAg complex is localized in the mesangial region. The association of posttransplantation IgAN and HBV infection has not been reported previously [2931]. In a literature review [32], there was a large discrepancy in the occurrence of posttransplantation IgAN in transplant patients—ranging between 9 and 61 %. Also, the influence of long-term outcomes were inconsistent among the series. Different racial and geographic distribution of primary IgAN may contribute to these variations. Many risk factors have been reported to be associated with the posttransplanttation IgAN, such as the status of the HLA haplotype [30, 31], the donor type (cadaveric or living kidney) [33], whether using anti-thymocyte immunoglobulin, and the underlying crescentic GN with rapid progression to ESRD. The association of posttransplantation IgAN and HBV in the prevalent area has never to be reported.

Although current consensus suggests that the type and intensity of modern immunosuppression (CsA or tacrolimus) does not influence the risk of posttransplantation IgA recurrence, which is consistent with the finding in our cohort as well, the relation between immunosuppressant, HBV activation, and posttransplantation IgAN remains to be elucidated in the future.

In transplant recipients, prophylactic or preemptive treatment with anti-HBV agents has been reported to reduce significantly the complications of virus hepatitis after transplantation [34, 35]. Hence, it would be interesting to investigate the effect of anti-HBV agents on posttransplantation IgAN.

Our PTGN group had a higher proportion of patients using cyclosporine, and some previous studies have reported that cyclosporine causes diverse renal injuries associated with transplant nephropathy [14]. However, our Cox multivariate analysis indicated that immunosuppressants had no significant or independent effect. This could be due to the relatively small size of our study groups or because a longer period of observation is necessary to detect differences in patients treated with tacrolimus and MMF agents. In addition, data of primary native kidney diseases, which could affect outcome, is usually not obtained. The other limitation of our study is that about half of our PTGN patients did not have primary kidney biopsy reports, and the primary GN was diagnosed clinically by the presence of proteinuria and hematuria.

Conclusions

Our results indicate that PTGN plays an important role in the long-term outcome of Chinese patients undergoing renal transplantation. Also, posttransplantation IgAN is the leading cause of long-term graft loss. Chronic HBV presence is a significant risk factor associated with the development of posttransplantation IgA in our cohort. It is unknown whether modern immunosuppressant regimens can help prevent PTGN.

Copyright information

© Société Internationale de Chirurgie 2012