Recurrent Idiopathic Membranous Nephropathy After Kidney Transplantation: A Surveillance Biopsy Study


* Corresponding author: Fernando G. Cosio,


Membranous nephropathy (MN) is a common cause of nephrotic syndrome in adults. MN can recur after kidney transplantation causing proteinuria, allograft dysfunction and graft failure. In this study we assessed the incidence of MN recurrence utilizing surveillance graft biopsies. The study included 1310 renal allograft recipients from 2000 to 2006. Glomerular diseases were the cause of kidney failure in 28% of patients and 23 (2%) had idiopathic MN. Recurrent MN was diagnosed in eight of 19 patients included in this analysis (42%) 13 ± 20 months (median = 4; range 2–61 months) after transplant. The initial clinical manifestations of recurrent MN were mild or absent. Urine protein excretion was 825 ± 959 (64–2286) mg/day and three patients had no proteinuria. Five of seven patients who did not receive additional immunosuppression for MN had significant increases in proteinuria during follow up and three became nephrotic. At diagnosis, light microscopic changes were subtle or absent. All patients had granular glomerular basement membrane deposits of IgG but little or absent C3 by immunofluorescence. Subepithelial deposits were observed in all cases by electron microscopy. In conclusion, idiopathic MN recurred in 42% of patients after transplantation. The initial clinical and histologic manifestations are subtle but the disease is progressive.


Glomerulopathies are generally considered as the third most common cause of renal failure in kidney transplant recipients (1,2). However, the true number of kidney transplant recipients with glomerulopathies is not entirely clear because a large proportion of patients with chronic kidney disease never undergo a diagnostic kidney biopsy. Several glomerulopathies tend to recur in the kidney allograft and at least in some cases recurrence is associated with reduced allograft survival (3,4). Recurrent disease is an important cause of kidney graft failure (5–7). Nevertheless, it is likely that the impact of disease recurrence on allograft survival is currently underestimated because many allografts with progressive deterioration of function are never biopsied. Furthermore, previous estimates of disease recurrence rely on clinical evidence of glomerular pathology, such as proteinuria. However, it is likely that allograft glomerular diseases remain subclinical for variable periods of time (8).

In this study, we used surveillance biopsies to assess the incidence, time and clinical presentation of idiopathic membranous nephropathy (MN). Despite improvements in the treatment of native MN approximately 40% of patients eventually develop kidney failure (2). It is estimated that from 15% to 30% of cases of MN recur in the renal allograft resulting in proteinuria and reduced graft survival (9). We postulated here that the study of surveillance biopsies not only would provide information about the early clinical and histologic presentation of MN but also would allow early diagnosis when the disease may theoretically be more responsive to therapy (10).


Patient population

Between October 2000 and October 2006, 1310 adult patients received kidney transplant at our institution. Among this cohort, 367 patients (28%) had renal failure due to a primary or secondary glomerulopathy and 23 patients (2%) had the primary renal diagnosis of idiopathic MN. Among the patients with MN, 19 were followed in our institution and had surveillance biopsies after transplantation. These patients are the subject of this study. The pretransplant diagnosis of MN was confirmed by review of native kidney biopsies, when done in our institution, and/or by review of biopsy slides or pathology reports if the original kidney biopsy was not done in our institution. Additional clinical and laboratory information was obtained from electronic databases. The search for this information and the publication of this study was approved by the Mayo Clinic IRB committee.

In our transplant program, all patients have surveillance allograft biopsies done at the time of kidney implantation (time 0) and at 4, 12, 24 and 60 months posttransplant (11). These biopsies are not motivated by clinical events; rather, they are done at fixed time points after transplantation. The light microscopic sections were evaluated in a blinded fashion by two of the authors (JPG and TSD). The biopsies were carefully examined for the presence of early subepithelial deposits in trichrome stained sections and the presence of early basement membrane spikes in silver stained sections. In all cases, the diagnosis of recurrent MN was based on light, immunofluorescence (IF) and electron microscopy (EM). Similarly, the determination of no MN recurrence was made in all cases based on the most recent biopsy available for each patient and all of those biopsies had IF and EM. In our transplant program 24 h urine protein and iothalamate glomerular filtration rate (GFR) are routinely measured at 3 weeks, 1 year and yearly after transplantation. Kidney allograft function is frequently monitored by serum creatinine and estimated GFR (12).

The immunosuppression protocols used in these patients have been described in detail in previous publications (13). In brief, most patients in this cohort received induction immunosuppression with thymoglobulin (94% of patients) and others received anti-CD25 monoclonal antibodies (5%). Maintenance immunosuppression consisted of low dose corticosteroids (tapering to 5 mg daily after the third month), mycophenolate mofetil (750 mg twice daily) and either tacrolimus (79%), cyclosporine (6%) or sirolimus (15%). Among patients with MN 84% received induction with thymoglobulin, 90% maintenance with tacrolimus and the remaining received cyclosporine.

Data analysis

Data are expressed throughout the manuscript as mean and standard deviation. Proportions were compared by chi square; numerical data were compared by Student's t-test if normally distributed and by nonparametric tests if not normally distributed data. Recurrence over time posttransplant was assessed by Kaplan–Meier plots.


Clinical presentation and evolution of recurrent MN

Table 1 displays the characteristics of patients with MN in their native kidneys. The primary diagnosis of MN was made from 1 to 31 years prior to the transplant. One patient had received a kidney transplant previously and had developed recurrent MN in the original graft. Two additional patients in the nonrecurrent group had received previous allografts without recurrence. Table 1 also compares the pretransplant characteristics of patients with or without MN recurrence during the period of observation. As can be seen, there were no statistically significant differences in any of the parameters collected. It should be noted that among patients with recurrence there was a numerically higher number of deceased donor kidneys.

Table 1.  Patient characteristics
VariableAll patientsNonrecurrentRecurrent
  1. 1Years from MN diagnosis in the native kidney to the transplant.

  2. 2IS = immunosuppression.

Number of patients19118
Age54 ± 1154.2 ± 1253.7 ± 9.8
Sex (% males)68%63%75%
Race (% Caucasian)100%100%100%
Years of MN1 (range)11 ± 8 (1–31)11 ± 911 ± 7
Dialysis pretransplant63%64%63%
Donor type (% living donor)32%45%13%
Induction IS2 (% thymoglobulin)84%91%75%
Maintenance IS2 (% tacrolimus)90%91%88%
Follow-up months44.9 ± 1948 ± 1440 ± 24

Thirteen of the 19 MN patients had time 0 biopsies and all of these tissues were considered to be normal by light microscopy. The six patients who did not have time 0 biopsies had 4-month biopsies and none of those revealed MN. All but one patient had surveillance biopsies at 4 months and all patients had protocol biopsies at 1 year. Surveillance biopsies were done in all patients followed for at least 24 months and in all patients followed for at least 5 years. In patients without MN surveillance biopsies were done 69% of the time at time 0 and at 4 months, in 79% of patients followed for 1 year, 52% in those followed for 2 years and 45% in those followed for 5 years. Eight of the 19 patients (42%) had recurrent MN on surveillance biopsies a mean of 13 ± 20 months (median, 4; range 2–61 months) after the transplant. The incidence of recurrence was the same if patients with second transplants were excluded from the analysis. Figure 1 displays the cumulative incidence of recurrent MN in this group of patients. Six of the eight recurrences were diagnosed during the first year posttransplant. Two additional patients had late recurrence of MN, one 22 months and another 61 months after transplantation.

Figure 1.

Cumulative incidence of recurrent MN after kidney transplantation. Number of patients at each time point is indicated in the figure.

The patient's clinical features at the time of histologic diagnosis of recurrent MN are shown in Table 2. For comparison clinical information obtained at 1 year posttransplant is shown in patients without recurrence. There were no significant differences in blood pressure, allograft function or proteinuria between patients with recurrence and those without. Three of the eight patients with recurrent MN had urine protein excretion within the normal range (Figure 2). The other five patients had abnormal proteinuria (>150 mg/day) but none of them had urine protein concentrations >3000 mg/day. In one patient proteinuria was 1544 mg/day 3 months following the transplant and 1 month prior to diagnosis of recurrence. This particular patient had >9000 mg/day of proteinuria immediately before the transplant thus the proteinuria prior to the diagnosis of recurrence likely represented residual proteinuria from native kidneys (14). In fact, proteinuria at 1 year (8 months after MN recurrence) was 111 mg/day but the patient did not receive additional immunosuppression designed to treat the MN.

Table 2.  Clinical presentation of recurrent MN
  1. 1In this analysis values obtained at diagnosis of recurrence were compared to values obtained 1 year posttransplant in patients without recurrence.

  2. 2p-Value comparing systolic blood pressure between the groups is shown. Similar results were obtained comparing diastolic BP.

  3. 3Mann–Whitney.

Blood pressure125 ± 15/68 ± 11128 ± 9/75 ± 80.6222
Serum creatinine (mg/dL)1.5 ± 0.31.5 ± 0.40.989
Estimated GFR (mL/min/L.72 m2)48 ± 1353 ± 210.468
Proteinuria (mg/day) (range)169 ± 117 (33–378)825 ± 959 (64–2286)0.2063
Figure 2.

Proteinuria at the time of diagnosis of recurrent MGN or at 1 year after transplantation in patients without recurrent disease. Urine protein levels (Y axis) are displayed in a logarithmic scale.

Following the diagnosis of recurrent MN, seven patients were managed with tight blood pressure control (target systolic blood pressure <130 mmHg achieved in all patients) and angiotensin II inhibition (except in one patient whose systolic blood pressure was consistently 100–110 mmHg) in addition to posttransplant maintenance immunosuppression. Figure 3 displays the evolution of proteinuria in these seven patients. In one patient proteinuria did not change significantly over a 1-year period of time and remained within the normal range. In contrast, proteinuria increased significantly over time in six patients. In three patients proteinuria reached levels greater than 3000 mg/day after 6, 12 and 18 months, respectively, while in the remaining three patients proteinuria remained at subnephrotic levels. Kidney function did not change significantly during the period of observation (data not shown).

Figure 3.

Changes in proteinuria following the diagnosis of recurrent MN. This group of seven patients continued to receive standard transplant immunosuppression and therapy with angiotensin II inhibitors and tight blood pressure control.

Histology of recurrent MN

The biopsy findings in the eight patients with recurrent MN are summarized in Table 3. Sub-epithelial deposits were identified by light microscopy in six of the eight cases, but well-established basement membrane spikes were only identified in four cases. Two of the eight biopsies had no subepithelial deposits or spikes by light microscopy. In these two cases the diagnosis of recurrence was made by IF and EM. By IF, granular deposition of IgG along the glomerular basement membrane was present in all cases. In contrast, complement C3 deposits were either weak or absent. EM was available in all but one case and demonstrated podocyte foot process fusion and sub-epithelial electron dense deposits. In all cases the EM changes were qualified as stage I MN. Follow-up biopsies were available in seven of these eight cases. However, in only three of these cases was the repeat biopsy done before initiation of additional immunosuppression designed to treat the MN. The results of these three rebiopsy cases are also shown in Table 3. Two of the three cases showed lesions of focal segmental glomerulosclerosis (FSGS) superimposed upon MN. These biopsies also showed increasing interstitial fibrosis. Furthermore, histologic changes of MN were more prominent in these repeat biopsies particularly in those patients with increasing levels of proteinuria. Review of biopsies from patients without recurrence did not show any of the light, IF or EM findings specific for MN.

Table 3.  Histologic features in biopsies from eight patients with recurrent MN
PatientMonth post-TxProteinuria (mg/day)Light microscopy3IFElectron microscopy
GlomeruliSpikesDepositsFibrosisIGGC3FPF1Subepi DD2
  1. The diagnostic biopsy is shown in all cases. In addition, for three patients follow-up biopsy prior to any specific intervention is also shown.

  2. 1FPF = glomerular epithelial cell (podocyte) foot process fusion; 2DD = electron dense deposits; 3‘spikes’ were evaluated in silver stained slides, deposits and interstitial fibrosis after trichrome stain.

JH2.5 1479NoNoNo3+1+3+1+
RK3.9 20525NoYesNo3+NegNo1+
CB251 10814 (FSGS)1+1+Mild3+NegNo glomeruli
PR4.1 21771+YesNo3+TrNot done
PR2910 114 6 (FSGS)2+2+Mild2+2+3+2 ± spikes
RA22.4   6413NoYesMild3+1+1+1+
JL61 10261+YesNoNo glomeruli2+1+


This study showed that in 42% of kidney transplant patients with idiopathic MN the disease recurs in the allograft and that recurrence most often occurs during the first year following transplantation. As in previous studies we noted that some patients with recurrent MN have overt proteinuria at presentation (3,4). However, most patients had minor or absent clinical manifestations even while the histologic picture of MN was unequivocal. The discrepancy between clinical and the histologic manifestations may be explained by the biopsy findings of early recurrent MN that includes strong granular basement membrane staining for IgG but little or no C3 deposition. Previous studies in experimental models of MN related proteinuria to the deposition of terminal pathway complement components in glomeruli (15). Thus, perhaps the low levels of proteinuria observed in cases of early recurrent MN relates to the relatively low level of complement activation and deposition in the glomerular basement membrane. The evidence presented here confirms that these cases indeed had recurrent MN. Thus, all of these patients had surveillance biopsies at time 0 and/or at 4 months and none of those tissues showed MN in the donor kidney. Furthermore, de novo MN (8) cannot be diagnosed in these patients since all of them had MN in their native kidneys.

As in previous studies (3,4), these results do not disclose any pretransplant variables that differentiate patients who will develop recurrent MN from those who will not. In this study recurrence was diagnosed histologically rather than clinically. Thus, it was important to establish whether or not subclinical histologic recurrence had clinical consequences. We showed that during the period of follow up six of seven patients (86%) with recurrent MN had increasing levels of proteinuria during follow up. Furthermore, histologic progression was noted in two of the three patients who underwent repeat biopsies. In native kidneys the clinical presentation of MN, particularly the amount of proteinuria relates to prognosis (16) and spontaneous remission of the proteinuria may occur in a substantial number of patients (17). These results suggest that these observations may not apply to recurrent MN, which appears to have a much higher chance of progressing even when only minimal proteinuria was present at diagnosis. This may not be surprising because MN in a transplant candidate is indicative of a MN sub-type capable of causing kidney failure. It is interesting to note that in our patients disease progression, judged by proteinuria, was quite slow over a periods of years. However, this is not reassuring because it indicates that MN progression may occur in the absence of a progressive increase in proteinuria. Whether these observations are indicative of a modulating effect of antirejection medications on proteinuria or representative of the natural evolution of idiopathic MN cannot be determined by these studies.

We consider these observations as a first step toward developing better management strategies for kidney transplant recipients with MN. These data and previous studies (3,4) argue that MN recurrence in the allograft is unpredictable by pre-transplant variables. Furthermore, because the initial clinical manifestations of the disease are subtle early diagnosis requires surveillance biopsies. We postulate that early diagnosis of recurrent MN may lead to a more successful therapy of the disease. This hypothesis awaits confirmation. Conversely, these data provide evidence that when left untreated patients with recurrent MN may develop focal glomerulosclerosis that in native kidneys (18) and in allografts (19) is associated with poor prognosis. Ultimately, in order to optimize the care of kidney transplant patients with recurrent MN, both prompt diagnosis and effective therapy will need to be accomplished. Protocol biopsies that include IF and EM appear to be safe and reliable methods to diagnose early recurrent MN histologically, prior to clinical presentation, at a time point where specific intervention may be most effective.


This work was supported by grants from the Mayo Clinic Transplant Center and Division of Nephrology and Hypertension. We thank the kidney-pancreas transplant coordinators for their dedication to the care of transplant recipients and their help in the collection of data from these patients. We also thank Ms. Cynthia Handberg for her excellent secretarial assistance.