Early Calcification of Renal Allografts Detected by Protocol Biopsies: Causes and Clinical Implications

Authors


*Corresponding author: Wilfried Gwinner, wgwinner@t-online.de

Abstract

Interstitial calcification has been described in renal allografts, however, the etiology and significance of this finding for the graft are unclear. The aim of this study was to examine calcification in serial protocol biopsies, to test the hypothesis that calcification is related to parameters of calcium homeostasis in these patients and to analyze a possible relation between calcification and graft function at 1 year. We studied 213 patients with 586 protocol biopsies obtained 6 weeks, 3 and 6 months after transplantation. Calcifications increased over time, from 6.1% at 6 weeks to 17.8% at 6 months. Out of the 213 patients, 56 had calcification in one or more biopsies. Patients age and gender, underlying renal disease, dialysis mode and duration, previous transplantations, donor type, age and gender, HLA matches and ischemia time had no influence on calcification. Calcification was not related to rejection episodes, acute tubular lesions, calcineurin inhibitor toxicity or tubulointerstitial fibrosis and tubular atrophy. Patients with calcification had significantly higher serum parathormone and calcium levels. In patients with calcification, high PTH levels correlated with an inferior outcome of graft function at 1 year after transplantation (p < 0.05). Therefore, treatment of hyperparathyroidism should be considered earlier and more often in these patients.

Introduction

Calcifications in the tubulointerstitium of renal allografts have been described in small series of patients with functioning grafts (1–5) and in cases with terminal graft failure (6). In a larger recent study, calcification of renal allografts within the first week after transplantation was observed in more than one-fourth of patients with delayed graft function (7). In patients with a combined pancreas–kidney transplantation, serial protocol biopsies over a period of 10 years after transplantation revealed an increasing rate of tubular microcalcifications which was related to calcineurin inhibitor therapy and toxicity (8). Collectively, these reports indicate that tubulointerstitial calcifications of renal allografts may be associated mainly with the dystrophic effects of tubular injury and toxicity (7–9).

In the present study, we examine calcification systematically in serial protocol biopsies of renal allografts, with the aim to test the hypothesis that calcification is related to parameters of calcium homeostasis and to examine a possible association between calcification and the graft function at 1 year after transplantation.

Materials and Methods

In the transplant center of the Medical School of Hannover, renal protocol biopsies are regularly performed 6 weeks, 3 and 6 months after kidney or combined kidney/pancreas transplantation since 2001. From patients participating in this program, demographic and clinical data were collected from the time before and at the time of transplantation. Clinical data and routine laboratory results after transplantation were collected corresponding to the time points of the three protocol biopsies and 1-year posttransplantation. All data were entered into a customized database (Oracle Enterprises, version 8.0.5). Data collection and usage were performed with informed consent of the patients and with approval of the ethics board of the Medical School of Hannover.

From the files of 366 patients that are currently in the database, 213 patients were selected for the analysis. Patients were selected, if a serum creatinine value at 1 year or information about graft failure was available. Extensive comparisons between the selected patients and the whole group of patients in the database with regard to demographic data and those variables that were used for the subsequent analyses confirmed that the selected patients were representative of the whole group.

Protocol biopsies were evaluated according to the updated BANFF classification (10). Analyses included 183 biopsies at 6 weeks, 206 and 197 biopsies at 3 and 6 months, respectively. Besides routine stainings, von Kossa stain (11) was performed on those cases which had tubular or interstitial crystalloid deposits. The deposits were classified according to their localization in the luminal or tubulointerstitial area. Serum values of creatinine, calcium, phosphate and alkaline phosphatase activity were measured with an autoanalyzer. Parathormone levels were analyzed as the 7–84 fragment.

For the statistical analyses, the SPSS statistical software package, version 12.0.1 (SPSS Inc., Chicago, Illinois, USA), was used. Comparisons of categorical data between the groups and different biopsy time points were performed with the chi-square test for two or more samples. Numerical data were compared with the t-test and one-way ANOVA or with the u-test and Kruskal–Wallis test, dependent on the results of the analysis for normal distribution. Correlation analyses were performed using the Spearman rank test. A stepwise negative logistic regression analysis was performed with the variables which were significantly different in the univariate analyses. Mean values are given with standard deviation, unless otherwise stated. Differences with p < 0.05 were considered as being statistical significant.

Results

The incidence of different pathological findings in protocol biopsies obtained 6 weeks, 3 and 6 months after transplantation is shown in Figure 1. Besides findings such as acute rejection and borderline rejection, chronic changes with tubular atrophy and tubulointerstitial fibrosis, acute tubular damage and signs of calcineurin inhibitor toxicity, an increasing rate of calcification was observed over time, reaching nearly 18% at 6 months after transplantation (Figure 1). Out of 213 patients, 56 patients had calcification in one or more biopsies. Calcifications were present in a focal pattern and were mainly located in the tubular lumina, besides from some deposits in the cortical and medullar tubulointerstitium. Vascular calcifications were not observed (Figure 2 and Table 1). Von Kossa stain confirmed the presence of calcium-containing deposits.

Figure 1.

Incidence of pathological findings according to the updated BANFF classification (10) in protocol biopsies at 6 weeks, 3 and 6 months after transplantation (n = 213 patients; **p < 0.01; ***p < 0.001). CNI toxicity: calcineurin inhibitor toxicity, as presented by isometric vacuolization of tubular cells.

Figure 2.

Presentation of renal allograft calcification. Deposits were mainly located in a focal pattern in the tubular lumina, and in the cortical and medullary tubulointerstitium.

Table 1.  Localization of allograft calcification in protocol biopsies obtained at 6 weeks, 3 and 6 months after transplantation
Localizationn at
6 weeks3 months6 months
Luminal101626
Tubulointerstitial123
Luminal and tubulointerstitial219

For the subsequent analyses, patients were divided into those who never had calcification and those with calcification in any of the biopsies. First, we examined whether calcifications were related to any differences in the pre-transplant variables of the patients, properties of the graft, immunosuppressive therapy and additional medication. Second, we analyzed whether other bioptic findings in the biopsies were linked to calcification and third, we addressed the role of hyperparathyroidism and hypercalcemia for calcifications. Finally, to assess the association between calcifications and graft outcome, changes in graft function at 1 year after transplantation were examined.

Pre-transplant demographic variables and clinical conditions as well as factors related to graft quality and concordance between donor and recipient were not different between both groups as shown in Tables 2 and 3. Considering the initial phase of transplantation (Table 3), induction therapy with ATG was little more prevalent in patients without calcification. Reasons for induction therapy with ATG were the use of calcineurin inhibitor-sparing or avoiding protocols in recipients with a graft from a donor older than 69 years in 16 cases, and an mycophenolate mofetil (MMF) and steroid-based immunosuppression because of hemolytic uremic syndrome as underlying renal disease in one case. Initial function of the graft was more often present in patients with calcification. Best renal function between transplantation and the first protocol biopsy, as defined by the lowest serum creatinine achieved within the first 6 weeks after transplantation, was similar in both groups indicating that patients with calcification were not those with an inferior graft from the beginning. Differences in ATG therapy and initial function of the graft were not significant in the multivariate analysis.

Table 2.  Demographic data and pre-transplant status of the patients without calcification and of the patients with calcification in any of the biopsies
 Without
calcification
(n = 157)
With
calcification
(n = 56)
Recipient age (years)49 ± 1446 ± 14
Recipient gender (%male/female)58/4262/38
Cause of ESRD 
 Glomerulonephritis (%)2638
 Tubulointerstitial disease (%)1216
 Other/unknown (%)6246
Renal replacement therapy 
 Hemodialysis (%)9089
 Peritoneal dialysis (%)811
 Pre-emptive transplantation (%)20
Time on dialysis (months)76 ± 5484 ± 48
Parathyroidectomy before transplantation (%)2218
Re-transplanted patients (%)1218
Combined kidney/pancreas transplantation (%)8.93.6
Table 3.  Transplant data and variables in the early post-transplant period of the patients without calcification and of the patients with calcification in any of the biopsies
 Without
calcification
(n = 157)
With
calcification
(n = 56)
  1. *p < 0.05.

Donor age (years)43 ± 1645 ± 18
Donor gender (% male/female)49/5141/59
Donor type 
 Cadaver (%)8588
 Living, blood-related (%)95
 Living, not blood-related (%)67
S-creatinine of the donor (μmol/L)94 ± 7679 ± 40
Number of HLA mismatches2.252.32
Cold ischemia time (h)15.3 ± 7.615.5 ± 8.1
ATG induction (%)112*
IL-2 antibody induction (%)8293
Initial graft function (%)6279*
Lowest S-creatinine within the first 6 weeks after tx (μmol/L)146 ± 73134 ± 78

The immunosuppression taken in the time interval before the respective biopsy is summarized for all the three protocol biopsies in Figure 3. Whereas use of cyclosporine, tacrolimus and steroids was equal in both groups, patients with calcification received more often sirolimus and less often MMF. In the multivariate analysis, this was significant for sirolimus in the 6-weeks biopsies and for MMF in the 3-months biopsies.

Figure 3.

Immunosuppressive drugs taken in the time interval before the respective biopsy in patients with and without calcification. For simplification, data are summarized from all three intervals (univariate analysis: *p = 0.03, **p = 0.017; multivariate analysis: $p = 0.003 for the 3-months biopsy, #p = 0.027 for the 6-weeks biopsy).

Treatment with vitamin D was similarly frequent in patients with and without calcifications at 6 weeks (31% vs. 26%) and at 3 months (37% vs. 42%) after transplantation. At 6 months, vitamin D therapy had to be stopped in many patients with calcification and in some cases, bisphosphonates were necessary to reduce calcium levels. Thus, vitamin D treatment at 6 months was present in only 32% of the patients with calcifications, compared with 52% of the patients without calcifications. On the other hand, oral supplementation with phosphate was more prevalent in patients with calcification, with 46% at 6 weeks, 60% at 3 months and 62% at 6 months after transplantation, compared with 6%, 13% and 17% at the three-time points in patients without calcifications. With regard to the different localization of calcifications, phosphate supplementation tended to be more frequent in patients with tubulointerstitial calcifications compared to the patients with luminal calcifications at 6 months after transplantation (92% vs. 48%; p < 0.06).

Associated findings simultaneously present in biopsies with and without calcification are summarized in Figure 4. The incidence of rejection, borderline rejection, chronic changes, acute tubular lesions or calcineurin inhibitor toxicity was not different between both groups. Also, episodes of acute rejection and borderline rejection additionally detected in biopsies that were performed because of impaired graft function were similarly frequent (acute rejection: 24% and 23%, borderline rejection: 10% and 13% of the patients with and without calcification, respectively). As such pathological alterations may precede calcification, the biopsies prior to the biopsy with calcification were analyzed for each biopsy time point and compared with the biopsies from the patients without calcification in any of the biopsies. No differences were present with respect to the incidence of rejection, chronic changes, acute tubular lesions or calcineurin inhibitor toxicity (not shown).

Figure 4.

Associated findings in the biopsies with and without calcification which were simultaneously present in the respective specimen. For simplification, biopsy results are summarized from all three time points. CNI toxicity: calcineurin inhibitor toxicity, as presented by isometric vacuolization of tubular cells.

Serum levels for intact parathormone at the time of the respective biopsy were higher in patients with calcification (Figure 5). Analysis of the mean calcium serum levels and alkaline phosphatase activities in the interval before the respective biopsy showed that these values were also higher in patients with calcification. Serum levels of phosphate were comparable in both patient groups. The product of serum calcium and phosphate was similar in patients with and without calcification at 6 weeks (1.6 ± 0.3 vs. 1.9 ± 0.7) and 6 months (2.5 ± 0.6 vs. 2.5 ± 0.6) after transplantation but higher in patients with calcification at 3 months (2.6 ± 0.6 vs. 2.2 ± 0.6; p < 0.01). In the multivariate analysis, differences in parathormone and calcium levels were significant at 3 and 6 months after transplantation. Further analysis showed that parathormone levels in patients with calcification were not correlated with the serum creatinine concentrations at the three biopsy time points (not shown). Pre-biopsy mean serum calcium levels correlated positively with parathormone levels at 3 months posttransplantation (r = 0.62; p < 0.02), indicating that hyperparathyroidism in these patients was not secondary in response to low serum calcium levels. Patients with tubulointerstitial calcifications tended to have higher serum levels for parathormone and calcium than patients with luminal calcifications, without reaching statistical significance (p = 0.1).

Figure 5.

Serum levels of intact parathormone hormone (A), calcium (B), phosphate (C) and activity of alkaline phosphatase (D) in patients with (open symbols) and without (closed symbols) calcification. Individual and median values are shown. Multivariate analysis: parathormone levels at 3 and 6 months posttransplantation: p < 0.001; calcium levels in the interval before 3- and 6-months biopsy: p = 0.001.

For the assessment of graft outcome, the difference between the lowest serum creatinine value within the first 6 weeks after transplantation and the 1-year value was calculated as percentage change (Figure 6). No significant difference was present between patients without calcification and the whole group of patients who had calcification in any of the biopsies. There was a trend to a greater increase in serum creatinine for patients with calcification in the 3-months biopsy suggesting that calcification could have a negative effect on outcome but only for some patients. Therefore, a correlation analysis between the creatinine change at 1-year posttransplantation and parathormone levels at 3 and 6 months was performed in the patients showing calcification at these biopsy time points. The results indicated that an inferior outcome can be expected for patients with higher parathormone levels (Figure 7). In patients without calcification such correlation could not be established (not shown).

Figure 6.

Functional graft outcome, as assessed by the difference between the lowest serum creatinine value within the first 6 weeks after transplantation and the 1-year serum creatinine value, expressed as percentage change. Values represent mean and SEM; *p < 0.05 vs. patients without calcification.

Figure 7.

Correlation of functional graft outcome (delta %-increase in serum creatinine at 1 year) and intact parathormone serum levels at 3 months (A) and 6 months (B) after transplantation in patients with calcification in the biopsies.

Discussion

The present study demonstrates that a significant proportion of renal allografts develop tubulointerstitial calcifications within the first 6 months after transplantation. Calcifications were mainly located in the tubular lumina and less often, in the cortical and medullar tubulointerstitium. By von Kossa stain, the presence of calcium phosphate containing material was confirmed. Unlike to recent observations with early calcification during the first week after transplantation in conjunction with delayed graft function (7) and chronic microcalcifications developing over years as a sequel of calcineurin inhibitor toxicity (8), hyperparathyroidism and hypercalcemia were identified as the key factors in our patients with allograft calcification. Among the patients with calcification, parathormone and calcium levels tended to be highest in cases with tubulointerstitial deposits suggesting that the severity of disturbed calcium metabolism could be related to the localization of allograft calcification. Our novel findings underscore the relevance of regular protocol biopsies to identify pathological entities previously not sufficiently appreciated.

Correlation analysis of serum calcium levels and parathormone levels showed that hyperparathyroidsm in patients with allograft calcification was clearly not secondary in response to low calcium levels. Non-resolving hyperparathyroidism as defined by a more than 2.5-fold persistent elevation above the upper normal limit for serum parathormone has been reported in 17% of patients after renal transplantation (12). In patients with persistent hyperparathyroidism, a higher set-point for calcium was noted suggesting a dysregulation of parathormone secretion (13). Hypercalcemia has been recognized as a frequent sequel after renal transplantation affecting approximately 30–64% of the patients (14–18), with a maximum of about 6 months after transplantation (14,16,18). Post-transplant hypercalcemia was shown to be related to abnormally high parathormone values (14) and diffuse or nodular hyperplasia of parathyroid glands (17,18). Patients with extraskeletal calcification or renal bone disease before transplantation may have more often post-transplant hypercalcemia (14,15), whereas patients with an inferior graft function (creatinine clearance <30 mL/min) may be less prone to hypercalcemia (15).

The highly significant differences in parathormone and calcium serum levels between patients with and without graft calcification point to an important role of these factors in the development of renal allograft calcification. However, it must be noted that there was a considerable overlap of these values in both patient groups, indicating that other precipitating factors may be important in the process of calcification.

Apart from hyperparathyroidism and hypercalcemia, several conditions may be related to renal calcification such as renal tubular acidosis (19) and tubular injury due to acute tubular necrosis (7), acute rejection (1,3) or toxicity by calcineurin inhibitors or other drugs (8,9).

Excluding hyperparathyroidism and hypercalcemia, patients with and without graft calcification showed very few differences in the variables that might be related to calcification. Unfortunately, renal tubular acidosis could not be examined because of lacking data. Factors that can be related to tubular injury, such as cold ischemia time, delayed graft function or acute tubular necrosis and acute rejection in the biopsies, were not more prevalent in patients with graft calcification. This is in contrast to recent results that allograft calcification within the first week after transplantation is present in 27% of patients with delayed graft function and in 11% of patients without delayed graft function (7). As we found no relation between delayed graft function and increased calcification and a much lower incidence of calcification at 6 weeks after transplantation (6.1%), such early calcifications within the first week may represent a transient phenomenon in response to graft injury which may resolve in the later course after transplantation.

Calcineurin inhibitor toxicity appeared to be not important for graft calcification as isometric vacuolization in protocol biopsies and use of calcineurin inhibitors were not different between both patient groups. Analysis of the additional immunosupressive medication revealed differences with regard to MMF and sirolimus in patients with calcification. Due to the small number of patients with sirolimus treatment, no further clear-cut conclusions could be made. Sirolimus is believed to have a low potential in causing nephrotoxicity; however, tubular toxicity of calcineurin inhibitors may be accentuated in combination therapy with sirolimus in man (20). Of note, in rats, sirolimus monotherapy caused tubular microcalcifications similar to those seen with calcineurin inhibitors (21).

Treatment with vitamin D might have contributed to allograft calcification, because patients with calcification received vitamin D therapy similarly frequent as patients without calcification at 6 weeks and 3 months, despite higher serum calcium levels. In pediatric patients on dialysis treatment and after renal transplantation, a higher prevalence of generalized soft tissue calcification among patients with vitamin D therapy has been reported (22). Thus, discontinuation of vitamin D and institution of bisphosphonate therapy as done in some patients because of hypercalcemia, might have been too late to obviate allograft calcification in these patients.

We are uncertain about the significance of oral phosphate supplementation for allograft calcification. Phosphate supplementation was more prevalent in patients with calcifications, particularly in those with tubulointerstitial deposits. On the other hand, serum phosphate levels were not different between patients with and without calcification. The product of serum calcium and phosphate was moderately higher in patients with calcification at 3 months after transplantation, but still in a range well below the upper limit which has been defined for patients on dialysis to avoid extraskeletal calcifications (23,24). Nevertheless, we cannot exclude that therapy with phosphate supplements facilitated allograft calcification. In view of the current guidelines which advise phosphate supplementation in patients with serum phosphate concentrations below 0.48 mmol/L with the goal to attain levels between 0.81 and 1.45 mmol/L (25), we are now very cautious with phosphate therapy in patients after renal transplantation, avoiding phosphate levels above 0.8 mmol/L.

Recent studies revealed that circulating serum factors and local accumulation of proteins involved in the regulation of bone mineralization might be important in the processes of extraskeletal calcification. In patients with end-stage renal failure, low serum levels of fetuin, an inhibitor of calcification, were correlated with an increased cardiovascular mortality (26). Arteries from dialysis patients displayed staining for osteopontin, Type I collagen and alkaline phosphatase in the vicinity of medial calcifications (27). Moreover, thrombospondin-1, matrix Gla protein and cartilage oligomeric matrix protein may be important in extraskeletal calcifications (28). It is tempting to speculate that such factors were involved in the graft calcification in our patients, especially in those cases with rather low parathormone and calcium levels.

In patients with calcification, impairment of graft function at 1 year after transplantation was correlated with high parathormone levels. Therefore, earlier therapeutical intervention in such patients might be beneficial. As discussed by Messa et al. (13), an important issue is to identify and treat patients who are prone to develop severe, non-resolving post-transplant hyperparathyroidism already before transplantation. In that study, patients with non-resolving severe hyperparathyroidism had parathormone levels at the time of transplantation that were considerably higher than in patients without persistent post-transplant hyperparathyroidism (mean: 538 pg/mL vs. 168 pg/mL). Pre-transplant parathormone levels were not available in our patients. However, plots of the individual parathormone levels at the three biopsy time points in the patients with calcification (Figure 8) indicate that cases with persistent and severe hyperparathyroidsm may be recognized quite early after transplantation. In most patients with parathormone levels greater as approximately 400 pg/mL at 6 weeks after transplantation, regression of hyperparathyroidism was not observed at 6 months. On these grounds, we suggest that transplant candidates should have parathormone levels below 400–500 pg/mL and parathyroidectomy should be considered before transplantation if these limits cannot be achieved with medical therapy within a reasonable period of time. This is in contrast to current guidelines which recommend parathyroidectomy primarily in cases with persistent levels of intact parathormone levels above 800 pg/mL in association with hypercalcemia and/or hyperphosphatemia which is refractory to medical therapy (29).

Figure 8.

Individual course of intact parathormone serum levels from 6 weeks to 6 months after transplantation in patients with calcification. An arbitrary line depicts a concentration of 400 pg/mL. Patients with an initial parathormone concentration above 400 pg/mL are indicated by open symbols.

For patients with severe hyperparathyroidism early after transplantation, current guidelines advocate to await spontaneous resolution of hyperparathyroidism for 1 year and to reserve early parathyroidectomy for cases with severe hypercalcemia (serum calcium >2.87 mmol/L) (25,30). However, waiting for spontaneous resolution of hyperparathyroidism in patients with graft calcification and high parathormone levels could facilitate further calcification and graft impairment as regression of parathormone levels will not occur in the short-term (Figure 8). Parathyroidectomy in the early post-transplant period might be indicated in these cases but yet, evidence is lacking that this could prevent or halt graft calcification and functional impairment. In our series, parathyroidectomy had to be performed in seven of the patients with calcification within the first year after transplantation and in 3 patients thereafter, which is 18% of all patients with calcification compared with 1% in patients without calcification. No clear-cut effects on subsequent graft function could be observed in these patients. Whether therapy with calcimimetics is an alternative approach to postpone or obviate parathyroidectomy in such patients with severe and persistent post-transplant hyperparathyroidism remains open to further studies.

Acknowledgments

The help of A. Wohlgemuth and H. Doehring in the collection of the data is greatly appreciated.

Ancillary