• Azathioprine;
  • chronic allograft nephropathy;
  • cyclosporine nephrotoxicity;
  • kidney transplantation;
  • mycophenolate mofetil


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Mycophenolate mofetil (MMF) reduces acute rejection in controlled trials of kidney transplantation and is associated with better registry graft survival. Recent experimental studies have demonstrated additional antifibrotic properties of MMF, however, human histological data are lacking. We evaluated sequential prospective protocol kidney biopsies from two historical cohorts treated with cyclosporine (CSA)-based triple therapy including prednisolone and either MMF (n = 25) or azathioprine (AZA, n = 25). Biopsies (n = 360) were taken from euglycemic kidney-pancreas transplant recipients. Histology was independently assessed by the Banff schema and electron microscopic morphometry. MMF reduced acute rejection and OKT3 use (p < 0.05) compared with AZA. MMF therapy was associated with limited chronic interstitial fibrosis, striped fibrosis and periglomerular fibrosis (p < 0.05–0.001), mesangial matrix accumulation (p < 0.01), chronic glomerulopathy scores (p < 0.05) and glomerulosclerosis (p < 0.05). MMF was associated with delayed expression of CSA nephrotoxicity, reduced arteriolar hyalinosis, striped fibrosis and tubular microcalcification (p < 0.05–0.001). The beneficial effects of MMF remained in recipients without acute rejection. Retrospective analysis shows that MMF therapy was associated with substantially reduced fibrosis in the glomerular, microvascular and interstitial compartments, and a delayed expression of CSA nephrotoxicity. These outcomes may be due to a limitation of immune-mediated injury and suggest a direct effect of reduced fibrogenesis.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Mycophenolate mofetil (MMF) is the prodrug of mycophenolic acid (MPA), a selective, noncompetitive inhibitor of inosine 5-monophosphate dehydrogenase (IMPDH), the rate-limiting enzyme for guanosine triphosphate synthesis. Inhibition of de novo purine synthesis and guanosine nucleotide depletion reduces T- and B-lymphocytes proliferation, primary (but not secondary) humoral and cellular immune responses (1). MPA induces apoptosis of activated T-lymphocytes with clonal elimination, suppresses glycosylation of adhesion molecules limiting recruitment of lymphocytes and monocytes, depletes tetrahydrobiopterin reducing inducible nitric oxide synthase activity, nitric oxide production and peroxynitrite-induced tissue damage (1). In prospective trials, induction immunosuppression with MMF reduced acute rejection rates in kidney transplantation (2), and was associated with improved graft survival in retrospective analyses (3).

Transplant renal function, blood pressure and lipid levels are predictably improved when MMF is used in established chronic allograft nephropathy (CAN) or cyclosporine (CSA) nephrotoxicity, in conjunction with dose reduction or elimination of calcineurin inhibitor therapy (4–6). Intriguingly, some data suggest that the addition of MMF alone, without alteration of the calcineurin inhibitor dose, also improves or stabilizes renal function in both human and experimental studies independent of CSA levels (7,8). Furthermore, the improved graft survival with MMF use from retrospective analysis could not be accounted for simply by immunological factors, and the presence of additional mechanisms have been postulated (3).

Mycophenolic acid exerts potent anti-proliferative effects on smooth muscle, mesangial, vascular endothelial and renal tubular cell cultures in vitro (9–15). MMF also reduces mesangial matrix deposition, glomerulosclerosis, regenerative proliferation, myofibroblast infiltration and interstitial fibrosis and collagen formation in models of glomerulonephritis, chronic rejection, CSA nephrotoxicity and several models of nonimmune renal disease (9,16–25). These data imply that MPA exerts an antiproliferative effect capable of actively modifying the response to injury, which are additional to its immunosuppressive properties. Histological data in human renal transplantation are, however, lacking.

We undertook this further evaluation of allograft pathology after consistently finding a protective effect of MMF against disparate histological parameters in long-term protocol kidney transplant biopsies by post hoc multivariate analysis (26). The aim of this study was formally to evaluate the histological presentation of CAN in CSA-treated recipients, comparing MMF with a historical control group using azathioprine (AZA) therapy.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Study population and design

The study patients were a consecutive series of type I diabetic recipients (spanning the era from 1993 to 1999) of a successful combined kidney-pancreas transplant as previously described (26). For inclusion, patients were to have three or more protocol kidney biopsies available for analysis, and under treatment using CSA and prednisolone combined with either MMF (MMF group n = 25, 1996–1999) or azathioprine (AZA group n = 25, 1993–1996). Induction antibody was not routinely used. Consecutive AZA and MMF patients were included in equal number, although excluded when baseline therapy was substantially altered through conversion to tacrolimus from CSA because of early acute rejection (n = 2 patients) or MMF pulmonary toxicity (n = 1), or because of pancreas thrombosis and recurrence of diabetes (n = 9) or calcineurin-mediated hemolytic uremic syndrome (n = 1).

Protocol biopsy study and histological definitions

Kidney biopsies were undertaken at implantation, 1 and 2 weeks, 1, 3, 6 and 12 months and then annually until 10 years. However only biopsies taken until 5 years after transplantation were analyzed to equalize the follow-up times between groups. Beginning in 1997, the protocol biopsy schedule was simplified to omit 2 and 4 years, resulting in slightly fewer total biopsy samples in the later MMF cohort.

Histological evaluation included assessment of global glomerulosclerosis and periglomerular fibrosis, and quantification of Banff qualifiers (27) including mesangial matrix (mm), chronic glomerulopathy (cg), arteriolar hyalinosis (ah), chronic interstitial fibrosis (ci) and tubular atrophy (ct) scores as described (26). CSA nephrotoxicity was defined by the presence of striped cortical fibrosis or de novo or increasing arteriolar hyalinosis (not attributable to renal ischemia or donor hyalinosis from implantation biopsies), which was often supported by the presence of tubular microcalcification (not related to preceding biopsy evidence of acute tubular necrosis). Acute rejection was defined clinically by a 25% increase in serum creatinine supported by biopsy, and subclinical rejection was defined as the finding on protocol biopsy of acute rejection or borderline changes without immediate functional deterioration. Needle-core protocol biopsies were obtained and assessed by two blinded observers using the Banff schema as previously described (26). Light microscopic evaluation included the use of periodic acid–Schiff stains.

In addition in selected samples (n = 10) with Banff cg score of 1 or greater, C4d was assessed as follows: dewaxing with histolene, rehydration in graded alcohols, antigen retrieval (Universal Decloaker, Biocare Medical, Concord, CA), application of primary polyclonal antihuman C4d antibody (1/200 dilution, Biomedica, Vienna), primary antibody enhancment (Labvision, Fremont, CA), then HRP, followed by DAB/H2O2 chromagen and counterstain by haematoxylin. A C4d positive control of severe antibody-mediated rejection and negative control without the primary antibody were also employed.

Electron microscopic glomerular evaluation

Electron microscopy was used to quantify the extent of mesangial matrix and glomerular ultrastructural abnormalities from 21 patients randomly selected from archived EM specimens of 3-month protocol biopsies from both groups. Renal cortical biopsy tissue was transported in modified Karnovsky fixative (2% formalin, 2.5% glutaraldehyde in 0.1M Mops buffer, pH 7.4), postfixed using osmium tetroxide, dehydrated using ascending ethanol series, embedded in Epon resin and polymerized for 20 h. Glomeruli for analysis were selected from sections taken from Epon blocks at 500 nm thickness and stained with toluidine blue in semi-thin sections, with sclerosed and incomplete glomeruli being discarded. Ultrathin sections (86 nm thickness) were placed on standard 400 mesh copper standard grids and slot grids, stained by lead citrate and uranyl acetate and examined by transmission electron microscopy (Philips CM120 Biotwin TEM, Philips, Eindhoven, Netherlands) for glomeruli (minimum of 2 per sample).

Photomicrographs of whole glomeruli were taken at 4200× to identify mesangial areas, and mesangial images at 9700× magnification were used to estimate mesangial matrix and mesangial cell volume fractions. Mesangium was defined by the following criteria: loss of parallelism between the endothelial and epithelial surface; presence of basement membrane like material; identification of a mesangial cell or mesangial cytoplasm; and presence of mesangial GBM. Mesangial cell bodies were identified by nuclear notching and dense cytoplasmic contents. Endothelial cells were differentiated by the presence of a capillary lumen and fenestrated endothelial membrane.

Images were acquired using a digital camera on the iTEM platform (Morada TEM digital camera, Soft Imaging Systems, Munster, Germany), archived as tiff files and analyzed by point counting using a 126-point randomly overlaid grid and morphometric analysis software (Soft Imaging Systems, analySIS, GmBh, Germany). Volume fractions of mesangial matrix and mesangial cells were computed with points falling on the entire sampled glomerular profile as a reference. The fractional volume of mesangium Vv (Mes/glom); fractional volume of mesangial matrix, Vv (MM/glom); and fractional volume of mesangial cells Vv (MC/glom) were thus defined (28,29).

Statistical analysis

An unpaired Student's t-test or a Wilcoxon test was used for nominal data and a conditional binomial test examined categorical data. Cox regression was used for survival analysis and a generalized estimating equation was used for analysis of repeated measurements. Multivariate analyses corrected for repeated measures were used to adjust for differences in initial group demographics and increased initial use of oil-based CSA compared with the microemulsion formulation in the AZA group. In addition, a subset analysis of all biopsies in patients only taking microemulsion CSA was undertaken to provide two `pure' groups to further eliminate any confounding effect of CSA formulation. Differences in histology associated with MMF treatment were evaluated for their relationship with immunological factors by subset analyses of patients without acute rejection. Data are expressed as mean ± SD unless otherwise stated. A probability of less than 0.05 was considered significant.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Clinical results and protocol histology

Overall, study recipients were 37.6 ± 7.3 years old and 56% were male. Kidney–pancreas recipients experienced sustained euglycemia with a mean HBA1C of 5.4 ± 0.9%. The numbers of acute cellular and vascular rejection episodes were 1.0 ± 0.9 (median = 1) and 0.10 ± 0.3 (median = 0) per patient, respectively, and the prevalence rates of subclinical rejection (including Banff `borderline') at 1, 3 and 12 months after transplantation were 61.3%, 28.0% and 7.9%, respectively. The 1 and 5-year isotopic GFR measurements were 64.2 ± 17.7 and 50.6 ± 20.0 mL/min, and serum creatinine measurements were 129 ± 28 and 148 ± 69 umol/L, respectively. The mean patient follow-up was 7 years and biopsy follow-up was 3.8 ± 2.5 years. No grafts failed during the 5-year histological follow-up.

Initial demographic parameters were comparable between groups, with the exception that many AZA-treated patients (72%) were initiated on the oil-based CSA formulation before subsequent universal conversion to the microemulsion formulation (at 1.6 ± 0.9 years after transplantation), which was then continued in all patients. The number of biopsies taken on oil-based CSA formulation was 89 (43% of AZA biopsies). The MMF-treated group was treated entirely with microemulsion CSA. The CSA dose and trough levels were equivalent between groups, through the study (Table 1, p = NS). The AZA group experienced more acute cellular and vascular rejection episodes and OKT3 use, compared with MMF (p < 0.05, Table 2). One patient each in both groups required posttransplant dialysis. There were no differences in the serum creatinine levels and isotopic GFR estimates between groups (Table 2). Overall, the mean AZA dose was 101 ± 29 mg/day and MMF dose was 2.7 ± 0.50 g/day.

Table 1.  Demographic characteristics and clinical outcomes of study patients according to the use of MMF and AZA. Mean ± SD (%)
Group ParameterAZAMMFp<
Number (patients)2525 
Recipient age (years)37.9 ± 8.237.4 ± 6.3NS
Recipient sex (n, % male)12 (48%)16 (64%)NS
Recipient weight (kg)63.8 ± 8.971.6 ± 14.6NS
Total HLA mismatch score4.1 ± 1.44.3 ± 1.4NS
Total ischemia time (h)12.9 ± 2.811.2 ± 3.4NS
Donor age (years)22.5 ± 8.922.3 ± 7.6NS
Duration dialysis (years)1.3 ± 0.81.4 ± 0.7NS
Cyclosporine dose (mg/kg/day)
  3 months6.4 ± 1.86.3 ± 1.8NS
  1 year6.1 ± 1.94.8 ± 1.5NS
  2 years5.1 ± 1.86.0 ± 1.7NS
  3 years5.3 ± 1.75.4 ± 1.9NS
  4 years5.1 ± 2.13.5 ± 3.4NS
  5 years4.3 ± 1.53.9 ± 0.8NS
Cyclosporine levels (ng/mL)
  3 months307 ± 116260 ± 67NS
  1 year233 ± 96213 ± 59NS
  2 years206 ± 106245 ± 81NS
  3 years202 ± 102223 ± 71NS
  4 years194 ± 116180 ± 32NS
  5 years126 ± 33159 ± 56NS
Prednisolone dose (mg/kg/day)
  3 months23.7 ± 4.022.7 ± 5.3NS
  1 year10.6 ± 2.29.5 ± 2.5NS
  5 years9.3 ± 1.510.0 ± 0NS
Table 2.  A clinical outcome of groups according to the use of MMF and AZA. Mean ± SD or number (%)
Group ParameterAZAMMFp<
Number (patients/biopsies)25 (207)25 (136) 
Delayed graft function01 (4%)NS
Acute cellular rejection (number)1.3 ± 0.90.7 ± 0.90.05
Vascular rejection (number)0.2 ± 0.40.0 ± 00.05
Use of OKT317 (68%)8 (32%)0.05
Persistent subclinical rejection6 (24%)7 (28%)NS
Duration of SCR (months)0.54 ± 1.70.42 ± 0.79NS
True chronic rejection2 (8%)3 (12%)NS
Clinical CSA nephrotoxicity7 (28%)2 (8%)NS (0.07)
Posttransplant hypertension15 (60%)17 (68%)NS
ACE inhibitor treatment7 (28%)5 (20%)NS
1 year systolic blood pressure134 ± 14137 ± 14NS
1 year diastolic blood pressure79 ± 782 ± 9NS
Serum creatinine (umol/L)
  Discharge151 ± 43139 ± 38NS
  3 months124 ± 25117 ± 21NS
5-year isotopic GFR (mL/min)48.8 ± 19.056.2 ± 25.5NS

Of 360 study biopsies analyzed (296 were collected as per protocol and 64 were clinically indicated but performed at protocol times), 13.6 ± 7.9 glomeruli and 2.2 ± 1.0 arteries were present per biopsy. Inadequate samples, defined as less than seven glomeruli or no artery, occurred in 16.6% of biopsy specimens. Each biopsy was scored without knowledge of the clinical details by two independent observers. The interobserver kappa statistics for the presence of CAN; chronic interstitial fibrosis, tubular atrophy, arteriolar hyalinosis and chronic glomerulopathy were 0.61, 0.80, 0.49, 0.51 and 0.35, respectively.

Time-course of tubulointerstitial damage

Compared with AZA, tubulointerstitial damage was reduced in the MMF group for Banff chronic interstitial fibrosis (coefficient from 1 to 5 years was –0.42 ± 0.15 by generalized estimating equation, p < 0.01) and tubular atrophy scores (coefficient =–0.32 ± 0.14, p < 0.05), (Figure 1). Banff chronic glomerulopathy scores at 3 months were less with MMF treatment compared with AZA (0.48 ± 0.43 vs. 0.78 ± 0.38, respectively, p < 0.05). CAN grade II or more was greater in the AZA group compared with MMF, although this did not reach significance (p = 0.15, Figure 2).


Figure 1. Time-course of tubulointerstitial damage according to treatment with either AZA or mycophenolate assessed by Banff chronic interstitial fibrosis scores (ci, top panel) and tubular atrophy scores (ct, bottom panel). Key: *p < 0.05, **p < 0.01.

Download figure to PowerPoint


Figure 2. Kaplan–Meier curves for the onset of moderate (grade II) CAN defined by the Banff criteria and based on its occurrence on sequential biopsies (p = NS between curves, top panel). Kaplan–Meier incidence of CSA nephrotoxicity defined by occurrence on sequential biopsies of increasing or de novo arteriolar hyalinosis and/or striped fibrosis (p < 0.05 between curves, bottom panel).

Download figure to PowerPoint

Microvascular injury and the expression of cyclosporine nephrotoxicity

The histological expression of CSA nephrotoxicity was ameliorated by MMF treatment; with reduced Banff arteriolar hyalinosis scores (coefficient =–0.35 ± 0.16, p < 0.05, Figure 3), and lesser prevalence of striped fibrosis (coefficient =–0.15 ± 0.05, p < 0.01, Figure 4) and microcalcification (coefficient =–0.21 ± 0.05, p < 0.001, Figure 4) compared with AZA treatment. MMF reduced the prevalence of total microcalcification and CSA-related microcalcification (defined as microcalcification that could not be attributed to earlier acute tubular necrosis on sequential biopsies).


Figure 3. Top panel: arteriolar hyalinosis scored by the Banff schema is reduced by mycophenolate treatment. Data displayed by time after transplantation. Mean ± SEM. Bottom panel: difference in global glomerulosclerosis according to mycophenolate or AZA treatment, by time after transplantation. Hatched bar is AZA-treated group and solid bar represents mycophenolate-treated biopsies. Key: *p<0.05, ***p<0.001.

Download figure to PowerPoint


Figure 4. Markers of tubular damage associated with the histological definition of CSA nephrotoxicity including the number of biopsies with tubular microcalcification (not explained by prior acute tubular necrosis, top panel) and the occurrence of striped fibrosis (bottom panel) presented according to time after transplantation. Hatched bar is AZA-treated group and solid bar represents mycophenolate-treated biopsies. Key: *p<0.05, **p<0.01.

Download figure to PowerPoint

By Kaplan–Meier analysis, MMF treatment retarded the early appearance of CSA nephrotoxicity on sequential histology (p < 0.05), although by 5 years after transplantation the curves had almost rejoined (Figure 2). There were no differences between CSA doses and levels between groups, and no effect when adjusted for the use of the microemulsion formulation of CSA (Table 1). Between 1 and 3 years after transplantation, MMF reduced arteriolar hyalinosis by 38.2 ± 14.9%, p < 0.05, chronic interstitial fibrosis by 36.5 ± 16.2%, p < 0.05, striped fibrosis by 17.6 ± 6%, p < 0.01, and microcalcification by 26.5 ± 4.8%, p < 0.001, respectively. By Cox proportional hazard modeling, the predictors of microcalcification were complex, with adjusted hazard ratios of 37.5 (95%CI 1.5–952.0, p < 0.05) for early post-transplant renal dysfunction (defined as serum creatinine failing to fall below 150 umol/L by day 8), 43.7 for use of microemulsion CSA (95%CI 1.35–1414, p < 0.05), with a protective effect of 0.1 for MMF use (95%CI 0.007–1.27, p = 0.07). This suggests that microcalcification may have resulted from ischemic-reperfusion injury to the tubules, aggravated by the more rapid CSA absorption from the microemulsion formulation—and that MMF still substantially moderated these effects. There were no differences between groups for serum calcium and phosphate (data not shown), but an effect from residual hyperparathyroidism cannot be entirely excluded because of the lack of PTH data.

When the entire data set was analyzed by generalized estimating equation, statistically corrected for patients taking CSA-microemulsion vs. CSA-oil formulation, MMF therapy was independently associated with reduced CSA nephrotoxicity as a composite diagnosis (p < 0.001), as well as striped fibrosis (p < 0.001), microcalcification not attributable to acute tubular necrosis (p < 0.01) and arteriolar hyalinosis scores (p < 0.001). In the CSA-microemulsion subgroup (where biopsies on CSA-S were excluded), significant time-point differences between MMF and AZA were preserved for the diagnosis of CSA nephrotoxicity (1–3 years after transplantation, p < 0.05–0.001), arteriolar hyalinosis scores (from 1 to 3 years, p < 0.05–0.01), striped fibrosis (years 1 and 2, p < 0.05), as well as periglomerular fibrosis (year 3 only significant, p < 0.001), and percentage sclerosed glomeruli (years 3–5, p < 0.05–0.001). Hence, histological differences between study groups persisted despite statistical correction of CSA formation and by subset analysis.

Evolution of chronic glomerular changes

The MMF treatment group initially retarded the expression of chronic transplant glomerular pathology and glomerulosclerosis. Banff mesangial matrix scores increased rapidly from 1 month after transplantation until 1 year, and then rates stabilized in both groups. MMF treatment was associated with substantially reduced mesangial matrix scores and a delayed increase within the first year after transplantation (coefficient =–0.21 ± 0.08, p < 0.01 vs. AZA, Figure 5). Banff mesangial matrix scores correlated with chronic interstitial fibrosis (r = 0.26, p < 0.001), tubular atrophy (r = 0.31, p < 0.001), chronic glomerulopathy (r = 0.21, p < 0.001) and arteriolar hyalinosis scores (r = 0.27, p < 0.001). By generalized estimating equation, mesangial matrix accumulation adjusted for time after transplantation was associated with chronic interstitial score (p < 0.001) and use of mycophenolate appeared protective (p < 0.001). Interstitial inflammation on biopsy had no effect. Chronic glomerulopathy scores (Banff cg) remained at relatively low levels by 5 years after transplantation in both groups (mean combined cg scores were 0.08 ± 0.27 from 1 month until 5 years inclusive), and so a treatment effect was difficult to discern. Chronic glomerulopathy scores by light microscopy were reduced in the MMF group (coefficient =–0.04 ± 0.02, p = 0.09 vs. AZA, data not shown).


Figure 5. Top panel: differences in of Banff mesangial matrix scores according to time after transplantation and treatment group. Bottom panel: the extent of periglomerular fibrosis according to time after transplantation and treatment. Group means ± SEM. Hatched bar is AZA-treated group and solid bar represents mycophenolate-treated biopsies. Key: *p<0.05, ***p<0.001.

Download figure to PowerPoint

These effects on glomerular microstructure were distinct from those seen with chronic antibody-mediated rejection and transplant glomerulopathy. Retrospective review did not reveal any preceding glomerulitis in sequential biopsies, PTC basement membrane multilamination by EM, or by retrospective C4d staining on paraffin sections, fixed by 10% buffered formalin. In limited samples (n = 10) with Banff cg score of 1 or greater, C4d was assessed by immunhistochemistry and found to be all negative (as defined as cut-off of 50% peritubular staining. Two biopsies (of 5) in the AZA group showed minimal 20–25% staining of PTC but were classified as negative, compared to all MMF cohort negative (p = NS). Donor-specific antibody levels were not available. Hence, the alterations in glomerular pathology probably are likely to represent a nonspecific effect of tissue response, and do not necessarily imply alteration of expression of chronic alloimmune transplant glomerulopathy (although numbers are small).

By electron microscopy evaluation at 3-months after transplantation, the extent of mesangial matrix deposition was reduced in the MMF compared with AZA biopsies (Figure 6)—without any overall change in the mesangial fractional volume. The fractional volume of mesangial matrix (Vv of MM/glom) of MMF was 0.25 ± 0.41 vs. 0.32 ± 0.76 with AZA, p = 0.017, with a trend in reduced fractional volume of mesangial cells (Vv MC/glom) of 0.21 ± 0.04 vs. 0.18 ± 0.04, p = 0.07, respectively. The mesangial fractional volumes (Vv Mes/glom) of 0.46 ± 0.52 vs. 0.49 ± 0.71, p = NS were comparable between groups.


Figure 6. Electron microscopy at 4700× of glomerular capillaries and mesangium with increased matrix in the AZA group (top panel) compared with MMF group (bottom panel).

Download figure to PowerPoint

The extent of global glomerulosclerosis was minimal until 1 year after transplantation when progressive increases occurred. By 5 years after transplantation, glomerulosclerosis was less in the MMF-treated group compared with AZA (12 ± 6% vs. 32 ± 22% sclerosed glomeruli, respectively, p < 0.05, Figure 3). Segmental glomerulosclerosis was not statistically different between groups assessed by light microscopy (p = 0.16 in favour of MMF, data not shown). Interestingly, there was also a marked reduction in periglomerular fibrosis with MMF treatment compared with AZA (2.7 ± 5.3% of glomeruli affected vs. 9.1 ± 12.3%, respectively, p < 0.001) in biopsies taken from one to 5 years after transplantation, inclusive. The difference was greatest by 3 years after transplantation (Figure 5), but its presence as a distinct entity was lessened at 5 years by the onset of substantial global glomerulosclerosis.

Subset analyses without rejection

In the subset of patients without acute clinical rejection (n = 16 patients/84 biopsies), histological differences remained between MMF and AZA treatment groups. By generalized estimating equation of biopsies between 1 month and 5 years, inclusive, MMF therapy was associated with reduced Banff chronic interstitial (ci) scores (coefficient –0.24 ± 0.12, p < 0.01), CSA nephrotoxicity (by composite histological diagnosis; coefficient –0.29 ± 0.09, p < 0.01), arteriolar hyalinosis scores (coefficient –0.52 ± 0.20, p < 0.05), mesangial matrix scores (coefficient –0.19 ± 0.05, p < 0.001), periglomerular fibrosis (coefficient –4.5 ± 1.6, p < 0.01), percentage sclerosed glomeruli (coefficient –6.83 ± 2.5, p < 0.01) but not chronic tubular atrophy scores (coefficient –0.14 ± 0.11, p = NS). In biopsies between 1 month and 3 years inclusive, MMF was associated with reduced individual histological markers of CSA nephrotoxicity, including striped fibrosis (coefficient –0.13 ± 0.03, p < 0.001), microcalcification not attributable to acute tubular necrosis (coefficient –0.16 ± 0.08, p < 0.05) and arteriolar hyalinosis scores (coefficient –0.40 ± 0.13, p < 0.01). Hence, histological protection attributable to MMF was also seen to the same extent in the subgroup without acute rejection—suggesting that nonimmune factors may have a role. The sole exception of tubular atrophy scores would imply that immune-mediated damage was an important contributing factor to tubular injury in those patients with rejection.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

This case-controlled cohort study demonstrated MMF was associated with reduced acute rejection and histological damage of the kidney transplant compared with AZA treatment. MMF appeared to limit the development of chronic interstitial fibrosis, tubular atrophy and periglomerular fibrosis on protocol biopsies. Despite equivalent CSA exposure in the two groups, MMF was also associated with a delay in the histological presentation of CSA nephrotoxicity lesions of arteriolar hyalinosis, striped fibrosis and tubular microcalcification. MMF treatment was associated with reduced glomerular injury, mesangial matrix accumulation, transplant glomerulopathy scores and subsequent glomerulosclerosis. The histological alteration of kidney transplant damage observed in the MMF therapy group was independent of acute rejection (differences being sustained in the subsets without rejection) and CSA formulation suggesting nonimmune mechanisms. We hypothesize that some of the pathological differences may be due directly to its antiproliferative and anti-fibrotic properties of MMF, outlined below.

There is increasing experimental evidence that at therapeutic concentrations used in human transplantation, MPA suppresses not only proliferating lymphocytes but a wide variety of other cells (15,30). Most cell types constitutively express the type I isoform of IMPDH, inhibited by MPA (although fivefold less than the type II isoform expressed by activated lymphocytes), and use both IMPDH-dependent de novo purine synthesis and the MPA-independent salvage pathways (1). The major clinical limiting toxicities of MMF are predominantly related to rapid cell turnover, such as hemopoietic suppression and gastrointestinal intolerance (enterocytes are 50% IMPDH dependent), indicating that at conventional doses other tissues are affected. Our study consistently showed widespread effects on glomerular, microvascular and tubulointerstitial histological compartments with MMF therapy.

MMF was associated with reduced chronic interstitial fibrosis and tubular atrophy compared with AZA and limited early tubulointerstitial damage. While reduced tubular atrophy scores were dependent on the presence of acute rejection, consistent with direct immune-mediated tubular destruction, chronic interstitial fibrosis generation was reduced in both rejecting and nonrejecting recipients, providing further supportive evidence of an additional nonimmune mechanism. An anti-fibrotic effect of MMF has been observed in experimental glomerulonephritis (31), models of true chronic transplant rejection and CSA nephrotoxicity (16,32,33), and in vitro fibroblast and mesangial cell cultures (31). In a renal allograft model of chronic rejection in Fischer to Lewis rats, MMF substantially reduced chronic tubulointerstitial damage, fibrointimal hyperplasia and glomerulosclerosis (16), although another study failed to see a suppression of chronic histology sum scores unless sirolimus was added, when a synergistic effect was observed (33). Interestingly, FK778, the active leflunomide metabolite and de novo pyrimidine synthesis inhibitor, also reduced chronic histological changes including the development of tubular atrophy, fibrointimal hyperplasia, transplant glomerulopathy and glomerulosclerosis in the same experimental chronic rejection model (34). In addition, mononuclear cell infiltration, serum allo-specific antibody production and intragraft TGFβ mRNA expression was decreased (34). This suggests that inhibition of de novo nucleotide synthesis through either of the pathways is effective in ameliorating damage from chronic rejection.

Even in nonimmune models of renal failure, including adriamycin-uninephrectomy, unilateral ureteric obstruction and the 5/6-remnant kidney models in rats, MMF reduces tubulointerstitial damage and glomerulosclerosis (9,17,19,23,24,35). The proliferative cellular healing response to injury, including lymphocyte, macrophage and myofibroblast infiltration was reduced, limiting deposition of interstitial collagen and fibronectin. Improved renal function was accompanied by less early tubulointerstitial damage and glomerulosclerosis (9,19,23), reduced profibrotic factor expression including TGFβ-1, TIMP and PAI-1 in these experimental models (17,21,24,36). In our study, MMF was associated with reduced chronic interstitial fibrosis, and also periglomerular and striped fibrosis. Because these fibrotic responses are etiologically disparate and occur within differing structural compartments of the transplanted kidney, MMF may operate by modifying a broad pathological response to injury—irrespective of the originating cause.

MMF also reduced markers of glomerular injury including mesangial matrix deposition, transplant glomerulopathy scores, periglomerular fibrosis and glomerulosclerosis, compared with AZA treatment. Increased mesangial matrix is characteristic of transplant glomerulopathy was ameliorated by MMF, independently of acute rejection status. The human mesangial cell is exquisitely sensitive to MPA, and several in vitro studies show maximal inhibition of stimulated proliferation by fetal calf serum at concentrations of 0.6–1 μM. MPA directly inhibits human and rodent mesangial cell proliferation in a dose-dependent manner, unaccompanied by cytotoxicity and reversed by guanosine, indicating mediation by IMPDH (12). Mesangial cellular turnover, apoptosis and mitogen-induced proliferation were also rapidly and substantially inhibited by MPA at low and clinically relevant drug concentrations. The IC50 of MPA for cultured human mesangial cells and immortalized clones ranged from 0.06 to 0.19 μM (11,12). At therapeutic does of MMF of 2 to 3 gm/day, MPA trough concentrations range between 1 and 5 mg/L (37), well above the concentrations used in vitro (the conversion of 1 μM is 0.32 mg/L). While the pharmacological activity of MPA is a function of unbound drug concentration and dependent on protein concentration (38), the cell culture data confirm that MPA inhibits a wide range of cells and tissues at clinically relevant concentrations.

MMF treatment was associated with reduced transplant chronic glomerulopathy scores (Banff cg) and glomerulosclerosis in our study. In anti-Thy 1.1 glomerulonephritis (31), MMF limited glomerular hypercellularity and hypertrophy, expression of α-smooth muscle actin (a marker of myofibroblasts) and extracellular matrix deposition; and it decreased renal cortical i-NOS mRNA, glomerular volume and glomerular cells and glomerulosclerosis in murine experimental glomerulonephritis, independent of the NF-kappaB pathway (25). Post-transplant glomerulosclerosis is likely to be exacerbated by ischemia from arteriolar and vascular insufficiency, mesangial matrix synthesis and from tubular destruction leading to atubular glomeruli formation (39). MMF may reduce glomerulosclerosis by limiting arteriolar hyalinosis and glomerular ischemia, mesangial matrix deposition, direct immune-mediated tubular destruction and the chronic interstitial fibrogenic healing response to injury.

MMF was also associated with delayed the histological expression of CSA nephrotoxicity in our study, with reduced arteriolar hyalinosis, striped fibrosis and tubular microcalcification, consistent with human (5–8,40,41) and experimental studies (14,16,21,32,33,42). CSA elimination or dose reduction using adjunctive MMF therapy improved renal transplant dysfunction (4–6,41), especially when CSA nephrotoxicity was the primary cause of impairment (6), but also stabilized the decline and sometimes improved renal function, independent of the CSA dose or levels (7,8) with the greatest benefit in the least damaged grafts (8).

Retrospective studies using historical controls need careful interpretation because of potential differences between eras. Our patients were demographically comparable at entry except for the initiation with oil-based CSA in some AZA-treated patients, prior to subsequent universal conversion to the microemulsion formulation. However, these histological differences persisted after statistical adjustment of formulation; and remained in the subgroup only used CSA-microemulsion therapy. The oil-based CSA formulation exposure was limited in duration in this study, and CSA doses and trough levels were not different between groups. So we cannot account for our observed differences by the increased absorption of microemulsion CSA. Indeed, it could be argued that the increased absorption and bioavailability (AUC) of microemulsion CSA (with equivalent trough levels between formulations) would tend to increase nephrotoxicity—making the (opposite) observed protective effect of MMF even more impressive in this group. In a rat model of CSA nephrotoxicity, MMF treatment substantially inhibited afferent arteriolopathy, TGF-B1, PAI-1 expression and extracellular protein deposition of biglycan and collagen types I and IV compared with controls (21). Thus, MMF therapy may have a potential adjunctive role in mitigating the response to arteriolar injury from CSA, although this protective effect was not sustained indefinitely, being limited to the first few years after transplantation. Another potential limitation of the study is the retrospective analyses (even though biopsies were collected prospectively), which limits the ability to prove causality. The patient cohorts were relatively high immunological risk kidney–pancreas recipients using CSA, limiting the generalizability to lower risk kidney transplants using more powerful tacrolimus-based therapy currently.

In conclusion, MMF therapy was associated with widespread and profound effects on the initiation and progression of damage to the transplanted kidney in this case-controlled study. MMF appeared to reduce tubulointerstitial, striped and periglomerular fibrosis; the evolution of glomerular abnormalities including transplant glomerulopathy, mesangial matrix deposition and glomerulosclerosis, and was associated with a delay in the histological presentation of progressive arteriolar hyalinosis and the lesions of CSA nephrotoxicity. These and other experiment data suggest that MMF, while lacking tissue specificity, may have a role as an intermediate or long-term immunosuppressive agent capable of limiting the pathogenic fibrotic process characteristic of CAN and partially ameliorating CSA nephrotoxicity.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The National Pancreas Transplant Unit at Westmead receives funding from the Commonwealth Government of Australia. Salary support is provided by a grant from the NHRMRC to CCRE for MDW and MV.

Potential conflicts of interest. Dr O'Connell received consulting fees from Wyeth, lecture fees from Janssen-Cilag and a grant from Roche, Dr Chapman has received consulting and lecture fees from Novartis, Roche, Fujisawa and Wyeth, lecture fees from Janssen-Cilag and a grant from Novartis.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  • 1
    Allison AC, Eugui EM. Mycophenolate mofetil and its mechanisms of action. Immunopharmacology 2000; 47: 85118.
  • 2
    Mathew TH. A blinded, long-term, randomized multicenter study of mycophenolate mofetil in cadaveric renal transplantation: results at three years. Tricontinental Mycophenolate Mofetil Renal Transplantation Study Group. Transplantation 1998; 65: 14501454.
  • 3
    Ojo AO, Meier-Kriesche HU, Hanson JA et al. Mycophenolate mofetil reduces late renal allograft loss independent of acute rejection. Transplantation 2000; 69: 24052409.
  • 4
    Schnuelle P, Van Der Heide JH, Tegzess A et al. Open randomized trial comparing early withdrawal of either cyclosporine or mycophenolate mofetil in stable renal transplant recipients initially treated with a triple drug regimen. J Am Soc Nephrol 2002; 13: 536543.
  • 5
    Francois H, Durrbach A, Amor M et al. The long-term effect of switching from cyclosporin A to mycophenolate mofetil in chronic renal graft dysfunction compared with conventional management. Nephrol Dial Transplant 2003; 18: 19091916.
  • 6
    Ducloux D, Motte G, Billerey C et al. Cyclosporin withdrawal with concomitant conversion from azathioprine to mycophenolate mofetil in renal transplant recipients with chronic allograft nephropathy: A 2-year follow-up. Transpl Int 2002; 15: 387392.
  • 7
    Gonzalez Molina M, Seron D, Garcia del Moral R et al. Mycophenolate mofetil reduces deterioration of renal function in patients with chronic allograft nephropathy. A follow-up study by the Spanish Cooperative Study Group of Chronic Allograft Nephropathy. Transplantation 2004; 77: 215220.
  • 8
    Henne T, Latta K, Strehlau J, Pape L, Ehrich JH, Offner G. Mycophenolate mofetil-induced reversal of glomerular filtration loss in children with chronic allograft nephropathy. Transplantation 2003; 76: 13261330.
  • 9
    Badid C, Vincent M, McGregor B et al. Mycophenolate mofetil reduces myofibroblast infiltration and collagen III deposition in rat remnant kidney. Kidney Int 2000; 58: 5161.
  • 10
    Dubus I, L'Azou B, Gordien M et al. Cytoskeletal reorganization by mycophenolic acid alters mesangial cell migration and contractility. Hypertension 2003; 42: 956961.
  • 11
    Dubus I, Vendrely B, Christophe I et al. Mycophenolic acid antagonizes the activation of cultured human mesangial cells. Kidney Int 2002; 62: 857867.
  • 12
    Hauser IA, Renders L, Radeke HH, Sterzel RB, Goppelt-Struebe M. Mycophenolate mofetil inhibits rat and human mesangial cell proliferation by guanosine depletion. Nephrol Dial Transplant 1999; 14: 5863.
  • 13
    Heinz C, Hudde T, Heise K, Steuhl KP. Antiproliferative effect of mycophenolate mofetil on cultured human Tenon fibroblasts. Graefes Arch Clin Exp Ophthalmol 2002; 240: 408414.
  • 14
    Mohacsi PJ, Tuller D, Hulliger B, Wijngaard PL. Different inhibitory effects of immunosuppressive drugs on human and rat aortic smooth muscle and endothelial cell proliferation stimulated by platelet-derived growth factor or endothelial cell growth factor. J Heart Lung Transplant 1997; 16: 484492.
  • 15
    Morath C, Zeier M. Review of the antiproliferative properties of mycophenolate mofetil in non-immune cells. Int J Clin Pharmacol Ther 2003; 41: 465469.
  • 16
    Azuma H, Binder J, Heemann U, Schmid C, Tullius SG, Tilney NL. Effects of RS61443 on functional and morphological changes in chronically rejecting rat kidney allografts. Transplantation 1995; 59: 460466.
  • 17
    Bayazit AK, Bayazit Y, Noyan A, Gonlusen G, Anarat A. Comparison of mycophenolate mofetil and azathioprine in obstructive nephropathy. Pediatr Nephrol 2003; 18: 100104.
  • 18
    Raisanen-Sokolowski A, Vuoristo P, Myllarniemi M, Yilmaz S, Kallio E, Hayry P. Mycophenolate mofetil (MMF, RS-61443) inhibits inflammation and smooth muscle cell proliferation in rat aortic allografts. Transpl Immunol 1995; 3: 342351.
  • 19
    Romero F, Rodriguez-Iturbe B, Parra G, Gonzalez L, Herrera-Acosta J, Tapia E. Mycophenolate mofetil prevents the progressive renal failure induced by 5/6 renal ablation in rats. Kidney Int 1999; 55: 945955.
  • 20
    Ryuzo M, Soares V. Effect of mycophenolate mofetil on the progression of adriamycin nephropathy. Ren Fail 2001; 23: 611619.
  • 21
    Shihab FS, Bennett WM, Yi H, Choi SO, Andoh TF. Mycophenolate mofetil ameliorates arteriolopathy and decreases transforming growth factor-beta1 in chronic cyclosporine nephrotoxicity. Am J Transplant 2003; 3: 15501559.
  • 22
    Tapia E, Franco M, Sanchez-Lozada LG et al. Mycophenolate mofetil prevents arteriolopathy and renal injury in subtotal ablation despite persistent hypertension. Kidney Int 2003; 63: 9941002.
  • 23
    Van den Branden C, Ceyssens B, Pauwels M et al. Effect of mycophenolate mofetil on glomerulosclerosis and renal oxidative stress in rats. Nephron Exp Nephrol 2003; 95: e9399.
  • 24
    Zhang C, Zhu Z, Wang G, Deng A. Effects of mycophenolate mofetil on renal interstitial fibrosis after Unilateral ureteral obstruction in rats. J Huazhong Univ Sci Technolog Med Sci 2003; 23: 269270, 282.
  • 25
    Yu CC, Yang CW, Wu MS et al. Mycophenolate mofetil reduces renal cortical inducible nitric oxide synthase mRNA expression and diminishes glomerulosclerosis in MRL/lpr mice. J Lab Clin Med 2001; 138: 6977.
  • 26
    Nankivell BJ, Borrows RJ, Fung CL, O'Connell PJ, Allen RD, Chapman JR. The natural history of chronic allograft nephropathy. N Engl J Med 2003; 349: 23262333.
  • 27
    Racusen LC, Solez K, Colvin RB et al. The Banff 97 working classification of renal allograft pathology. Kidney Int 1999; 55: 713723.
  • 28
    Caramori ML, Basgen JM, Mauer M. Glomerular structure in the normal human kidney: Differences between living and cadaver donors. J Am Soc Nephrol 2003; 14: 19011903.
  • 29
    Fioretto P, Steffes MW, Mauer M. Glomerular structure in nonproteinuric IDDM patients with various levels of albuminuria. Diabetes 1994; 43: 13581364.
  • 30
    Deierhoi MH, Sollinger HW, Diethelm AG, Belzer FO, Kauffman RS. One-year follow-up results of a phase I trial of mycophenolate mofetil (RS61443) in cadaveric renal transplantation. Transplant Proc 1993; 25: 693694.
  • 31
    Ziswiler R, Steinmann-Niggli K, Kappeler A, Daniel C, Marti HP. Mycophenolic acid: a new approach to the therapy of experimental mesangial proliferative glomerulonephritis. J Am Soc Nephrol 1998; 9: 20552066.
  • 32
    Viklicky O, Zou H, Muller V, Lacha J, Szabo A, Heemann U. SDZ-RAD prevents manifestation of chronic rejection in rat renal allografts. Transplantation 2000; 69: 497502.
  • 33
    Jolicoeur EM, Qi S, Xu D, Dumont L, Daloze P, Chen H. Combination therapy of mycophenolate mofetil and rapamycin in prevention of chronic renal allograft rejection in the rat. Transplantation 2003; 75: 5459.
  • 34
    Pan F, Ebbs A, Wynn C et al. FK778, a powerful new immunosuppressant, effectively reduces functional and histologic changes of chronic rejection in rat renal allografts. Transplantation 2003; 75: 11101114.
  • 35
    Fujihara CK, Malheiros DM, Zatz R, Noronha ID. Mycophenolate mofetil attenuates renal injury in the rat remnant kidney. Kidney Int 1998; 54: 15101519.
  • 36
    Shihab FS, Bennett WM, Yi H, Choi SO, Andoh TF. Combination therapy with sirolimus and mycophenolate mofetil: effects on the kidney and on transforming growth factor-beta1. Transplantation 2004; 77: 683686.
  • 37
    Kuypers DR, Vanrenterghem Y, Squifflet JP et al. Twelve-month evaluation of the clinical pharmacokinetics of total and free mycophenolic acid and its glucuronide metabolites in renal allograft recipients on low dose tacrolimus in combination with mycophenolate mofetil. Ther Drug Monit 2003; 25: 609622.
  • 38
    Nowak I, Shaw LM. Mycophenolic acid binding to human serum albumin: characterization and relation to pharmacodynamics. Clin Chem 1995; 41: 10111017.
  • 39
    Nankivell BJ, Borrows RJ, Fung CL, O'Connell PJ, Allen RD, Chapman JR. Evolution and pathophysiology of renal-transplant glomerulosclerosis. Transplantation 2004; 78: 461468.
  • 40
    Hueso M, Bover J, Seron D et al. Low-dose cyclosporine and mycophenolate mofetil in renal allograft recipients with suboptimal renal function. Transplantation 1998; 66: 17271731.
  • 41
    Weir MR, Ward MT, Blahut SA, et al. Long-term impact of discontinued or reduced calcineurin inhibitor in patients with chronic allograft nephropathy. Kidney Int 2001; 59: 15671573.
  • 42
    Haug C, Schmid-Kotsas A, Linder T et al. The immunosuppressive drug mycophenolic acid reduces endothelin-1 synthesis in endothelial cells and renal epithelial cells. Clin Sci (Lond) 2002; 103: 76S80S.