Lower Incidence of Chronic Allograft Nephropathy at 1 Year Post-Transplantation in Patients Treated with Mycophenolate Mofetil
Abstract
Chronic allograft nephropathy (CAN) is the main cause of graft failure after the first year of transplantation.
This prospective, centrally randomized, open-label study was conducted to examine the possibility that mycophenolate mofetil (MMF) can prevent the emergence of CAN. The incidence of biopsy-proven CAN at 1 year was compared between two cyclosporine-based regimens comprising either mycophenolate mofetil (MMF) or azathioprine (AZA). The AZA group (n = 34) and the MMF group (n = 37) were balanced for all baseline characteristics of donors and recipients, the pre-existence of renal lesions on donor biopsy, the incidence of delayed graft function and acute rejection. Based on an intent-to-treat analysis, the number of patients with CAN at 1 year post-transplantation was significantly reduced in the MMF group (17/37–46%) compared with the AZA group (24/34–71%) (p = 0.03). When observed data were considered, 56/71 (78.8%) patients had a 1-year biopsy, and the number of patients with CAN was significantly lowered in the MMF group (9/29–31%) compared with the AZA group (17/27–63%) (p = 0.01).
These results suggest a beneficial effect of MMF on the incidence of CAN at 1 year post-transplantation.
Introduction
The number of cadaveric kidneys available has not increased at the same rate as the increasing number of patients awaiting transplantation, underscoring the need to maximize graft survival (1). Chronic allograft nephropathy (CAN), formerly called chronic rejection, is the main cause of graft failure after the first year of transplantation (2,3). It corresponds to an incompletely understood clinico-pathological phenomenon, which has been defined as a renal allograft dysfunction at least 3 months post-transplant, independent of acute rejection, drug toxicity and specific disease entities, with typical features on biopsy (4). Protocol biopsies-based studies have demonstrated that approximately 40% of renal allograft will develop CAN within 2 years post-transplant (5) and even as soon as 3 months after grafting in 25% of patients (6). However, chronic allograft lesions are nonspecific and may be the end result of a large palette of events, as different as lesions linked to aging or hypertension in the donor, delayed graft function, drug toxicity, viral infections or acute rejection. This difficulty could be overcome if a comparison is possible with a ‘zero-hour’ biopsy, performed preoperatively in the donor.
Mycophenolate mofetil (MMF), the morpholinoethylester of mycophenolic acid, is a potent, uncompetitive, and reversible inhibitor of eukaryotic inosine monophosphate dehydrogenase, the rate-limiting enzyme for the de novo pathway of purine synthesis (7). Mycophenolate mofetil has a selective antiproliferative effect on T and B lymphocytes and is now part of standard immunosuppression in renal transplantation (8,9). In vitro, MMF also inhibits vascular smooth cell proliferation, at concentrations that may be reached in clinical settings (10). In vivo, the administration of MMF limited the intragraft infiltration of monocytes/macrophages and T cells (11). Such effects of MMF could be of relevance in the prevention of CAN, which is characterized morphologicallly by obliterative vasculopathy, interstitial fibrosis, and a variable degree of mononuclear cell infiltration. However direct evidence of an effect of this molecule on the onset of CAN has not yet been investigated.
This study was designed to examine the possibility that a MMF-based immunosuppressive regimen treatment could decrease the emergence of lesions of CAN in adult cadaveric kidney allograft recipients.
Methods
Study design
This prospective, open-label, centrally randomized study was designed to compare the incidence of biopsy-proven chronic allograft nephropathy (CAN) at 1 year post-transplantation between two cyclosporine A-based immunosuppressive regimens comprising either MMF or azathioprine (AZA). Patients were randomized in a 1 : 1 ratio before transplantation and were stratified by the study site. Approval was obtained from the Bordeaux University Hospital Ethics Committee.
Patient selection
Men and women, 18–70 years of age, receiving their first ABO-compatible cadaver kidney transplant were eligible for enrollment. Patients were not included if they were: (1) a recipient of a previous organ transplant, (2) currently receiving a multiorgan transplant, (3) had a high immunological risk defined as having a previously measured panel-reactive antibody grade of ≥80%; (4) transplanted with a kidney from a donor of ≥65 years; or (5) were currently receiving immunosuppressants.
Patients were excluded after randomization if the randomized drug therapy was discontinued for more than 14 days.
Treatment plan
Patients were randomized before transplant to receive either MMF (CellCept, Roche, Neuilly sur-Seine, France) or AZA (Imurel, Glaxo-Wellcome, Paris, France). Both medications were started immediately before surgery. Oral MMF was administered at 1000 mg twice daily. Azathioprine was administered at a starting IV dose of 3 mg/kg for day 0 only, then at an oral dose of 1.5–2 mg/kg/day.
All patients also received additional immunosuppressive therapy with methylprednisolone (Solu-médrol, Médrol, Pharmacia & Upjohn, Saint-Quentin-en-Yvelines, France), Cyclosporine A (Néoral, Novartis, Rueil Malmaison, France) and an induction course of a polyclonal antithymocyte globulin (ATG, Fresenius, Sèvres, France). Methylprednisolone was administered at 500 mg on day 0, 2 mg/kg on day 1, 0.5 mg/kg/day by day 14, 0.25 mg/kg/day by day 28 and a maintenance dose of 0.1–0.15 mg/kg/day. Cyclosporine A was started at day 10 at a starting dose of 4 mg/kg/day in two divided doses. Trough cyclosporine whole-blood monitoring was performed to achieve target concentrations of 150–200 ng/mL for the first 3 months, and 100–150 thereafter based on a radioimmunoassay specific to the parent molecule [CYCLO-Trac(r) SP-Whole Blood, Incstar, Stillwater, MN]. Antithymocyte globulins were administered from day 0 to day 10, at an initial dose of 1.25 mg/kg/day, with monitoring based on a daily CD2+ T lymphocyte determination.
The diagnosis of acute rejection was based on a biopsy and scored using the Banff schema. Treatment of rejection episodes used corticosteroids as a first-line therapy.
Cytomegalovirus (CMV) infection was diagnosed based on CMV pp65 antigen positivity in peripheral blood leukocytes (Argene Biosoft, Varilhes, France). CMV disease was defined as previously described (12). Preemptive rather than preventive treatment was performed based on an iterative monitoring of CMV antigenemia during the first few months, and patients were treated using IV ganciclovir in case of positive pp65 antigenemia.
Histological analysis
Chronic allograft nephropathy is characterized by varying degrees of interstitial fibrosis, tubular atrophy, fibrointimal hyperplasia of the renal vessels and glomerulosclerosis. The biopsies had been scored using Banff schema, in which the severity of changes of the glomeruli, vessels, tubuli, and interstium of a renal allograft is assigned a value of 0–3 (13) The primary efficacy variable was the comparison of the proportion of patients in each group with lesions of CAN on a 1-year protocol biopsy. All biopsies (from donors and recipients) were blindly and centrally examined by the same pathologist (CD). One specimen was used for a conventional histological examination and the second specimen was divided for immunofluorescence (in order to eliminate any possible recurrence of the primary glomerulopathy) and an electron microscopy examination, focused on peritubular capillary membrane splitting, which has been described as strongly associated with lesions of CAN (10,14). This latter examination was performed in order to help to distinguish CAN from cyclosporine-induced nephrotoxicity, which are two overlapping biopsy entities. In cases in which only one core could be obtained, this material was assigned for conventional histology as a matter of priority. The total mean (± SD) number of glomeruli present in the biopsies was 8.8 ± 5.3. All biopsies were scored by the revised Banff 97 classification by the central pathologist (13).
As far as possible, a preoperative biopsy of kidney donor was realized for the conventional histology examination and the results were compared with those of the 1-year protocol biopsy.
Statistical analysis
It was planned to enroll 34 patients in each group in order to provide a statistical power of 80% with an a-level of 5% for the analysis of the primary criterion. Chronic allograft nephropathy at 1 year post transplantation estimated at 40% in the MMF group and 75% in the AZA group: expected delta = 35%. A mean drop-out rate of 10% was assumed. To handle missing data, two different analyses were conducted using data from all patients who had at least one postbaseline efficacy observation:
- 1
One analysis by affecting ‘failure’ (i.e. chronic allograft nephropathy at 1 year post transplantation) to missing data (intent-to-treat). This analysis (primary) underestimates the true difference between groups.
- 2
One analysis with observed data using all the available data without replacing missing data (observed data). This analysis (secondary) only concerned patients who underwent a biopsy at 1 year and overestimates the true difference between groups.
Data analysis
Data are represented as the mean ± SD (unless otherwise stated). Univariate associations of categorical variables were examined with the Chi-square test. Univariate associations of continuous variables were examined with the nonparametric Mann–Whitney U-test for unpaired variables or Wilcoxon rank-test for paired variables. Correlation statistical significance was tested with the Fisher z-test. Tests were considered significant when the p-value was below 0.05.
Results
Baseline characteristics of donors and recipients before transplantation did not differ between the two treatment groups
Three patients, two in the MMF group and one in the AZA group, were excluded after randomization for discontinuation of the randomized drug and a total of 71 individuals was finally available for analysis, distributed in the AZA group (n = 34) and the MMF group (n = 37). Table 1 summarizes the donors' characteristics of the two treatment groups, which appeared to be well matched with respect to age, sex ratio, cause of death, cardiovascular history, hemodynamic events and renal function. Unless there was a logistic obstacle, a preoperative donor biopsy was performed in 80% of the patients in both groups. Mild interstitial fibrosis and segmental glomerulosclerosis were observed in donor-biopsy specimens of 20.6% of AZA patients and 16% of MMF patients, which was not significantly different. As depicted in Table 2, the two treatment groups of recipients were balanced for all baseline characteristics and immediate post-transplant status.
| AZA (n = 34) | MMF (n = 37) | p | |
|---|---|---|---|
| Age (yr) | 42 ± 14 | 41 ± 16 | NS |
| Sex : male/female (n) | 23/11 | 26/11 | NS |
| Cause of death (n) | NS | ||
| Head injury | 9 | 17 | |
| Stroke | 14 | 14 | |
| Others | 11 | 6 | |
| History of cardiovascular disease (n) | 1 | 2 | NS |
| Hypotensive period (n) | 5 | 5 | NS |
| Cardiac arrest (n) | 4 | 3 | NS |
| Serum creatinine (μM/L) | 87 ± 30 | 86 ± 27 | NS |
| Lesions on preoperative renal biopsy (n) | NS | ||
| Yes | 6 | 5 | |
| No | 22 | 25 | |
| Not performed | 6 | 7 |
- MMF = mycophenolate mofetil; AZA = azathioprine.
- Plus–minus values are means ± SD.
| AZA (n = 34) | MMF (n = 37) | p | |
|---|---|---|---|
| Age (yr) | 47 ± 13 | 44 ± 10 | NS |
| Sex : male/female (n) | 20/14 | 29/8 | NS |
| Primary renal disease (n) | NS | ||
| Glomerular | 14 | 17 | |
| Hereditary | 7 | 6 | |
| Tubulointerstitial | 3 | 4 | |
| Vascular | 1 | 3 | |
| Unknown | 9 | 7 | |
| History of cardiovascular disease (n) | 18 | 23 | NS |
| Time on dialysis (mo) | 42 ± 42 | 27 ± 24 | NS |
| Prior blood transfusion (n) | 13 | 14 | NS |
| Pretransplant | 4 | 2 | NS |
| PRA = 20% (n) | |||
| HLA A, B, DR mismatching | 2.8 ± 1.1 | 2.7 ± 0.9 | NS |
| Cold ischemia time (hr) | 21.8 ± 6 | 21.6 ± 8 | NS |
| Kidney anastomosis (min) | 48 ± 17 | 49 ± 19 | NS |
| Delayed graft function (n) | 13 | 17 | NS |
| Dialysis in first week (n) | 2 | 2 | NS |
- Plus−minus values are means ±SD.
- Panel-reactive antibodies (PRAs) are anti-HLA antibodies that have a cytotoxic effect on lymphocytes obtained from a panel of donors among the general population.
- Delayed graft function was defined as a level of serum creatinine above 200 µM/L at day 3.
- MMF = mycophenolate mofetil; AZA = azathioprine.
Incidence of acute rejection and biological follow up were not different between the two treatment groups
The incidence of acute rejection was not significantly different between the two groups of patients (Table 3). In the AZA group, the incidence was 20.5% with a mean delay of 43.7 days post-transplantation. These seven acute rejections were scored as grade I (n = 2), grade II (n = 3), or grade III (n = 2). Graft was lost in two patients with grade III and one patient with grade II. In the MMF group, the incidence of acute rejection was not significantly different (13.5%), with a similar mean delay of 48.5 days post-transplantation. These five acute rejections were scored as grade I (n = 4) and grade II (n = 1). All the acute rejections in the MMF group were corticosteroid-sensitive, whereas only three out of seven from the AZA group responded favorably to corticosteroid (p = 0.01).
| AZA (n = 34) | MMF (n = 37) | p | |
|---|---|---|---|
| Acute rejection episodes | 7 (20.5) | 5 (13.5) | NS |
| Moderate to severe (Grade II–III) | 5 (15) | 1 (2.7) | NS |
| Corticosteroid-sensitivity | 3 (42) | 5 (100) | 0.01 |
- MMF = mycophenolate mofetil; AZA = azathioprine.
To properly evaluate biopsy data it is relevant to describe the parameters that may affect the outcome of the graft, such as serum creatinine, creatinine clearance (estimated by the Cockcroft-Gault formula), proteinuria, and cyclosporine A levels. These biological parameters and others (serum cholesterol and triglycerides, hemoglobin, platelet and white blood cell counts) were not found to be significantly different between the two groups at 3, 6, 9, and 12 months (Table 4 showing these values at month 6 and month 12).
| AZA (n = 34) | MMF (n = 37) | p | |
|---|---|---|---|
| Serum creatinine (μmol/L) | |||
| 6 months | 137 ± 87 | 132 ± 37 | NS |
| 12 months | 121 ± 36 | 138 ± 50 | NS |
| Creatinine clearance1 (mL/mn) | |||
| 6 months | 58.5 ± 20.1 | 60.4 ± 17.3 | NS |
| 12 months | 63.1 ± 16.8 | 61.3 ± 15.8 | NS |
| Proteinuria (g/L) | |||
| 6 months | 0.2 ± 0.4 | 0.1 ± 0.2 | NS |
| 12 months | 0.2 ± 0.3 | 0.2 ± 0.6 | NS |
| Trough concentration CsA (ng/mL) | |||
| 6 months | 150 ± 69 | 140 ± 131 | NS |
| 12 months | 136 ± 38 | 140 ± 42 | NS |
| Serum cholesterol (mmol/L) | |||
| 6 months | 5.8 ± 1.3 | 5.9 ± 1.2 | NS |
| 12 months | 6.0 ± 1.5 | 5.9 ± 1.2 | NS |
| Serum triglycerides (mmol/L) | |||
| 6 months | 1.9 ± 1.3 | 1.9 ± 1.0 | NS |
| 12 months | 2.0 ± 1.5 | 2.1 ± 1.3 | NS |
- Plus–minus values are means ± SD.
- 1Creatinine clearance was estimated by the Cockcroft-Gault formula.
- MMF = mycophenolate mofetil; AZA = azathioprine.
Higher incidence of CMV infections in the AZA group
As indicated in Table 5, the incidence of CMV infection was more frequent in the AZA group of patients (50%) compared with the MMF group (29.7%) (p = 0.02), whereas the donor/recipient serologic status was not significantly different between the two groups. As most of these CMV infections were symptomatic (approximately 82% of them in the two groups), a higher number of CMV diseases was observed in the AZA group (41.2%) than in the MMF group (24.3%) (p = 0.007).
| AZA (n = 34) | MMF (n = 37) | p | |
|---|---|---|---|
| Donor/Recipient CMV serologic status | NS | ||
| D−/R− | 4 (11.8) | 9 (24.3) | |
| D+/R+ | 9 (26) | 6 (16.2) | |
| D+/R− | 7 (20.7) | 7 (18.9) | |
| D−/R+ | 10 (29.4) | 15 (40.5) | |
| Unknown | 4 (12) | 0 (0) | |
| CMV infections | 17 (50) | 11 (29.7) | 0.02 |
| CMV diseases | 14 (41.2) | 9 (24.3) | 0.007 |
- MMF = mycophenolate mofetil; AZA = azathioprine; CMV = cytomegalovirus.
Lower incidence of chronic allograft nephropathy at 1 year post-transplantation in the MMF-treated group
Two different analyses of results were conducted and are depicted in Table 6. A primary analysis, based on the intent-to-treat principle, was first performed among the 71 patients with missing data, i.e. result of the protocol biopsy, considered as CAN. The number of patients with CAN at 1 year post-transplantation was significantly reduced in the MMF group (17/37, 46%) compared with the AZA group (24/34, 71%) (p = 0.03). As donor kidney quality is a known factor influencing graft survival, we also examined a possible link between the results of the donor biopsy and those of the 1-year protocol biopsy. Among the 71 patients, a donor biopsy was available for 58 (81.7%), showing kidney damage in 11/58 (19%) patients. These recipients with lesions on a ‘zero-hour’ biopsy were equally distributed between the AZA group (n = 6) and the MMF group (n = 5). Owing to the fact that a donor biopsy was not available in 13/71 (18.3.%) patients, mainly for logistic reasons, there was no significant correlation (p = 0.24) between the pre-existence of renal lesions on donor biopsy and the occurrence of CAN on 1-year protocol biopsy.
| Type of analysis | Intent-to-treat n (%) | p | Observed data n (%) | |||
|---|---|---|---|---|---|---|
| AZA (n = 34) | MMF (n = 37) | AZA (n = 27) | MMF (n = 29) | p | ||
| Chronic allograft nephropathy | 24 (71) | 17 (46) | 0.03 | 17 (63) | 9 (31) | 0.01 |
| Banff 97 chronic grade | ||||||
| No lesion | NA | NA | NA | 10 (37) | 20 (69) | 0.01 |
| Grade I | NA | NA | NA | 15 (55) | 7 (24) | 0.01 |
| Grade II | NA | NA | NA | 2 (7) | 2 (7) | NS |
- MMF = mycophenolate mofetil; AZA = azathioprine.
A secondary analysis based on observed data, which only concerned patients who underwent a biopsy at 1 year, was also performed. The average sample of this 1-year biopsy is marginal in the light of Banff 1997 criteria, as the total mean number of glomeruli was 8.8 ± 5.3. However, this parameter was not significantly different between the two groups of patients, AZA (8.8 ± 5.8) and MMF (8.8 ± 4.9), allowing a histological comparison. A histopathological result was obtained in 56 out of 71 patients (78.8%) and the reasons why a renal biopsy was not performed at month 12 post-transplantation are summarized in Table 7. Twenty-six out of 56 (46.4%) renal allograft recipients had histological signs of mild or moderate CAN after the first year. Again, the number of patients with CAN was significantly lowered in the MMF group (p = 0.01), 9/29 (31%), compared with the AZA group, 17/27 (63%) (Table 6). Accordingly, twofold more patients had no kidney damage in the MMF group, whereas the occurrence of grade I lesions of CAN was more than twofold higher in the AZA group (p = 0.01). More severe grade II lesions were only observed in four out of the 56 patients (7%), equally distributed between the two treatment groups, and no patients presented with grade III changes. Using the Banff 97 criteria for lesion scoring, we did not find any apparent histopathologic differences in the characteristics of CAN between the AZA and MMF biopsies (Table 8). Arteriolar hyaline thickening, which is considered as an indicator of anticalcineurin nephrotoxicity, was only found in three patients (two AZA, one MMF), who apparently had no history of cyclosporine A nephrotoxicity. Results of immunofluorescence studies were obtained in 35 out of 56 biopsied patients (62,5%) for logistic reasons. An equal number of patients in the AZA and MMF groups (approximately 32%) presented with some nonspecific patterns of immunoglobulin deposition, whereas only two patients, one in each group, had a recurrent nephropathy. Immunoglobulin deposits and the presence of CAN were found strongly associated (p = 0001). Only eight patients (three AZA, five MMF) showed peritubular capillary membrane splitting using electronic microscopy, obtained in 50 out of 56 biopsied patients, and the difference was not significant.
| AZA (n = 7) | MMF (n = 8) | p | |
|---|---|---|---|
| Death | 1 | 1 | NS |
| Graft loss | 5 | 0 | 0.01 |
| Refractory acute rejection | 3 | — | |
| Recurrent disease | 1 | — | |
| Vascular occlusion | 1 | — | |
| Technical reason | 0 | 3 | NS |
| Refusal of patient | 1 | 4 | NS |
- MMF = mycophenolate mofetil; AZA = azathioprine.
| Scoring of chronic allograft nephropathy, mean/3 (± SE) | AZA (n = 27) | MMF (n = 29) | p |
|---|---|---|---|
| Vascular changes | 0.48 ± 0.11 | 0.51 ± 0.11 | NS |
| Tubular atrophy | 0.59 ± 0.10 | 0.65 ± 0.11 | NS |
| Interstitial fibrosis | 0.44 ± 0.10 | 0.48 ± 0.11 | NS |
| Allograft glomerulopathy | 0.14 ± 0.06 | 0.03 ± 0.03 | NS |
| Mesangial matrix increase | 0.37 ± 0.10 | 0.48 ± 0.10 | NS |
- MMF = mycophenolate mofetil; AZA = azathioprine.
Discussion
This study was based on the result of a protocol biopsy showing that an already high prevalence (almost 50%) of mild or moderate lesions of CAN are observed 1 year following kidney transplantation. However, this result was not really surprising in the light of recent studies, showing an even more elevated prevalence of 94% at 1 year in kidney-pancreas recipients (16), and reinforcing our initial prestudy hypothesis of 75% of patients with CAN at 1 year in the control group.
We found that the patients receiving MMF in addition to maintenance therapy with cyclosporine A and corticosteroids had a lower frequency of biopsy-proven CAN 1 year after transplantation than the patients receiving AZA with the same maintenance therapy.
Despite a relatively limited number of patients, the two treatment groups were homogeneous. Donors' characteristics were similar for factors known to influence graft prognosis, such as age or donor illness (17). More importantly, a similar number of donors in the two groups had lesions on ‘zero-hour’ biopsy, and we did not find any significant association between these initial lesions and those observed 1 year later. However, this biopsy was usually performed at the time of organ removal and not implantation, and does not reflect acute tubular lesions linked to delayed graft function. This complex parameter is the result of different types of injuries, such as brain death, donor maintenance, organ preservation or rewarming, and not only ischemia-reperfusion injuries. Delayed graft function is generally considered, alone (18) or in the presence of acute rejection (19) as a critical risk factor affecting long-term outcome. In our study, delayed graft function as well as ischemia time or requirement for dialysis in the first week were comparable in both treatment groups and not significantly associated with rejection neither acute nor chronic (data not shown). A nonsignificant trend of a longer waiting time was observed in the AZA group, which has been previously described as a risk factor for CAN (15). Other nonimmunological variables, potential therapeutical targets of CAN, such as hypertension, hyperlipemia, or toxic effect of anticalcineurin agents (20) were also not significantly different in the two groups. The distinction between calcineurin inhibitor nephrotoxicity and CAN may be in some cases extremely difficult. Immunologic parameters have also a major effect on the incidence of CAN. Again, the two treatment groups were homogeneous for variables such as histocompatibility differences, presensitization and acute rejection episodes. With our limited number of patients in each arm, it was not that surprising to observe an identical effect of AZA and MMF on the incidence of acute rejection, whereas numerous studies have convincingly demonstrated that MMF is superior to AZA in the prevention of acute rejection episodes. However, we believe that one cannot rule out the possibility that part of the effect of MMF on CAN is mediated by a better control of acute rejection. We observed a nonsignificant trend of more severe acute rejection in the AZA group compared with the MMF group. Moreover, acute rejections in the MMF group were more prone to favorably respond to corticosteroid treatment. In fact, once censored patients with acute rejection, there was no significant difference between the two groups regarding the onset of CAN at 1 year post-transplantation (data not shown). In this context, subclinical acute rejection could also play an important role in the development of CAN at 1 year. Nankivell et al. have recently demonstrated the role of early subclinical acute rejection as a risk factor of CAN and found a protective effect of MMF in the development of CAN (16). In our study, the general incidence of CMV infection was high probably because of the use of ATG and the absence of CMV prophylaxis. However, among the few variables that differentiate MMF from AZA patients, the significantly higher incidence of CMV infection in the latter group, and even more marked CMV disease, is important to consider. This surprisingly higher incidence of CMV infection in the AZA group could have originated from the fact that overall the patients in this group received more antirejection therapy, as they were more prone to acute rejection. There was also a nonsignificant trend of observing more patients with a very low risk of CMV infection (D–R–) in the MMF group compared with the AZA group. The importance of CMV infection in the development of CAN has been previously suggested in a study based on protocol biopsies with a different immunosuppressive regimen (21). Analyzing their historical cohort, Giral et al. (22) demonstrated that (i) CMV disease is a risk factor for graft loss in patients treated with AZA, and (ii) MMF therapy abrogates the CMV-related long-term kidney graft alterations. Even if the effect we observed of MMF on CAN could be CMV-mediated, we cannot rule out a direct influence of MMF. In vitro, mycophenolic acid inhibits vascular smooth cell proliferation at concentrations that are in the range of therapeutical use (10). In various animal models, it has been demonstrated that MMF could beneficially influence the onset of CAN by its specific action on smooth muscle cell and intragraft cell migration and proliferation (11,23,24).
Finally, although this effect of MMF had not been previously demonstrated in a randomized study, it is interesting to note that Ojo et al. (25), using a very different approach, reached similar conclusion. Indeed, they analyzed the risk factors for CAN in the large database of the US Scientific Transplant Registry and showed a protective effect of MMF on long-term survival, partly caused by the decrease in the incidence of acute rejection. This result was confirmed in a recent study, based on the analysis of almost 50 000 primary renal allograft recipients reported to the United States Renal Data System, demonstrating that continuous use of MMF vs. azathioprine was associated with an improved renal function and a decrease in the risk and incidence of renal deterioration (26).
In conclusion, we provide data indicating that kidney grafts treated with MMF had a lower incidence of CAN compared with patients treated with AZA, which suggests, in addition to already published studies, a beneficial effect of this immunosuppressant in the course of kidney transplantation. These findings also underline the interest of protocol biopsies as an irreplaceable tool to analyze the prophylactic intervention of CAN.
Acknowledgments
The authors thank the surgical team, especially Prs J.L. Pariente and J.M. Ferrière, for performing donor biopsy, and Dr A. Modesto for performing renal histology. We also thank Pr C. Combe, and Drs Ph. Chauveau and J. Déchanet for critical reading of the manuscript.
This work was financially supported partly by the Association Promotion Transplantation Rénale and the Centre Hospitalier Universitaire de Bordeaux.
