Abstract
Maintenance immunosuppression with cyclosporine (CsA) is associated with nephrotoxicity, hyperlipidemia, and hypertension. This long-term study (core study + 4 yr of follow-up) investigated the long-term efficacy and safety of CsA withdrawal from a mycophenolate mofetil (MMF)-based regimen. Seventy-seven patients were maintained on CsA, MMF, and steroids (CsA-MMF group), and 74 were given a CsA-free regimen of MMF and steroids (MMF group). Serum creatinine and creatinine clearance were measured at 6-month intervals. Patient and graft survival, acute rejection episodes, malignancies, BP, and lipid profile were also recorded. At 5 yr, patient and graft survival was 93 and 88%, respectively, for the MMF group and 95 and 92%, respectively, for the CsA-MMF group. During follow-up, seven MMF patients experienced acute rejection episodes compared with one CsA-MMF patient (P = 0.0283). Nine grafts were lost to chronic rejection in the MMF group versus three in the CsA-MMF group. No demographic or immunologic characteristics were associated with acute or chronic rejection in the MMF group, but the doses of both MMF and steroids decreased significantly between 1 and 5 yr. The MMF group showed a trend toward improved creatinine clearance (67.4 versus 61.7 ml/min; P = 0.0500). Withdrawal of CsA from an MMF-containing immunosuppressive regimen resulted in an increased risk for acute rejection episodes and graft loss as a result of rejection throughout the 5-yr study period. The creatinine clearance–confirmed improvement in renal function observed at year 1 was maintained at 5 yr BP and cholesterol levels were well controlled in both groups.
Since 1984, cyclosporine (CsA) (Neoral, R.P. Scherer GmbH, Eberbach/Baden, Germany) has been used routinely in renal immunosuppressive regimens to help prevent acute rejection and improve patient survival. Early results were extremely encouraging, with CsA-steroid combinations evoking a 10 to 20% improvement in graft survival after 1 to 3 yr when compared with azathioprine (AZA)-containing dual-therapy regimens (1,2). Indeed, calcineurin inhibitor–based regimens have now become the cornerstone for preventing renal graft rejection. Unfortunately, despite their efficacy benefits, calcineurin inhibitors are associated with nephrotoxicity, hyperlipidemia, and hypertension (3–6). All of these factors may ultimately affect both graft and patient survival, with a long-term study indicating that any initial benefit is lost after 10 yr (7).
Concerns about CsA toxicity encouraged numerous investigators to evaluate CsA withdrawal strategies, via a conversion from CsA to AZA or through withdrawal of CsA from an AZA-containing regimen. A large meta-analysis of 13 studies that were completed before 1999 found that after CsA discontinuation, there was an incremental increase of 11% in the proportion of patients with acute rejection, when compared with patients who continued on CsA, AZA, and steroids (8).
Despite the increased risk for acute rejection, CsA withdrawal does not adversely affect the risk for graft failure, and long-term data have revealed a trend toward improved graft survival in patients who are weaned from CsA (relative risk for graft failure 0.92) (8). Nevertheless, maintenance therapy with AZA and steroids has not become routine practice, mainly because of the concern about rejection after CsA discontinuation.
Mycophenolate mofetil (MMF) (CellCept, F. Hoffman-LaRoche Ltd., Basel, Switzerland) is more immunosuppressive than AZA when used in combination with CsA and steroids (9). As MMF does not adversely affect kidney function, BP, or lipid levels, it can potentially facilitate safe and effective CsA withdrawal. Several prospective, randomized studies have investigated the short-term effect of CsA withdrawal from MMF-containing regimens. As with studies involving AZA, the investigators reported an improvement in renal function. However, incremental increases in acute rejection were also observed after CsA withdrawal (e.g., 6.3% [10] and 20.6% [11]).
In 1997, a 1-year, randomized, controlled, multicenter study was initiated to investigate the withdrawal of CsA in stable renal transplant recipients who received CsA, MMF, and corticosteroids (12). Rejection incidence was 10.6% in the CsA-withdrawal group versus 2.4% in the CsA-continuation group (P < 0.05). There was a trend to improved renal function in the CsA-withdrawal group.
This article describes the 4-yr follow-up of that core study and seeks to present a clearer picture of the long-term efficacy and safety benefits of CsA withdrawal in the presence of MMF. Specifically, the principal objectives of the study were to observe the long-term rates of acute rejection and graft loss and to analyze the long-term effects on serum creatinine levels, creatinine clearance, cholesterol levels, and BP.
Materials and Methods
Patients
During the core study, 170 stable patients who were treated with CsA + MMF 2 g + steroids were randomized to either continue CsA (n = 85; CsA-MMF group) or have CsA weaned over a period of 12 wk (n = 85; MMF group). Patients were followed for 9 mo from the start of CsA withdrawal.
As described previously (12), patients who were eligible for the core study were recipients of first or second cadaveric or living donor kidney transplants, were between 12 and 30 mo posttransplantation, were maintained on a CsA-based regimen, had stable renal function (serum creatinine <300 μmol/L for 3 mo before study entry), had experienced no more than one rejection episode after transplantation and no rejection episodes for 3 months before enrollment, and had panel reactive antibodies <50% at the time of transplantation.
The trial was performed in accordance with the Declaration of Helsinki (as amended in Tokyo, Venice, and Hong Kong), Good Clinical Practice guidelines, and all local laws and regulations concerning clinical trials. Informed consent was obtained from all patients.
The 157 patients who completed the core study were invited to continue to follow-up, and 151 entered the 4-yr follow-up. The MMF group comprised 74 patients; the CsA-MMF group contained the remaining 77 patients (Figure 1). Demographic data were identical to the core study, with no significant intergroup differences in age (total mean 46.7 yr; range 18 to 70 yr), race (total percentage 98% white), transplant type (total percentage 89% cadaveric), and number of HLA A+B+DR mismatches (total mean 2.69; range 0 to 6).
Flow chart of patient randomization.
At the start of the 4-yr follow-up period, the renal functions of the two study groups, as assessed by creatinine clearance calculations, were similar (MMF group mean 65.5 ml/min and median 63.2 ml/min, CsA group mean 63.0 ml/min and median 61.7 ml/min). During the follow-up period, patients remained on their initial randomized treatment regimen. Changes in immunosuppression therapy were permitted at the physicians’ discretion.
Recorded Events
Results from the core study have been published elsewhere (12). Unless stated, all data presented here relate to the 4-yr follow-up period only. The clinical events and parameters that were recorded during follow-up included graft loss as defined by a return to chronic dialysis or retransplantation, acute rejection, reason for graft loss (acute or chronic graft rejection, technical complications, recurrence of underlying disease, or other causes), serum creatinine levels, creatinine clearance, cholesterol levels, BP, malignancy, and death. All maintenance doses of MMF, CsA, and steroids were recorded during follow-up, with visits taking place at 6-month intervals. Six patients (three from each group) were lost to follow-up.
Statistical Analyses
The percentage of patients who experienced acute rejections during the 4-yr follow-up period (from the end of the core study to the 5-yr study completion) was analyzed using the log-rank statistic from the Kaplan-Meier analysis. Analysis of covariance (ANCOVA) was used to analyze creatinine clearance and serum creatinine during follow-up. The ANCOVA method tested for differences between the two treatment groups, using the baseline value as covariate, thereby taking into account the differences between the study groups at the start of the study. These results provide a more accurate estimate of the true difference between the groups, when compared with t test calculations. Data are presented with missing values not replaced, using a baseline of time of randomization in the core study. Only patients with functioning grafts were considered. Arbitrary values were not assigned to patients who lost their grafts. Creatinine clearance was calculated using the method proposed by Cockcroft and Gault: creatinine clearance (ml/min) = [(140 − age in years) × weight in kg/(72 × serum creatinine (mg/dl)] (× 0.85 in women).
BP and total cholesterol levels are presented using arithmetic means. The t test and χ2 test (Fisher Exact test) were also used to make comparisons between the groups. All analyses were performed on an intention-to-treat basis, which was defined as all patients entering follow-up (n = 151).
Results
Maintenance Immunosuppression
During the 4-yr follow-up period, steady and statistically significant decreases in MMF and steroids doses were observed in both groups. For patients who had data available at both 1 and 5 yr in the MMF group, the mean daily doses of MMF and steroids were 2.0 g and 10.0 mg at year 1, decreasing to 1.9 g and 7.4 mg by year 5 (Table 1). In the CsA-MMF group, the corresponding daily doses of MMF and steroids were 1.9 g and 7.8 mg, decreasing to 1.7 g and 5.9 mg. In addition, the CsA-MMF group showed statistically significant reductions in CsA dosing and trough levels. For patients in the CsA-MMF group who had data available at both 1 and 5 yr, the mean CsA dose was approximately 2.5 mg/kg per d, with a mean trough level of 122 ng/ml at 5 yr. At 5 yr, 19 of the 39 patients with levels available at this time had trough levels <100 ng/ml. Eleven of these 19 patients had trough levels <75 ng/ml. Thirteen patients in the MMF group received additional CsA during follow-up. For five of these patients, CsA maintenance therapy was started within 6 mo of a rejection episode.
Comparison of the dosing of MMF, steroids, and CsA (dose and trough levels) across the four-year follow-up perioda
Acute Rejection Episodes, Graft Loss, and Death
During the 4-yr follow-up period, seven (10%) patients in the MMF group and one (1%) patient in the CsA-MMF group experienced acute rejection episodes (Table 2). One of the seven patients in the MMF group started receiving CsA in response to an acute rejection episode. For four patients, CsA was added to their regimen within the 6-mo period during which the acute rejection episode occurred, and the remaining two patients were not administered CsA. When the core study and follow-up were combined, the total number of patients who experienced acute rejection episodes was 12 (16%) in the MMF group and 1 (1%) in the CsA-MMF group (P = 0.0029).
Incidences of acute rejection, graft loss, and death rates during the 4-yr follow-up period (1 to 5 yr after randomization)a
Graft loss (excluding death) occurred in nine (12%) MMF-group patients and six (8%) CsA-MMF group patients. The nine graft losses in the MMF group were due to acute (n = 2) or chronic (n = 7) rejection. Cross-analysis showed that when compared with patients with functioning grafts (n = 132), patients who experienced graft losses (n = 15) had significantly reduced renal function at entry to the 4-yr follow-up period (serum creatinine levels 186.2 versus 129.0 μmol/L; P < 0.0001).
Of the seven patients who lost their graft from chronic rejection, no acute rejection episodes were documented during follow-up. Three of the six graft losses in the CsA-MMF group were due to chronic rejection, and three were reported to be due to technical reasons. Of the three patients who lost their graft from chronic rejection, no acute rejection episodes were documented during follow-up. The combined end point of acute rejection or graft loss as a result of rejection occurred in 14 patients in the MMF group and four patients in the CsA-MMF group (log-rank test, P = 0.0099; Figure 2). Of these 14 MMF-group patients, only one had also experienced an acute rejection episode during the core study. Of the four CsA-MMF-group patients, one experienced an acute rejection episode during the core study.
Cumulative incidence of time to first acute rejection or graft loss as a result of rejection. A, acute rejection; G, graft loss.
During the 4-yr follow-up period, there were five (7%) deaths in the MMF group and four (5%) deaths in the CsA-MMF group. As a result, patient survival during this period was 93% in the MMF group and 95% in the CsA-MMF group (NS, log-rank test P = 0.6281). During the core study, one patient had died in the MMF group versus none in the CsA-MMF group. At 5 yr, among the patients who entered follow-up, the proportion of surviving patients who did not experience graft loss was 81% for the MMF group and 87% for the CsA-MMF group (t test, P = 0.3222).
Explorative analyses were performed to identify possible risk factors for acute rejection or graft loss from rejection among patients in the MMF group (n = 74). First, the 14 patients who experienced acute rejection or graft loss from rejection were compared with the 60 who did not (Table 3). Parameters such as the average steroid and MMF doses, living or cadaveric graft, primary or secondary transplant, cold ischemia time, donor and recipient age, HLA compatibility, or panel-reactive antibodies were no different between rejectors and nonrejectors. Furthermore, the occurrence of any rejection episode, either before or during the core study, was not associated with acute rejection or graft loss from rejection during follow-up.
Comparison of variables between patients in the MMF group with or without acute rejection or graft loss from rejection during follow-up
Second, patients were categorized arbitrarily according to their maintenance doses of MMF and steroid. “Higher dose” patients were defined as those who were taking at least a 2-g/d dose of MMF and at least a 10-mg/d dose of steroid for the entire 4-yr follow-up period (14 patients). “Lower dose” patients were defined as those who were taking lower doses at any time during the observation (60 patients). Differences in the number and percentage of patients who had acute rejection, graft loss, or both during the 4-yr follow-up period were compared across these two patient categories. One higher dose patient (1 of 14; 7%) and six lower dose patients (6 of 60; 10%) experienced acute rejection episodes (Fisher exact test, P = 1.000). No higher dose patients (0 of 14; 0%) and nine lower dose patients (9 of 60; 15%) experienced graft loss (Fisher exact test, P = 0.1930). One higher dose patient (1 of 14; 7%) and 13 lower dose patients (13 of 60; 22%) experienced either acute rejections or graft loss (Fisher exact test, P = 0.2817). Because of the small numbers of patients involved, it is difficult to draw strong conclusions. The results nevertheless suggest that lower dose patients experience more acute rejection episodes or graft loss.
Renal Function
At the end of the 5-yr total study period, serum creatinine was 133.8 μmol/L in the MMF group and 135.4 μmol/L in the CsA-MMF group (NS; Figure 3). During the 5-yr total study period, creatinine clearance was consistently higher for patients in the MMF group, when compared with the CsA-MMF group. At 5 yr, creatinine clearance was 67.4 ml/min in the MMF group and 61.7 ml/min in the CsA-MMF group (P = 0.0500; Figure 4).
Least-square means of serum creatinine by visit (not replacing missing values and using randomization visit as baseline).
Least-square means of creatinine clearance (calculated) by visit (not replacing missing values and using randomization visit as baseline). P values for the differences between treatment groups according to the analysis of covariance; *P < 0.05; at 5 yr, P = 0.0500.
Lipid Profile, BP, and Malignancies
At the end of the follow-up period, total cholesterol was 5.23 mmol/L in the MMF group and 5.38 mmol/L in the CsA-MMF group. Systolic BP was 133.2 mmHg in the MMF group and 136.8 mmHg in the CsA-MMF group. Diastolic BP in the MMF group was 78.3 mmHg versus 80.0 mmHg in the CsA-MMF group. Four (5%) patients in the MMF group and seven (9%) patients in the CsA-MMF group reported malignancies. Malignancies in the MMF group comprised skin and appendages (n = 2) and urogenital system malignancies (n = 2). Malignancies in the CsA-MMF group comprised whole body (n = 2), digestive system (n = 1), breast (n = 1), and skin and appendages (n = 2).
Discussion
This is the first article to report 5-yr CsA-withdrawal data from a triple-therapy regimen that comprised CsA, MMF, and steroids in a randomized, prospective trial. Our initial hypothesis was that the increased immunosuppressive potency of MMF, relative to AZA, would minimize the long-term risk for rejection episodes after CsA withdrawal. However, at 5 yr, the incidence of acute rejection and graft loss from rejection was significantly higher after CsA withdrawal. These events occurred throughout the total study period, during both the 1-year core study and the 4-year follow-up. Further analyses were performed to identify possible risk factors for acute rejection and graft loss from rejection among MMF-group patients.
Comparisons between rejectors and nonrejectors for demographic parameters such as the proportion of cadaveric grafts, secondary transplants, cold ischemia time, donor/recipient age, and HLA mismatches revealed no significant differences. A history of rejection before or during the core study was not a predictor for subsequent acute rejection episodes or graft loss from rejection for up to 5 yr. In addition, there were no significant differences in the average daily MMF or steroid doses of rejectors and nonrejectors. Nevertheless, for all patients in the study, doses of MMF and steroids decreased steadily throughout the study, with the mean steroid dose decreasing by approximately 25% between year 1 and year 5. It therefore is possible that the reduced level of immunosuppression was a contributory factor for the occurrence of acute rejections and graft losses from rejection within the MMF group. In support of this hypothesis, patients in the MMF group who received <2 g/d MMF and/or 10 mg/d steroids at any time during the course of the study follow-up seemed to be at a higher risk for acute rejection and graft loss.
If MMF or steroid dose reductions are critical factors for increased rejection, then one may speculate that mycophenolic acid monitoring may have helped to avoid underimmunosuppression and rejection in this setting. Accordingly, a 12-mo study of 77 patients found that when CsA was withdrawn at 3 mo from a triple-therapy regimen that comprised MMF + steroids, a low mycophenolic acid AUC0 to 12 at the time of CsA discontinuation was a predictor for later acute rejection (13). The hypothesis that therapeutic drug monitoring of MMF may be advantageous when CsA is withdrawn is also being investigated in an ongoing study by Dr. J. de Fijter in Leiden, Netherlands. With regard to steroids, as they are not routinely monitored, it may be wise to maintain these dual-therapy patients on at least 10 mg/d steroids.
There are no 5-yr data investigating CsA withdrawal strategies involving sirolimus, which is at present the main other non-nephrotoxic immunosuppressive drug used for primary immunosuppression. A 3-yr CsA withdrawal trial involving 525 patients who took a sirolimus + steroid combination has been reported (14). Because of the differences in study design and patient population, a direct comparison of these findings with our results is difficult. However, the sirolimus study is an example of successful therapeutic drug monitoring when CsA is withdrawn. Two to 3 mo after transplantation, patients were randomized to have CsA continued or withdrawn. At 3 yr, compared with that of the CsA-continuation group (5.6%), there was only a modest, nonsignificant higher rejection rate (10.2%) in the CsA withdrawal group. These favorable results were likely to be due to therapeutic drug monitoring of sirolimus, which allowed for adequate drug exposure and probably helped to avoid underimmunosuppression. At the end of the 3-yr study period, intention-to-treat analysis showed that serum creatinine was 168 μmol/L in the CsA-sirolimus group and 145 μmol/L in the sirolimus group (P = 0.002) (14). This study indicates that improved renal function can be seen after CsA withdrawal, provided that the incidence of kidney graft rejection remains under control.
Our findings showed that CsA withdrawal was associated with an improvement in calculated creatinine clearance of 6 ml/min after 5 yr. This is similar to the improvement observed at the end of the core study (12), indicating that improvements were maintained over the 5-yr total study period. Minimizing or eliminating the use of calcineurin inhibitors remains an important goal in transplantation, as impaired renal function has emerged as a cardiovascular risk factor both in the general population and in transplant recipients. A retrospective analysis of 58,900 patients who had sustained a fatal cardiovascular event beyond 1 yr after transplantation found that serum creatinine at 1 yr was strongly associated with the incidence of cardiovascular death and was independent of any known risk factors (15).
At 5 yr, our findings show no evidence to suggest that there was an increase in cardiovascular-related deaths in the CsA-MMF group. In contrast, an earlier long-term, CsA-withdrawal study by Hollander et al. (16) found that the frequency of cardiovascular death with a functioning graft was 8% higher in the group that remained on CsA. There are at least two cardiovascular risk factors that were less well controlled in the Hollander trial, compared with our current study: BP and cholesterol. During the Hollander trial, relatively high 5-yr mean cholesterol levels in the CsA group were recorded (7.5 mmol/L). In our study, the cholesterol levels were well controlled, resulting in a mean value at 5 yr for the CsA-MMF group of 5.4 mmol/L. Similarly, the mean BP levels at 5 yr in the Hollander trial were 145/88 mmHg, as compared with 137/80 mmHg in our study. Although the use of concomitant therapies was not documented in our study, it is likely that the tight BP and cholesterol control contributed to the low cardiovascular death rate among these patients. Furthermore, the average CsA dose in the CsA-MMF group was relatively low (2.5 mg/kg per d). This may have contributed to the improved survival rate and the maintenance of renal function during the 4-yr follow-up period.
The intergroup differences in malignancy rate observed in our study (5% in the MMF group versus 9% in the CsA-MMF group) are similar to differences reported in other CsA-withdrawal trials. For example, in the sirolimus trial detailed earlier, the total malignancy rate of the CsA-free arm was 5.6 versus 11.2% in the CsA group, underscoring the advantage of a CsA-free regimen to prevent cancer occurrence (14).
In conclusion, our results show that the tendency of MMF-steroid dual therapy to improve renal function is at the cost of increased acute and chronic rejection. Although no statistically significant immunologic risk factors have been identified, our results suggest that insufficient immunosuppression may have played a role in triggering rejection in a small subset of dual-therapy patients. We propose that sufficient immunosuppression should be maintained for patients on MMF-steroid dual therapy via MMF therapeutic drug monitoring or an adherence to doses of at least 2 g of MMF and 10 mg of steroids. Alternatively, a triple-therapy regimen of MMF, steroids, and low-dose CsA has been reported to be successful in allowing for improved renal function and good prevention of acute rejection episodes (17).
Acknowledgments
This study was funded by a grant from F. Hoffmann-La Roche Ltd., Basel, Switzerland. We acknowledge the participation of the following centers:
Prof. D. Abramowicz, Hôpital Erasme, Belgium; Prof. J. Klempnauer, Medizinische Hochschule Hannover, Germany; Dr. M. del Rosario Brunnet, CEMIC, Argentina; Dr. D. Cassadei, INBA, Argentina; Dr. D. del Castillo, Hospital Reina Sofía, Spain; Dr. A. Spallini Ferraz, HCFMRP Camprio Universitário, Brazil; Dr. J. Grinyó, Hospital de Bellvitge, Spain; Dr. A. Kribben, Universitätsklinikum Essen, Germany; Dr. P. Peeters, Dr. W. Terryn, Universitair Ziekenhuis, Belgium; Prof. M. Lao, Warsaw Medical School, Poland; Dr. N. Lefrançois, Dr. R. Gallard Hôpital Edouard Herriot, France; Mr. D. Manas, Freeman Hospital, UK; Prof. R. Margreiter, Universitätsklinik, Innsbruck, Austria; Dr. J. Morales, Hospital 12 de Octobre, Spain; Dr. A. Mota, Hospitais da Universidade de Coimbra, Portugal; Prof. G. Mourad, Hôpital Lapeyronie, France; Dr. F. Oppenheimer, Hospital Clinico, Spain; Prof. E. Pohanka, Universitätsklinik für Innere Medizin II, Wien, Austria; Prof. Y. Vanrenterghem, Universitair Ziekenhuis Gasthuisberg, Belgium; and Dr. S. Vitko, Institute for Clinical and Experimental Medicine, Prague, Czech Republic.
- © 2005 American Society of Nephrology
References
Jump to section
More in this TOC Section
Cited By...
- Adverse Outcomes of Tacrolimus Withdrawal in Immune-Quiescent Kidney Transplant Recipients
- Minimizing Immunosuppression, an Alternative Approach to Reducing Side Effects: Objectives and Interim Result
- Untreated Rejection in 6-Month Protocol Biopsies Is Not Associated with Fibrosis in Serial Biopsies or with Loss of Graft Function