Mycophenolate mofetil in combination with reduction of calcineurin inhibitors for chronic renal dysfunction after liver transplantation


  • Presented in part at the American Transplant Congress, Seattle, WA, May 21–25, 2005.


The purpose of the study was to introduce mycophenolate mofetil (MMF) in liver transplant recipients with renal dysfunction to decrease calcineurin inhibitor (CNI) dosages without increasing rejection risk. In this prospective, multicenter, randomized study, chronic CNI-related renal dysfunction was defined by an increase in serum creatinine with values >140 μmol/L and <300 μmol/L. Patients were randomized in 2 groups. Study group: combination of MMF (2 to 3 g/day) and reduced dose of CNI ≥50% of initial dose; control group: no MMF, but with the ability to reduce CNI doses, but not below 75% of initial dose. Fifty-six patients were included, 27 in the study group and 29 in the control group. In the study group, there was a significant decrease in serum creatinine values, from 171.7 ± 24.2 μmol/L at day 0 to 143.4 ± 19 μmol/L at month 12 and a significant increase in creatinine clearance, from 42.6 ± 10.9 mL/min to 51.7 ± 13.8 mL/min. No rejection episode was observed in the study group. In the control group, there was no improvement of renal function, assessed by the changes in serum creatinine values, from 175.4 ± 23.4 μmol/L at day 0 to 181.6 ± 63 μmol/L at month 12, and in creatinine clearance, from 42.8 ± 12.8 mL/min to 44.8 ± 19.7 mL/min. The differences between the 2 groups were significant: P = 0.001 for serum creatinine, and P = 0.04 for creatinine clearance. In conclusion, the introduction of MMF combined with the reduction of at least 50% of CNI dose allowed the renal function of liver transplant recipients to significantly improve at 1 year, without any rejection episode and without significant secondary effects. Liver Transpl 12:1755–1760, 2006. © 2006 AASLD.

Renal dysfunction is a major problem after liver transplantation, and it has an impact on long-term morbidity and mortality. The incidence of chronic renal disease among liver transplant recipients varies widely, from 10 to 78%.1–3 These variations have many explanations, including lack of a standard definition of posttransplant renal disease, confusion between acute and chronic dysfunction, and variable periods of follow-up. Ojo et al. have reported the results of a population-based cohort analysis involving recipients of non-renal solid organs to determine the incidence of chronic renal failure and the risk of death associated with it.4 Among them, 36,849 patients were liver transplant recipients. The primary endpoint analyzed was chronic renal failure, defined as a glomerular filtration rate of 29 mL/minute/1.73 m2 or less, or the onset of end-stage renal disease, as determined by the initiation of dialysis therapy or pre-emptive kidney transplantation. The 18.1% incidence of chronic renal failure at 5 years among liver transplant recipients, associated with an elevated risk of death after transplantation (relative risk, 4.55), was the key point of this report. This renal dysfunction is multifactorial in origin, but the main cause is calcineurin inhibitor (CNI) acute dose-dependent and chronic non-dose-dependant nephrotoxicity.5

Immunosuppressive drugs without renal side effects have been used increasingly as CNI-sparing agents. Mycophenolate mofetil (MMF) is a non-nephrotoxic drug that inhibits proliferation of T and B lymphocytes and has a proven efficacy in the field of liver transplantation.6 In several prospective (most often open labeled) and retrospective studies, the partial or complete replacement of CNI with MMF, in patients with chronic renal dysfunction, resulted in an increased risk of acute and chronic rejection.7–14 On the other hand, most of these studies showed significant improvement in renal function.11–14

We designed a prospective, multicenter, randomized study to analyze the consequences of introducing MMF and tapering CNI in liver transplant recipients with chronic renal dysfunction secondary to CNI.


CNI, calcineurin inhibitor; MMF, mycophenolate mofetil.


The study protocol was approved by an internal ethical review board. Patients who received liver transplant more than 1 year prior and who developed CNI-related chronic renal dysfunction were enrolled in the study. CNI-related chronic renal dysfunction was defined as follows: persistent increase in serum creatinine with values >140 μmol/L and <300 μmol/L measured on at least 2 successive occasions more than 1 month apart, proteinuria <1g/24 hours, absence of hematuria, absence of renal arteries stenosis, or urinary tract disease. Graft biopsy was performed in all patients during the 6 months before inclusion; presence of acute rejection signs or significant ductopenia were exclusion criteria. A previous history of rejection, before the 6 months prior to enrollment, was not an exclusion criterion. After informed consent was obtained, patients were randomized. The study was designed for a duration of 24 months, with analyses at 12 and 24 months. The primary endpoint was the evolution of renal function assessed by serum creatinine and creatinine clearance estimated using the formula of Cockcroft and Gault. Secondary endpoints were rejection episodes, biopsy-proven or presumptive, patient and graft survival, and side effects after 12 and 24 months. Patients were assessed with physical examination and laboratory testing every 15 days during the first 3 months, then every 3 months.

In the study group, there was a first phase of 3 months. During this phase, MMF was introduced at 500 mg twice daily up to 2,000 to 3,000 mg/day. MMF dose adjustment was allowed for patients with adverse events. After achieving the full dose of MMF, CNI was slowly tapered up to at least 50% of the initial dose. At the end of this phase, all patients were under the full dose of MMF and a 50% reduced dose of CNI. The controls remained on their previous CNI treatment with the option to reduce dosage, but never below 75% of the initial dose. If an increase of serum creatinine ≥ 30% of baseline was observed, the patient was withdrawn from the study.

Statistical Analysis

All analyses were done with SAS (version 8) software (SAS Institute, Cary, NC). Comparisons were made using chi-square and Fisher exact test for qualitative variables, and Student's t test for quantitative variables. A P value of 0.05 was regarded as significant.


Fifty-six patients were included in the study: 27 study patients and 29 controls. There were no significant differences between study group and control group baseline data (Table 1). The majority of patients were transplanted for alcoholic liver disease. In each group, 20 patients were under CNI monotherapy, with an equal distribution between cyclosporin and tacrolimus. The median time since transplantation was more than 5 years for both groups. The evolution of immunosuppressant doses and CNI concentrations are shown in Table 2. The protocol design was totally respected with a decrease of at least 50% of the initial dose in the study group (−62% for cyclosporin, and −54% for tacrolimus) and not more than 25% in the controls (−8.4% and −11.3%, respectively). At month 12, CNI blood trough levels were infratherapeutic. In the study group, at month 12, CNI blood trough levels were subtherapeutic and the median dose of MMF.

Table 1. Baseline Data of Patients and Controls
 Study Patients (n = 27)Controls (n = 29)
  • NOTE: There were no significant differences between study and control group baseline data.

  • Abbreviations: LT, liver transplantation; HCV, hepatitis C virus; HBV, hepatitis B virus; PBC, primary biliary cirrhosis; HCC, hepatocellular carcinoma.

  • *

    Some patients had more than 1 indication for LT.

Mean age (years)58.8 ± 8.657.7 ± 10.7
Indications for LT*  
Median time after LT (years)5.2 ± 3.35.7 ± 3.2
Immunosuppressants at study entry  
 Steroids (mg/day)7 (5.2 ± 2.3)7 (5 ± 1.4)
 Azathioprine (mg/day)03 (50)
Table 2. Immunosuppressive Treatments During the Study
 Study Patients (n = 27)Controls (n = 29)
  1. NOTE: Results are expressed as median values.

  2. Abbreviation: NA, not applicable.

 Dose [Concentration C0] mg/d [ng/mL]
Cyclosporin (n = 32)n = 15n = 17
 Day 0162.3 [122]165.8 [204]
 Month 674.3 [32]145.3 [260]
 Month 1257.5 [26]147.8 [263]
Tacrolimus (n = 24)n = 12n = 12
 Day 04.4 [7.4]3.2 [6.4]
 Month 62.1 [4.5]2.9 [5.5]
 Month 122.01 [4.4]2.84 [5.5]
 δ−54%−11.3 %
MMF (n = 27)  
 Day 00NA
 Month 61,951.9NA
 Month 122,000.0NA

Serum creatinine and creatinine clearance values are shown in Tables 3A and 3B. In both, per protocol and intention-to-treat analyses, there was a significant improvement of renal function at month 12 in the study group. The improvement was significant from month 6. The reasons for study withdrawal were 1 death in the study group, and 3 patients with an increase of serum creatinine >30% of baseline in the control group. An analysis of individual patient serum creatinine and creatinine clearance changes was performed. Concerning serum creatinine, we observed the following results in the study group: >10% improvement in 73.1% of patients, no change in 23.1% of patients, and >10% deterioration in 3.8% of patients. In the control group, we observed >10% improvement in 36% of patients, no change in 28% of patients, and >10% deterioration in 36% of patients. Concerning creatinine clearance, we observed the following results in the study group: >10% improvement in 80% of patients, no change in 16% of patients, and >10% deterioration in 4% of patients. In the control group, we observed >10% improvement in 28% of patients, no change in 36% of patients, and >10% deterioration in 36% of patients.

Table 3A. Evolution of Renal Function (Intention-to-Treat Analysis)
 Study Patients (n = 27)P ValueControls (n = 29)
  1. NOTE: Values are expressed as mean ± SD.

Creatinine (μmol/L)   
 Day 0171.7 ± 24.20.37175.4 ± 23.4
 Month 3149.2 ± 25.60.10171.5 ± 60.4
 Month 6144.1 ± 20.10.02169.8 ± 34.9
 Month 12143.4 ± 19.00.001181.6 ± 63
Clearance (mL/min)   
 Day 042.6 ± 10.90.9342.8 ± 12.8
 Month 1251.7 ± 13.80.0444.8 ± 19.7
Table 3B. Evolution of Renal Function (Per Protocol Analysis)
 Study Patients (n = 26)P ValueControls (n = 26)
  1. NOTE: Values are expressed as mean ± SD.

Creatinine (μmol/L)   
 Day 0170.9 ± 24.30.5173.2 ± 21.8
 Month 3148.1 ± 24.10.10173.1 ± 21.7
 Month 6143.9 ± 230.01172.8 ± 38.2
 Month 12142.8 ± 19.10.004170.5 ± 42.9
Clearance (mL/min)   
 Day 042.6 ± 11.10.9242.9 ± 12.6
 Month 1251.6 ± 14.080.0444.6 ± 19.3

No rejection episode was observed in the study group; 1 episode was observed in the control group and successfully treated with steroid pulse. There were no differences between the 2 groups concerning the evolution of liver function tests (Table 4). However, phosphatase alkaline level was significantly higher in the control group at baseline and at month 12.

Table 4. Evolution of Function Liver Function Test Results
 Study Patients (n = 27)Controls (n = 29)P Value
Month 0Month 12Month 0Month 12
  1. NOTE: Results are expressed as median values.

  2. Abbreviations: AST, aspartate aminotransferase; ALT, alanine aminotransferase; GGT, gamma-glutamyltransferase; Al Ph, alkaline phosphatase.

AST (IU/L)282324230.78
ALT (IU/L)242522240.87
GGT (IU/L)374133400.73
Al Ph (IU/L)80951171210.05
Total bilirubin (μmol/L)121211120.17

Both patient and graft survival rates were 96% in the study group and 100% in the control group. One patient from the study group died 6 months after beginning MMF treatment, as a result of respiratory failure related to pulmonary fibrosis diagnosed before introduction of MMF.

Non-graft-related adverse events are reported in Table 5. There was no significant difference between the 2 groups concerning the occurrence of adverse events.

Table 5. Non-Graft-Related Adverse Events
 Study Patients (n = 27)Controls (n = 29)P Value
↑ Creatinine >30%140.35
Gastrointestinal complaints520.24
Neuropsychiatric symptoms140.35

De novo malignancies were diagnosed in 4 patients: lung, tongue, and oropharynx in the study group, and prostate in the control group. Gastrointestinal complaints were reported in both groups: diarrhea, nausea, and gastritis in the study group, and diarrhea and bloody stools in the control group. Severe gastritis was observed in 1 patient in the study group, necessitating MMF cessation during 5 days and dose reduction thereafter. Hematologic complications were rare, with only 1 case of mild thrombocytopenia in the study group. Neuropsychiatric symptoms included nonspecific malaise in 1 patient in the study group, and depression in 3 patients and nonspecific malaise in 1 patient in the control group.


Herein we have presented the first prospective, randomized study with untreated arm designed to evaluate the safety and efficacy of the combination of MMF with the reduction of CNI dosage in liver transplant recipients with CNI-related chronic renal dysfunction. This regimen was found to result in significant improvement of renal function, as measured by serum creatinine levels, without increasing the risk of rejection. This study warrants several comments.

First, the definition and the causes of chronic renal failure after liver transplantation need to be clarified. Numerous studies used a definition of chronic renal failure as an elevated serum creatinine level. For instance, using a serum creatinine level above 2.5 mg/dL, Gonwa et al. found the incidence of chronic renal failure at 13 years to be 6%.2 Another study used direct measurements of glomerular filtration rate, such as isotopic or nonisotopic iothalamate clearances, and found the incidence of chronic renal failure at 10 years to be 10%.3 These techniques are cumbersome, time-consuming, and too expensive for use in clinical practice. Ojo et al. used a calculated glomerular filtration rate (from the Modification of Diet in Renal Disease).15 This technique is probably more accurate than use of a serum creatinine measurement alone, which could underestimate the real incidence of chronic renal failure. However, from a practical point of view, intervention in patients experiencing renal dysfunction is implemented when there is a significant rise in serum creatinine, and this was the way we selected the patients in this study.

The majority of authors have attributed the development of chronic renal dysfunction after transplantation to CNI therapy.10, 12–14 In fact, chronic renal dysfunction is more complex than originally sought and should not be classified as CNI nephrotoxicity without eliminating other causes, such as hypertension, diabetes, or drug-induced kidney diseases, questioning the need for renal biopsy.16 In our study, we postulated that diabetic nephropathy was unlikely in the absence of significant proteinuria. None of the patients received nephrotoxic drugs during follow-up after liver transplantation. Thus, from a benefit/risk standpoint, renal biopsy was not performed to support the diagnosis of CNI-related chronic renal dysfunction.

Second, the significant, 50% reduction of CNI dosages was sufficient enough to improve renal function. It must be noted that the cyclosporin levels at study entry were rather high. However, in the control group, despite of the possibility of reducing the CNI dose, without exceeding 25% of initial dose, only 36% of patients had a > 10% improvement of renal function, compared to 73.1% in the study group. There have been several reports on the management of CNI-induced nephrotoxicity, with either reduction or complete withdrawal of CNI.7–14 Most resulted in improved renal function. The point is to assess whether reduction is sufficient enough to reverse renal lesions. Chronic renal nephrotoxicity is partly dose-dependent, but it can also occur in the presence of low blood levels.17 Thus, it has often been considered to represent irreversible damage.18 It is now suggested, by the efficacy of CNI dosage reduction on improvement of renal function, that there is a component of reversible functional impairment in chronic CNI renal dysfunction.11, 19 In our study, we chose to reduce and not to withdraw CNI to avoid the risk of rejection, and approval was validated.

Third, no rejection episode was observed in the study group. Two points are relevant in our study: we included patients with long-term follow-up, more than 1 year after transplantation, and the majority of patients in our population had alcoholic liver disease. Long-term and alcoholic patients are less prone to rejection.14 We feel that this good result is also related to the maintenance of low doses of CNI. Indeed, MMF monotherapy carries a significant risk of rejection, described in 6 to 38% of patients.7–14 Some authors have reported severe acute and chronic rejection episodes, with the need for retransplantation in some cases.8 We have to be very cautious with the late-onset rejection episodes, which are responsible for a decreased graft survival, contrary to the early episodes. Moreover, it is critical to consider the risks associated with rejection therapy, such as a worsening of recurrent hepatitis C with corticosteroid pulse.20 We consider that CNI reduction is preferable to complete withdrawal especially in the absence of validated monitoring of mycophenolate therapy in the setting of liver transplantation.

A moderate pattern of side effects occurred in our study group. Most side effects were gastrointestinal, common with MMF, but none were responsible for MMF withdrawal. It is surprising that side effects of MMF were not a relevant problem in our patients, because the effects are usually observed after introduction in the long-term course after liver transplantation.11 One explanation is that we used a 2 g mean daily dose of MMF rather than the 3 g dose usually proposed.

One patient in the study group died from acute respiratory failure complicating severe pulmonary fibrosis. It must be emphasized that pulmonary fibrosis existed before the patient's enrollment in the study and the administration of MMF. However, some cases of progressive pulmonary fibrosis secondary to MMF administration have already been described.21, 22 We cannot exclude a worsening of the pulmonary fibrosis in our patient, related to MMF administration.

In conclusion, in this prospective, multicenter, randomized study, which is the first with an untreated control arm, the results at 1 year have shown that the introduction of MMF combined with at least a 50% reduction of CNI dose allowed the renal function of liver transplant recipients to improve significantly without any rejection episodes or significant side effects.


Members of the French multicenter group included Lionel Rostaing, CHU Toulouse; Christophe Duvoux, Daniel Cherqui, CHU Créteil; Yvon Calmus, Filomena Conti, CHU Cochin; Claire Vanlemmens, Solange Bresson-Hadni, Jean-Philippe Miguet, CHU Besançon; Anne Minello, Patrick Hillon, CHU Dijon; Sylvie Radenne, CHU Lyon; Danielle Botta-Friedlund, Jean Hardgwissen, Paul Castellani, Pierre-Yves Le Treut, CHU Marseille; Pierre-Henri Bernard, Martine Neau-Cranssac, CHU Bordeaux; Georges-Philippe Pageaux, Lucille Vercambre, Emilie Raynaud, Pierre Puche, Michaël Bismuth, Françis Navarro, Dominique Larrey, CHU Montpellier France.

Following is the number of patients from each of the 9 institutions, for a total of 56 patients: Montpellier, 20; Toulouse, 13; Cochin, 5; Créteil, 5; Besançon, 4; Marseille, 4; Bordeaux, 3; Dijon, 1; Lyon, 1.