NDT Advance Access originally published online on March 30, 2006
Nephrology Dialysis Transplantation 2006 21(8):2256-2262; doi:10.1093/ndt/gfl134
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© The Author [2006]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org
Original Articles: Dialysis and Transplantation
Switching immunosuppression medications after renal transplantation—a common practice
1 Department of Internal Medicine, University of Florida and 2 ProSanos Corporation, La Jolla, California, USA
Correspondence and offprint requests to: Herwig-Ulf Meier-Kriesche, Division of Nephrology, University of Florida College of Medicine, 1600 SW Archer Road, Box 100224, Gainesville, FL 32610-0224, USA. Email: Meierhu{at}medicine.ufl.edu
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and methodsResultsDiscussionConclusionReferences |
|---|
Background. The rate of change to immunosuppression discharge regimens over time is unknown. We examined the frequency of changes to initial drug treatment regimens and factors associated with a drug change following renal transplantation.
Methods. Scientific Registry of Transplant Recipients data from adult recipients who underwent primary renal transplantation between January 1998 and December 2002 were analysed. The Kaplan–Meier analysis was used to determine the frequency of regimen changes for the most common immunosuppression discharge regimens, type of change, and to examine switching between the calcineurin inhibitors tacrolimus (Tacro) and ciclosporin United States Pharmacopera (USP) modified (CsA). Cox proportional hazard regression was used to examine recipient, donor and transplant characteristics associated with a drug change.
Results. The majority of patients experienced a change to their discharge regimen post-transplantation, and more changes were observed within higher-risk sub-groups of patients. Switching from CsA to Tacro was more common than Tacro to CsA. Significant factors associated with a drug change included those associated with graft loss.
Conclusions. Significant immunosuppression regimen changes occur during the first 4 years post-transplantation. It is possible that early graft survival benefits proven in prospective clinical trials may not translate into long-term success in clinical practice, possibly in part because efficacious regimens are not necessarily maintained long-term.
Keywords: immunosuppression; mycophenolate mofetil; renal transplantation; tacrolimus
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionConclusionReferences |
|---|
The mid-1980s and the early 1990s offered few post-transplantation drug treatment choices; kidney transplant recipients generally received corticosteroids (CS) in combination with ciclosporin and azathioprine. In the US, since 1993, a variety of immunosuppression drugs have been approved by the Food and Drug Administration (FDA) [e.g. tacrolimus and ciclosporin for micro-emulsion, approved in 1994; mycophenolate mofetil (MMF) in 1995, sirolimus (Rapa) in 1999, rabbit antithymocyte globulin (ATG) in 1999, daclizumab (Zenapax) in 1999, and basiliximab (Simulect) in 2000]. The availability of these, often more potent, drugs has offered more options for specific clinical needs, and their use has contributed to a substantial decrease in acute rejection. In fact, acute rejection rates have been reduced by more than half between 1995 and 2000 [1,2].
Acute rejection has been demonstrated to be strongly associated with decreased graft survival, and thus it was reasonable to expect that the decrease in acute rejection would translate into substantial improvements in graft survival. During the period when the significant reductions in acute rejection were achieved, 1-year graft survival continued to improve. Yet surprisingly, long-term graft survival beyond 1 year did not follow this trend [1].
Why the successful reduction of early acute rejection has not resulted in the improvement in long-term graft survival is not fully understood. One hypothesis is that acute rejection may be a mere marker of high-risk patients, rather than a critical event in the process that leads to graft failure. High-risk patients are defined by other factors such as a high panel reactive antibody (PRA), re-transplantation, being an African American recipient, delayed graft function (DGF), in addition to acute rejection. Preventing acute rejection therefore, does not alter these intrinsic and other immunological factors, which are still in play to impact graft failure. Another hypothesis is that the more potent immunosuppression used to prevent acute rejection may result in other insults (e.g. nephrotoxicity, infections) which negate a survival benefit of preventing acute rejection.
Like elucidating the role of acute rejection in long-term graft survival, it is also difficult to evaluate the relationship between immunosuppression and long-term graft survival. Since long-term clinical trial data are not available, cause and effect cannot be described, but possibilities can be considered. First, the efficacy of immunosuppression agents to late occurring immunological injury may be different from that for earlier occurring insults. Second, driven by the desire to make therapy regimens more tolerable for patients, physicians could be under-immunosuppressing patients, leading to more late and sub-clinical rejections that ultimately impact long-term graft survival. Thus, the perceived benefit of minimization and withdrawal strategies to prevent long-term nephrotoxic effects and other toxicities may, unfortunately, contribute to graft injury and ultimate graft loss.
Patterns of long-term use of maintenance immunosuppressive drugs have not been previously reported. We examined the rate of changes to five common renal transplant drug regimens prescribed at hospital discharge between 1998 and 2002. Specific type of change was also examined for the two most common drug regimens, as well as the clinical, recipient and donor factors associated with a drug change.
|
Subjects and methods |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionConclusionReferences |
|---|
Patient population and study cohorts
This observational study analysed data from adult renal transplant recipients (age 18–80 years) in the Scientific Registry of Transplant Recipients (SRTR) who received a primary or repeated, single-organ renal transplant between 1 January 1998 and 31 December 2002.
Five study cohorts were created based on a post-transplantation immunosuppression regimen received at hospital discharge: (i) MMF CellCept, Roche Laboratories Inc., Nutley NJ) + tacrolimus (Tacro, Prograf, Astellas Pharma Inc. (formerly Fujisawa Healthcare Inc.), Deerfield, ILL), (ii) MMF + ciclosporin USP modified (CsA, Neoral, Novartis Pharmaceuticals Corp., East Hanover, NJ), (iii) MMF + sirolimus (Rapa, Rapamune, Wyeth Pharmaceuticals Inc., Philadelphia, PA), (iv) Rapa + Tacro and (v) Rapa + CsA. The majority of recipients also received CS; changes to CS, however, were not considered in this analysis.
Outcomes of interest
The objectives of the study were to identify and quantify changes to discharge regimens over time. Therapeutic drugs were considered in two categories: antiproliferatives (MMF, Rapa and azathioprine (AZA)) and calcineurin inhibitors (CNIs) (Tacro, CsA and other ciclosporins). A regimen change was defined as switching to or adding other immunosuppression agent(s) within the same category, or stopping the drug, at each follow-up year. Data on follow-up immunosuppression is collected at 6 months post-transplantation, and then on a yearly basis. If data for a particular follow-up year were available but no immunosuppressive drugs were recorded, it was assumed that no immunosuppressants were taken at this follow-up year and thus counted as a regimen change.
Regimen changes were examined for the five selected discharge regimens described above and were also analysed within sub-sets by donor type (living, deceased and expanded criteria donors), and, were also examined for recipients undergoing re-transplantation. Change patterns within MMF + Tacro and MMF + CsA cohorts were further investigated by examining antiproliferative and CNI components separately as well as examining ‘stopping’ (stopped and not switched to alternatives), ‘switching’ (stopped and switched to alternatives) and ‘addition’ (continued and added alternatives; this category was only applicable for antiproliferative component) categories independently. The rate of switching from Tacro to CsA vs CsA to Tacro was also evaluated. Time to events (e.g. regimen change, switching from Tacro to CsA or vice versa) were computed and analysed using survival analysis methods described in the next section.
Statistical methods
All statistical analyses were performed using the SAS software package (SAS version 8.2, SAS Institute Inc., Cary, North Carolina).
The Kaplan–Meier survival curves were computed to determine the proportion of recipients without changes to their initial discharge regimen. Log-rank statistics were used to compare the Kaplan–Meier curves for time to switching from Tacro to CsA or vice versa between MMF + Tacro and MMF + CsA cohorts. Statistical significance was evaluated at a two-sided level of 5%.
Cox regression models were used to assess the independent effect of covariates on time-to-event endpoints. Covariates in the model included: Human Leukocyte Antigen (HLA) mismatches [yes vs no; missing (n = 338) assumed as yes], recipient race [African American, other than African American or Caucasian vs Caucasian; missing (n = 14) assumed as other], recipient gender (female vs male), recipient age in decades, recipient body mass index (BMI) 
30 [yes vs no; missing height (n = 5406) and weight (n = 6078) were imputed using class-level means based on age, gender and race], induction treatment (yes vs no), cold ischaemia time in hours [missing (n = 9189) data were substituted with population mean], peak percent PRA [% PRA; missing (n = 485) data were substituted with population mean], pre-transplant diabetes [yes vs no; missing (n = 1035) assumed as no], time on dialysis prior to transplantation in years [missing (n = 7499) data were substituted with population mean], donor race [African American vs other than African American; missing (n = 22) assumed as other], donor age in decades [missing (n = 10) data were substituted with population mean], donor type (living vs deceased), expanded donor [yes vs no; missing (n = 16 311) assumed as no], delayed graft function [yes vs no; missing assumed (n = 72) as no] and cytomegalovirus (CMV) infection status [any positive donor (D)/recipient (R) combination (+D/–R; +D/+R; –D/+R) vs negative (–D/–R); missing or unknown (n = 1108) assumed as negative]. Induction treatments included antilymphocyte globulin, anti-thymocyte Globulin (ATG), anti-CD3 monoclonal antibody (OKT3), interleukin-2-toxin (DAB486 – IL–2), Aldesleukin (IL–2), Thymoglobulin, Zenapax and Simulect.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionConclusionReferences |
|---|
Study population
Study cohorts were formed based on a discharge regimen for 60 941 patients between the ages of 18 and 80, who received primary or repeated single-organ renal transplantation between 1998 and 2002. Within the primary renal-transplant population (n = 54 914), immunosuppressive drug change patterns were evaluated for the five most frequently prescribed combination regimens (accounting for 73% of all the discharge regimens) which utilized immunosuppression drugs available after 1995: MMF + Tacro (n = 17 407; 31.7%), MMF+CsA (n = 17 489; 31.8%), MMF+Rapa (n = 1881; 3.4%), Rapa+Tacro (n = 1972; 3.6%) and Rapa+CsA (n = 1609; 2.9%). A total of 40 358 patients (mean age 47.4±13.2 years) were included in these five study cohorts. Demographic and baseline clinical characteristics were similar among these cohorts (Table 1) with the exception of DGF which was higher in the MMF+Rapa cohort; this is likely because centres often switch to non-nephrotoxic alternatives if they suspect or observe DGF. There were 5127 repeated-renal-transplant recipients who were discharged on the five combination regimens mentioned above: MMF+Tacro (n = 3022; 58.9%), MMF+CsA (n = 1368; 26.7%), MMF+Rapa (n = 264; 5.1%), Rapa+Tacro (n = 304; 5.9%) and Rapa+CsA (n = 169; 3.3%).
|
The following results are from the primary renal-transplant population unless otherwise indicated.
Regimen change for top five cohorts and factors associated with regimen change
The Kaplan–Meier estimates of proportions of patients continuing discharge immunosuppressive regimen at each follow-up year within 4 years post-transplantation for the five study cohorts are presented in Figure 1. The results indicated that by 4 years post-transplantation, less than half of all recipients remained on their discharge regimen, accounting for patients who were lost to follow-up or had not reached 4-year follow-up visits (% remaining on initial regimen: MMF + Tacro: 40.7%; MMF+CsA: 30.8%; MMF+Rapa: 25.7%; Rapa+Tacro: 18.5%; Rapa+CsA: 20.2%).
|
Factors associated with an increased (or decreased) chance for regimen change by 4 years post-transplantation, within the five cohorts are presented in Table 2. Recipient factors associated with an increased likelihood for change included being an African American or a female recipient. Donor factors associated with increased likelihood for change included receiving a transplant from an African American donor, an expanded criteria donor or older donor. Transplant factors associated with increased likelihood for a regimen change included any HLA mismatches, a longer cold ischaemia time or delayed graft function. Factors associated with a decreased chance of regimen change included older recipient age, receiving induction therapy and having pre-transplant diabetes.
|
Regimen change patterns for donor type sub-sets
Change patterns were also examined within living (40.4% donors), deceased (59.6% donors) and expanded criteria donors. Patients who received kidneys from living donors remained on their discharge immunosuppressive regimens in higher proportions, compared with patients who received kidneys from deceased donors, in all five cohorts at each follow-up year, except Rapa+Tacro recipients at year 4 (Figure 2). Of the 24 047 patients who received kidneys from deceased donors, 3849 (16% of the deceased donors) were from expanded criteria donors. Proportions of patients remaining on discharge immunosuppressive regimen were even lower for patients who received kidneys from expanded criteria donors: at year 4, only 30.8% MMF+Tacro, 25.1%, MMF + CsA, 19.1% MMF+Rapa and 7.9% Rapa+CsA recipients remained on their initial discharge regimen (Figure 3). No Rapa+Tacro recipients remained on initial discharge regimen at year 4 and this is probably due to a high loss-to-follow-up rate in this cohort (only one patient in this cohort has follow-up duration longer than 4 years).
|
|
Regimen change patterns for repeated transplant recipients
The Kaplan–Meier estimates of proportions of repeated-renal-transplant recipients continuing discharge immunosuppressive regimens at each follow-up year within 4 years post-transplant for the five study cohorts were similar to primary-renal-transplant population (% remaining on initial regimen at year 4: MMF+Tacro: 37.3%; MMF+CsA: 27.8%; MMF+Rapa: 21.5%; Rapa+Tacro: 21.8%; Rapa+CsA: 27.9%).
Regimen, antiproliferative and CNI changes for MMF+Tacro and MMF + CsA cohorts
Regimen changes
The Kaplan–Meier estimates of proportions of patients continuing discharge immunosuppressive regimen at each follow-up year within 4 years post-transplant for the MMF+Tacro cohort were 66.7, 55.6, 47.6 and 40.7%, and for the MMF+CsA cohort were 59.7, 48.0, 38.6 and 30.8%.
Antiproliferative and CNI changes
For the antiproliferative component (MMF), the Kaplan–Meier estimates of proportions of patients continuing MMF without adding or switching to Rapa or AZA, by 4 years post-transplant, for the MMF+Tacro cohort were 46.6% and for the MMF+CsA cohort were 53.7%. Eighty-six percent of MMF+Tacro patients did not switch from MMF to Rapa or AZA by 4-years follow-up, compared with 89% in the MMF+CsA cohort. Eighty-nine percent of MMF+Tacro patients made it through 4 years post-transplant without adding Rapa or AZA to the regimen, compared with 92% in the MMF+CsA cohort. Fifty-eight percent of MMF+Tacro patients never stopped taking MMF by 4 years post-transplant, compared with 63% in the MMF+CsA cohort.
For the CNI component (Tacro and CsA), the Kaplan–Meier estimates of proportions of patients continuing their discharge CNI (without switching to other CNI(s) or stopping the discharge CNI), by 4 years post-transplant, for the MMF+Tacro cohort were 62.8% and for the MMF+CsA cohort were 41.3%. About 89% of MMF+Tacro patients did not switch from Tacro to CsA or another ciclosporin by 4-years follow-up, compared with 56.6% in the MMF+CsA cohort. The proportion of MMF+Tacro patients who never stopped taking Tacro through 4 years post-transplantation was slightly lower than in the MMF+CsA cohort (69.5 vs 71.6%).
Tacro/CsA switching for MMF + Tacro and MMF + CsA cohorts
Rates of switching specifically from Tacro to CsA and CsA to Tacro were evaluated. The Kaplan–Meier estimate of the proportion of patients who did not switch from Tacro to CsA in the MMF+Tacro cohort was statistically significantly higher than the proportion of patients who did not switch from CsA to Tacro in the MMF+CsA cohort by 4 years post-transplantation (91.3 vs 77.3%; P<0.001; Figure 4).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionConclusionReferences |
|---|
In this study, we observed a surprisingly high rate of changes to initial hospital-discharge drug regimens. By 4 years after kidney transplantation, less than half of the recipients were recorded as still being on their initial regimen. We found that the largest proportion of regimen changes occurred by 1 year post-transplantation. This finding is consistent with observations from single centre studies and clinical trials where the majority of dosing and discontinuation changes, whether prompted by recipient side effects or CNI-sparing protocols, are reported to occur within 1 year post-transplantation [3–5]. Many different reasons may motivate a provider to change a treatment regimen, including: poor graft function, adverse side effects, efficacy failure in general, individual centre protocols, economic reasons and availability of newer drugs; non-compliance is also a cause of drug changes. The data recorded in the SRTR database did not allow for a causal assessment of the reasons for a treatment change. The data did allow for observations of recipient, donor and transplant characteristics which are associated with a higher likelihood for a drug change, and we found that these associated factors (i.e. older donor age, longer cold ischaemia time, HLA mismatches, presence of delayed graft function and African American recipient) are also known risk factors for graft loss. This is not surprising, as clinical practice has evolved to make a change to the regimen when efficacy failures occur, in an attempt to boost efficacy. In fact as a group, recipients of grafts from expanded criteria donors, deceased donors and repeat transplant recipients (i.e. the higher risk groups) had a higher proportion of drug changes compared with recipients of grafts from living donors. Within the two most commonly used regimens, MMF+Tacro and MMF+CsA, we observed that about one-third of recipients stopped Tacro or CsA. These CNI stopping patterns likely reflect the increased emphasis on immunosuppression minimization protocols, designed to avert long-term nephrotoxicity. However, whether CNI minimization plays a role in the unrealized improvements in long-term graft survival is yet to be determined, and the impact of drug changes on late rejection and long-term graft survival cannot be determined by the observational data analysed here.
In cases where the CNI was not stopped, but rather one substituted for another, switching was much more common in the MMF+CsA-discharged group than in the MMF + Tacro-discharged group, and was largely driven by the switch from CsA to Tacro. This is concordant with clinical trial data which have shown a higher degree of switching from CsA to Tacro than vice versa [6], and may be indicative of a shift in perception of the best CNI treatment strategy.
From the data available in the SRTR, we are only able to calculate an approximate frequency of drug changes, since we had to make assumptions about missing follow-up data. We did perform a sensitivity analysis to test the assumptions we used in our analyses. If the data for a particular follow-up year were available, but no immunosuppressive medications were recorded followed by another year's record, showing that the patient is still remaining on discharge regimen, given no record showing there was a prior regimen change, it was assumed that immunosuppressive medications were continuously taken at this follow-up year rather than stopped. Results from the sensitivity analysis confirmed that the original analysis results were robust regardless of assumptions for missing immunosuppression data. In this study we were unable to assess changes in dosing, since these data are not reported to the SRTR, thus all observations are in respect to stopping, switching or adding drugs, and it is acknowledged that dose increases or decreases can similarly be hypothesized to possibly impact long-term efficacy outcomes.
|
Conclusion |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionConclusionReferences |
|---|
From the data we present in this study, it is clear that only a minority of patients remain on the immunosuppressive regimens they were originally assigned to after transplantation. Whether these changes are ultimately beneficial or detrimental for long-term graft survival is unknown. Elucidating the role of immunosuppression, and understanding the impact of long-term usage patterns is an important endeavour. Only extended prospective studies will be able to answer the question if continuing maintenance immunosuppression unchanged in the long-term might improve long-term graft survival.
|
Acknowledgments |
|---|
The data reported here have been supplied by the United Network for Organ Sharing and University Renal Research and Education Association under contract with the Department of Health and Human Services. The interpretation and reporting of these data are the responsibility of the authors and do not represent an official policy or interpretation of the United States Government or any of its representatives. The IRB approval or exemption determination is the responsibility of the authors as well. The analyses reported here have been conducted in ProSanos Corporation facilities, and were funded by Astellas Pharma US, Inc. This work was presented in part at the 2005 American Transplant Congress, Seattle, Washington.
Conflict of interest statement. H.U.M.K. serves as a consultant to ProSanos Corporation, Astellas, Novartis, Roche, and has received research support from: Astellas, Genzyme, Novartis, Roche, and Wyeth.
|
References |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionConclusionReferences |
|---|
- Meier-Kriesche HU, Schold JD, Kaplan B. Long-term renal allograft survival: have we made significant progress or is it time to rethink our analytic and therapeutic strategies? Am J Transpl 2004; 4: 1289–1295
- Hariharan S, Johnson CP, Bresnahan BA, Taranto SE, McIntosh MJ, Stablein D. Improved graft survival after renal transplantation in the United States, 1988 to 1996. N Engl J Med 2000; 342: 605–612
[Abstract/Free Full Text] - Pelletier RP, Akin B, Henry ML et al. The impact of mycophenolate mofetil dosing patterns on clinical outcome after renal transplantation. Clin Transpl 2003; 17: 200–205
- Knoll GA, MacDonald I, Khan A, Van Walraven C. Mycophenolate mofetil dose reduction and the risk of acute rejection after renal transplantation. J Am Soc Nephrol 2003; 14: 2381–2386
[Abstract/Free Full Text] - Vincenti F. Immunosuppression minimization: current and fture trends in transplant immunosuppression. J Am Soc Nephrol 2003; 14: 1940–1948
[Free Full Text] - Margreiter R. The European Tacrolimus vs. Ciclosporin Microemulsion Renal Transplant Study Group. Efficacy and safety of tacrolimus compared with ciclosporin microemulsion in renal transplantation: a randomized multicentre study. Lancet 2002; 359: 741–746[CrossRef][ISI][Medline]
Accepted in revised form: 2. 3.06
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  E. H. Cole, O. Johnston, C. L. Rose, and J. S. Gill Impact of Acute Rejection and New-Onset Diabetes on Long-Term Transplant Graft and Patient Survival Clin. J. Am. Soc. Nephrol., May 1, 2008; 3(3): 814 - 821. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||