In an effort to reduce rejection, extend allograft survival and minimize complications, we hypothesized that robust immunosuppression during the first 6 months after transplantation would allow for the safe withdrawal of steroids. A total of 274 pediatric subjects were enrolled and received an anti-CD25 antibody, sirolimus, calcineurin inhibitor and steroids. At 6 months after transplantation, subjects were randomized to steroid withdrawal (n = 73) versus continued low-dose steroids (n = 59). This study was stopped prior to target enrollment because of the incidence of post-transplant lymphoproliferative disorder. At the time of study termination, 132 subjects had been randomized and were available for analysis. At 18 months after transplantation, there was no difference in the standardized height z score; however, the standardized height velocity was greater in the steroid withdrawal group compared to the control group (p = 0.033). There were no differences in acute rejection episodes between treatment groups. The 3-year allograft survival rate was 84.5% in the control group and 98.6% in the steroid withdrawal group (p = 0.002). The immunosuppressive protocol utilized in this study allowed for the withdrawal of steroids without an increased risk of rejection or allograft loss. However, the complications associated with the use of this immunosuppressive protocol were too high to recommend its routine use in pediatric patients.
Pediatric patients receiving renal allografts continue to have remarkably good short-term outcomes. However, long-term outcomes have been only minimally improved while long-term complications have increased. To avoid complications associated with immunosuppression, the recent focus of care has been on minimizing immunosuppressive medications while maintaining and maximizing long-term allograft survival. Although corticosteroids were important for the early success of transplantation, they contribute to significant complications including osteoporosis, avascular necrosis, fractures, hypertension, myopathy, cataracts, glaucoma, diabetes mellitus, hypercholesterolemia, Cushingoid appearance, acne, skin atrophy, weight gain and decreased linear growth velocity with subsequent short stature. For many years patients, parents and transplant physicians have sought to minimize or eliminate the use of corticosteroids after pediatric renal transplantation.
Early analysis of data from the North American Pediatric Renal Trials and Collaborative Studies (NAPRTCS) suggested that a transition from daily to alternate day steroids was associated with no increased risk of rejection and with improved growth (1). However, the most recent annual report from NAPRTCS shows that 92% of patients continue to receive long-term daily steroids as part of their immunosuppressive regimen (http://www.naprtcs.org). After the introduction of cyclosporine, many centers explored the possibility of steroid avoidance and/or steroid withdrawal. These trials demonstrated a significantly increased rate of rejection in adults with reported ranges from 15 to 81% (2). Of even more concern, some trials demonstrated increased rates of allograft failure in adult patients withdrawn from steroids (3). A meta-analysis of studies during this era reported an increased risk of rejection and decreased allograft survival when steroids were withdrawn (4). With the introduction of newer immunosuppressive agents such as tacrolimus, mycophenolate mofetil, sirolimus and newer antibodies, such as thymoglobulin, basiliximab and daclizumab, many centers have again attempted to avoid and/or withdraw steroids. However, meta-analysis of these more recent adult trials has also shown an increased risk of acute rejection (5).
Analysis of the NAPRTCS registry data demonstrated that children with no rejection during the first 6 months after transplantation were at much lower risk of subsequent rejection. In this study, we hypothesized that robust immunosuppression in children at the time of transplantation would prevent early acute rejection, decrease the risk of late rejection thereby allowing for steroid withdrawal. To test this hypothesis, we conducted a randomized, double-blind, placebo controlled study of steroid withdrawal at 6 months after renal transplantation in pediatric renal transplant recipients.
This study was a randomized double-blind, placebo-controlled trial of steroid withdrawal starting 6 months after renal transplantation in pediatric recipients. As the primary outcome variable, we compared change in standardized height z score between subjects withdrawn from steroids (SW) and those continued on low-dose maintenance steroids (control). The study was conducted at 17 pediatric renal transplant centers across the USA and Mexico (see Appendix). The participating centers provided a study sample that reflects the demography (age, gender and race) of children receiving a renal transplant in the USA. The study protocol was approved by the Institutional Review Board of each center, by the transplant data safety and monitoring board (DSMB) of the National Institute of Allergy and Infectious Disease (NIAID) and was conducted under an Investigational New Drug (IND) status of the US Food and Drug Administration (FDA). Under the terms of the study grant, an independent NIH contractor collected, analyzed and submitted all the data to the DSMB for review. Research subjects were accrued from January 2001 through August 2004, and the results through August 2006 are reported here.
Subjects between the ages of 0 and 20 years old receiving their first living related, living unrelated or deceased donor renal transplant were eligible for enrollment.
All subjects received induction therapy with the chimeric anti-CD25 monoclonal antibody basiliximab, administered in two doses: the first dose was given preoperatively and the second dose was administered on post-transplant day 4. Maintenance immunosuppression included sirolimus, cyclosporine or tacrolimus, and corticosteroids. Sirolimus was administered daily starting on postoperative day 1 at a dose of 6 mg/m2 and was adjusted to maintain a trough level of 10–20 ng/mL. Subjects receiving cyclosporine microemulsion formula had the dose adjusted to maintain a whole blood trough TDx monoclonal level of 175–400 ng/mL (or an equivalent HPLC level) for the first 2 weeks after transplant. The dose was subsequently tapered to maintain a trough level of 175–300 ng/mL from week 3 to month 3, and 50–250 ng/mL from month 3 through the end of the study at month 36. Subjects receiving tacrolimus had the dose adjusted to maintain a whole blood trough level between 10 and 15 ng/mL for the first 4 weeks after transplant. Trough levels were maintained between 5 and 10 ng/mL thereafter. Methylprednisolone was administered at 10 mg/kg intravenously perioperatively and on postoperative day 1. Prednisone was administered orally beginning on postoperative day 2 starting at 2 mg/kg/day and tapering to 0.15 mg/kg/day as a single morning dose by post-transplant day 74.
All subjects received pneumocystis jiroveci prophylaxis with trimethoprim/sulfamethoxazole for 6 months following transplant. All recipients of a kidney from a CMV positive donor received center-specific anti-cytomegalovirus therapy. There was no protocol-driven surveillance for CMV, EBV or BK viruses.
6-month biopsy and randomization
All enrolled subjects who did not have an episode of acute rejection or other event resulting in removal from the study in the first 6 months after transplantation underwent a protocol-driven biopsy at 6 months. Subjects with no clinical or histologic evidence of rejection were centrally randomized in a 1:1 fashion to either undergo complete steroid withdrawal or to be maintained on low-dose (0.15 mg/kg/day) daily steroids in a placebo controlled, double-blinded fashion. Steroid withdrawal was completed by the end of 12 months post-transplant.
Effective August 13, 2004, the study was permanently terminated due to an unanticipated high incidence of post-transplant lymphoproliferative disorder. Enrolled subjects subsequently received site-specific immunosuppression. A safety follow-up study was implemented per the recommendation of the US Food and Drug Administration. As part of the safety follow-up study, subjects were followed for evidence of Epstein-Barr-related disease, serious adverse events, adverse events, hospitalizations, infections, graft failure, deaths and rejection for a period of 2.5 years following randomization or for 2 years following study closure. These data have previously been analyzed and reported (6).
The comparability of the two treatment groups at the time of randomization was assessed in terms of demographic characteristics, age, gender and race and renal allograft characteristics including cold ischemia time, donor's serum creatinine, source of the allograft (living related donor, living unrelated donor, deceased donor), human leukocyte antigen matching, blood group system matching and delayed allograft function. Wilcoxon tests and t-tests were used to test for group differences in continuous variables. Fisher exact tests were used to test for group differences in categorical variables. Time to event analyses was based on the time of randomization until the first occurrence of the event being described. The median time to event and its associated 95% confidence interval were computed. The survival distribution of time to event data was estimated using the Kaplan–Meier product limit estimator and the median time to event and its associated 95% confidence interval were compared. Differences between treatment groups in time to event endpoints were assessed by the log-rank test. A descriptive summary of the number and percentage, the mean number and days from transplant until the first, second, third, or more of biopsy-proven rejection episodes were prepared. The distribution of severity of acute rejection (mild, moderate, severe) that occurred after randomization was tabulated for both treatment groups and the statistical significance of the differences was assessed by Cochran–Mantel–Haenzel methods. To estimate and test for treatment differences in standardized height and growth velocity from 6 to 18 months post-transplant, analysis of covariance models were used adjusting for the 6-month value.
After informed consent, 274 subjects were enrolled from 17 centers and 273 subjects received a renal transplant (Figure 1) (one enrolled subject died prior to transplantation). This study was stopped prior to target enrollment by the Medical Monitor and the Data Safety Monitoring Board (DSMB) because of an unanticipated high incidence of post-transplant lymphoproliferative disorder (PTLD).
During the first 6 months after transplantation, all transplanted subjects were treated with basiliximab, tacrolimus or cyclosporine, sirolimus and prednisone. At 6 months, subjects underwent a surveillance renal allograft biopsy and those with no clinical or histological evidence for rejection were randomized to receive either maintenance low-dose steroids (control group) or steroid withdrawal (SW group) over the following 6 months. Of the 273 subjects transplanted, 35 (12.8%) were ineligible for randomization because they had not yet reached 6 months post-transplantation at the time the study was terminated. Of the remaining subjects, 106 were not eligible for randomization for other reasons (Table 1). The recipient and donor characteristics for subjects enrolled but not randomized (n = 142), randomized to control group (n = 59) and randomized to steroid withdrawal (n = 73) are shown in Table 2. There were no demonstrable differences in age, gender, race/ethnicity, primary cause of kidney disease, donor source or HLA matching among these groups.
Table 1. Reasons for subject termination during the first 6 months after transplantation
Subjects enrolled but not randomized
(n = 142)
Died prior to transplantation
Did not reach 6 months post-transplantation at time of study termination
Experienced an event during the first 6 months after transplantation
Rejection and allograft failure
Failure to return/refused treatment/withdrew consent
Table 2. Summary of demographics and baseline characteristics (all enrolled populations)
Withdrawal (N = 73)
Maintenance (N = 59)
Randomized total (N = 132)
Not randomized (N = 142)
11. 76 (4.90)
Age categories (years), n (%)
Sex, n (%)
Race, n (%)
American Indian or Alaskan native
Ethnicity, n (%)
Hispanic or Latino origin
Non-Hispanic or non-Latino origin
Source of allograft, n (%)
Live donor/other related
During the first 6 months after transplantation, 40 subjects experienced at least one episode of acute rejection, resulting in a rejection rate of 14.7%. There were no differences in patient or donor characteristics in those subjects with or without rejection episodes (data not shown). Sixteen subjects (5.8%) were withdrawn from study secondary to adverse events. Ten subjects developed PTLD and were withdrawn from further study participation (reported in detail elsewhere (6)).
At 6 months after transplantation, 132 subjects were randomized, 59 to continue low-dose maintenance steroids (control group) and 73 to undergo steroid withdrawal (SW group) over the ensuing 6 months. The disparity in numbers randomized to the two treatment groups was due to an unfortunate combination of imbalances within partially completed randomization blocks at the time the study was stopped by the DSMB.
After randomization, there were 13 subjects with rejection episodes, 9 in the control group (15.3%) and 4 in the SW group (5.5%), but these differences were not significant (p = NS). Further, these rejection rates were not different from those seen prior to randomization. The time to first rejection episode is shown in Figure 2. Although there was a trend toward more episodes of rejection in the control group compared to the SW group, this did not reach statistical significance (p = 0.081). The total number of rejection events was too small for meaningful subgroup analysis. There were no differences in the rejection rate between treatment groups in African American and other race; young subjects (<5 years of age) and older subjects (>5 years of age); adolescent subjects and other age groups or with increasing number of HLA mismatches. All subjects enrolled in this study received calcineurin inhibitors (CNI) as part of their immunosuppressive regimen. However, the choice of CNI (cyclosporine versus tacrolimus) was center specific. There were no differences in the number of subjects receiving cyclosporine versus tacrolimus in each study group or the number of rejection episodes in subjects treated with cyclosporine (4) versus tacrolimus (6) in any treatment group (three subjects with rejection episodes received both calcineurin inhibitors).
Allograft function and survival
Renal allograft function as calculated using the Schwartz formula was determined at randomization and every 3 months during study. The mean creatinine clearance at 18 months post-transplantation (12 months post-randomization) was 88.9 ± 27.1 mL/min/1.73 m2 in the control group and 81.7 ± 26.2 mL/min/1.73 m2 in the SW group (p = NS). Further, outcomes were assessed in terms of allograft survival and the composite outcome of allograft failure and death. There were four allograft failures and five deaths in the control group and no allograft failures and one death in the SW group. At 3 years post-transplantation, the composite of patient and allograft survival rate was 84.5% in the control group and 98.6% in the SW group (p = 0.002; Figure 3). There were no differences in 3-year allograft survival rate between treatment groups in African Americans, young subjects (<5 years), adolescent subjects or those with increased HLA mismatches. The causes of death were infection (n = 5) and malignancy (n = 1) (Table 3).
Table 3. Cause of death
Cause of death
Benefits of steroid withdrawal—growth, BP, lipids
Although subjects in the control group of this double-blind, placebo controlled design were on low-dose corticosteroids (0.15 mg/kg), there was a statistically significant difference between groups in the mean Cushingoid Features Assessment Score which has a range of 0–10 where 0 reflects no features, −3.0 ± 2.3 in the control group versus 1.6 ± 1.3 in the withdrawal group (p = 0.031).
The primary end-point of this study was changed in standardized height z score. Height was measured at randomization and every 3 months during the study (Figure 4). At 1 year after randomization, the mean height z score was −1.51 ± 0.95 for the control group and −1.27 ± 1.16 for the SW group (p = NS). However, standardized height growth velocity was significantly higher in the SW (0.16 ± 0.507 z score/year) as compared to the control group (−0.04 ± 0.372 z score/year) (p = 0.033 unadjusted and p < 0.001 adjusted for baseline effect).
Lipid levels including total cholesterol, high-density lipoprotein, low-density lipoprotein and triglyceride were measured every 3 months, and there were no statistically significant differences between treatment groups. Further, there was no difference in the use of lipid-lowering agents between the treatment groups. Finally, systolic and diastolic blood pressures were determined every 3 months after randomization and there were no statistically significant differences between treatment groups.
In this study, we hypothesized that robust immunosuppression at the time of transplantation would prevent early acute rejection, decrease the risk of late rejection and allow for steroid withdrawal. The primary outcome variable measured was change in linear growth in those subjects withdrawn from steroids versus those continued on low-dose steroids. At 1 year after steroid withdrawal, there was no statistical difference in age-adjusted height (z score) of the subjects withdrawn as compared to those maintained on steroids. However, there was a significant improvement in height velocity in those subjects withdrawn from steroids. The study design involved slow steroid withdrawal between 6 and 12 months after transplantation. At the 18-month time point, patients would have been completely off steroids for only 6 months. It is possible that with longer follow-up, a difference in height z score would have become evident between treatment groups. However, it is also possible that the effects of very small doses of steroids (0.15 mg/kg) do not contribute significantly to the growth failure seen in these children. Other factors have been shown to contribute to growth failure in this patient population including chronic allograft dysfunction, acid–base disorders, abnormalities in the growth hormone/insulin-like growth factor axis, and other immunosuppressive medications (7).
The data presented here support the hypothesis that robust immunosuppression at the time of transplantation prevents early acute rejection, decreases the risk of late rejection and allows for steroid withdrawal. Subjects treated with basiliximab, corticosteroid, calcineurin inhibitor and sirolimus had a rejection rate of 14.7% at 6 months after transplantation. This represents one of the lowest rates of rejection reported in either single-center or multi-centered trials. When these subjects were slowly withdrawn from steroids, they experienced no increased incidence of rejection or allograft loss. However, the complications seen with the robust immunosuppressive protocol used in this trial, especially the rate of PTLD, exceeded those in patients treated with other regimens. This was particularly noted in the youngest subjects and those naïve to EBV receiving an EBV positive allograft. These data have been explored more deeply and reported elsewhere (6). Because of this, we cannot recommend the use of this protocol as a means to withdraw steroids in pediatric patients after renal transplantation.
Recent uncontrolled trials have suggested that steroid avoidance or withdrawal is safe after transplantation, demonstrating no increased risk of rejection or allograft survival in adult (8,9) and pediatric (10–13) patients. Complete steroid avoidance has also been reported in uncontrolled studies that have suggested no increased risk of rejection or allograft failure (14). Several recent randomized controlled trials in adults have reported steroid withdrawal with no increased rejection and no change in short-term allograft survival (15–17). Our results demonstrated a decreased rate of rejection and improved allograft and patient survival in patients withdrawn from steroids. This study, especially at 3 years of follow-up, involves a small number of patients and the results should be interpreted with caution. However, similar results of improved allograft outcomes after steroid withdrawal have been reported in adult studies (18). It is possible that the use of steroids either directly or indirectly have an adverse effect on outcomes after transplantation in this group of patients with aggressive immunosuppression and no rejection during the first 6 months after transplantation. Of interest, the rejection rate after randomization was high, similar to that seen during the first 6 months after transplantation, and in the control group continued throughout the entire 3 years of analysis. The numbers were quite small and sub-group analysis did not identify particular groups at higher risk. However, it is possible that continued use of steroids and the associated Cushingoid features contributed to decreased compliance with the control subjects, increased rejection and decreased allograft survival.
We examined other possible beneficial effects of steroid withdrawal in our subjects. Many clinicians and investigators have suggested that low-dose steroids are not associated with significant cosmetic or clinically relevant side effects especially when used in an every-other-day regimen. In this double-blind, placebo controlled study, there was a significant difference in the Cushingoid facies in those patients who continued on daily steroids as compared to those who withdrawn. This is particularly relevant since Cushingoid facies may be a visible sign of other steroid toxicities and, in addition, may contribute to non-adherence with the immunosuppressive regimen. There were no statistically significant differences in the blood pressure or lipid profiles in these subjects at 1 year after steroid withdrawal.
This study represents the only controlled trial of steroid withdrawal in pediatric transplant recipients. We demonstrated that robust immunosuppression at the time of transplantation can allow steroid withdrawal at 6 months post-transplantation without an increased risk of rejection or allograft failure. However, this protocol of basiliximab, sirolimus, calcineurin inhibitor (cyclosporine or tacrolimus) and steroids was too toxic to be recommended for routine use. Further study is needed to find protocols that can safely be used to allow steroid withdrawal or avoidance in pediatric patients.
Results of a study conducted by the Cooperative Clinical Trials in Pediatric Transplantation (CCTPT) of the National Institute of Allergy and Infectious Diseases, NIH. Support provided by NIH U01-A1-46135 and Wyeth Pharmaceuticals, Inc.