A Randomized Trial to Assess the Impact of Early Steroid Withdrawal on Growth in Pediatric Renal Transplantation: The TWIST Study


* Corresponding author: Ryszard Grenda, r.grenda@czd.pl


Minimizing steroid exposure in pediatric renal transplant recipients can improve linear growth and reduce metabolic disorders. This randomized multicenter study investigated the impact of early steroid withdrawal on mean change in height standard deviation score (SDS) and the safety and efficacy of two immunosuppressive regimens during the first 6 months after transplantation. Children received tacrolimus, MMF, two doses of daclizumab and steroids until day 4 (TAC/MMF/DAC, n=98) or tacrolimus, MMF and standard-dose steroids (TAC/MMF/STR, n=98). Mean change in height SDS was 0.16 ± 0.32 with TAC/MMF/DAC and 0.03 ± 0.32 with TAC/MMF/STR. The mean treatment group difference was 0.13 (p < 0.005 [95% CI 0.04–0.22]), 0.21 in prepubertal (p = 0.009 [95% CI 0.05–0.36]) and 0.05 in pubertal children (p = ns). Frequency of biopsy-proven acute rejection was 10.2%, TAC/MMF/DAC, and 7.1%, TAC/MMF/STR. Patient and graft survival and renal function were similar. Significantly greater reductions in total cholesterol and triglycerides but significantly higher incidences of infection and anemia were found with TAC/MMF/DAC (p < 0.05 all comparisons). Early steroid withdrawal significantly aided growth at 6 months more so in prepubertal than pubertal children. This was accompanied by significantly better lipid and glucose metabolism profiles without increases in graft rejection or loss.


At the time of transplantation, the average pediatric patient with chronic kidney disease is nearly two standard deviations below the appropriate age- and sex-adjusted height or is shorter than the third percentile of their peers, a deficit which is greater for males than females (1). While younger children experience better catch-up of growth deficits than older ones, about 30% of recipients of a renal transplant during their youth never attain normal adult height (2).

Linear growth after transplantation is affected by the interrelated effects of various pre- as well as posttransplant factors. The administration of steroids is known to exert growth-suppressing effects (3) and contributes to growth retardation in pediatric transplant recipients after transplantation (4–7). Previous uncontrolled studies have suggested that newer immunosuppressive agents, such as IL2R antagonists, allow minimization of steroids (5,8,9) and these agents have demonstrated some benefit in pediatric patients without compromising safety or efficacy (10–14). Nevertheless, clinicians seem to have taken a conservative approach to eliminating corticosteroids; maintenance steroid use in pediatric transplantation was estimated at 94% in 2004, decreasing only slightly to 89% 2 years later (15).

Several uncontrolled single-center studies (5,9,16–20) and one small randomized clinical trial (21) evaluated the effects of decreasing corticosteroid exposure in children. Results are encouraging and provide evidence that steroid avoidance or elimination in pediatric transplantation is possible while maintaining equivalent or superior outcomes to standard steroid-containing protocols.

This large, randomized, open-label, comparative multicenter trial was conducted to evaluate the impact of early steroid withdrawal on growth expressed as change in height standard deviation score (SDS) from baseline to the end of 6 months after transplantation in pediatric renal allograft recipients as well as the impact of early steroid withdrawal on other steroid-related metabolic complications. The experimental regimen consisted of two doses of IL2R antagonist induction (daclizumab) administered in combination with tacrolimus and mycophenolate mofetil (MMF) with steroids discontinued at day 4. Results from noncontrolled pediatric trials (8,22) and a controlled adult trial (23) demonstrated efficacy using daclizumab to facilitate steroid withdrawal. The combination of tacrolimus, MMF and steroids, the reference treatment, has been widely and effectively used and evaluated in previous pediatric trials (4,5) and retrospective analysis (12). Safety and efficacy outcomes of both regimens were evaluated.



Review and approval of the study protocol was obtained from the institutional review board at each center, and the study was conducted according to the Declaration of Helsinki. Written informed consent was obtained from parents or legal guardians and assent from the older patients if appropriate.

Eligible for enrollment were patients aged 2–18 years with end-stage renal disease who were suitable candidates for primary renal transplantation or retransplantation, receiving a kidney from a living or deceased donor with compatible ABO blood type. For eligibility, the skeletal age for boys was ≤17 and for girls ≤15 years. Excluded from the study were patients with a panel reactive antibody (PRA) grade ≥50% within the past 6 months and/or retransplantation due to organ rejection.


Tacrolimus was administered to both groups at an initial daily dose of 0.3 mg/kg given in two divided doses postoperatively. Subsequent doses were adjusted based on clinical evidence of efficacy and occurrence of adverse events (AEs) and guided by the following recommended trough-level ranges: 10–20 ng/mL for days 0–21 and 5–15 ng/mL for days 22–183. These tacrolimus exposure levels were established in a previous study (10).

MMF was administered preoperatively to both groups at an initial dose of 600 mg/m2. The daily dose was 1200 mg/m2 given in two doses for the first 2 weeks. Thereafter, the daily dose was 600 mg/m2 given in two doses (adjusted if medically indicated).

Patients in the experimental group (TAC/MMF/DAC) received the first dose of daclizumab 1.0 mg/kg within ±24 hours of reperfusion and the second and final dose of 1.0 mg/kg on day 14.

Both the dose and the duration of corticosteroid administration differed between the groups. The TAC/MMF/DAC group received corticosteroids as a bolus dose of 300–600 mg/m2 i.v. on day 0. The dose on day 1 was 60 mg/m2 reduced to 40 mg/m2 on day 2, 30 mg/m2 on day 3, and 20 mg/m2 on day 4 and then discontinued on day 5. The corticosteroid bolus dose and the day 1 dose were the same for the reference group (TAC/MMF/STR). Corticosteroid dose was reduced to 40 mg/m2 starting on days 2–7, and then reduced in decrements of 10 mg/m2 per week to <10 mg/m2 on days 43–183 at the discretion of the investigator.

Episodes of rejection were confirmed by biopsy and graded using the Banff 97 (24) classification. First-line therapy for acute rejection was corticosteroids according to local practice. Monoclonal or polyclonal antibodies could be administered for severe vascular rejection (Banff IIb or III).

Growth hormone treatment was prohibited during the study.

Outcome measures

Growth expressed as change in height SDS from baseline to end of 6 months was the primary endpoint of the study. The change in height SDS was calculated according to the following formula:


where SDS = (height measured – height standard population)/SD standard population. The values for height and standard deviation (SD) of height in the standard population were given per the age group and sex according to Reinken (25). On the consensus of the senior investigators, it was agreed that a period of 6 months would be sufficient to assess a change in height. Positive growth trends at 6 months were reported in the preliminary results of a single-center phase 1 trial (4) with that study group reporting improved growth by 6 months in patients <5 and 5–15 years compared to historical controls (5).

Height was measured in the children while standing against the vertical plane of a stadiometer with feet flat on the floor without shoes and socks. The child's heels, buttocks, shoulders and back of the head touched the vertical plane, and the eye–ear plane was perpendicular to the vertical plane. Three height measurements were taken by the same investigator at each of the eight protocol-defined assessment visits (days 0, 1, 7, 14, 28, 60, 91 and 183), and the average of these three measurements was used. If not already available, a wall-mounted Harpenden stadiometer was provided by the study sponsor to participating centers.

The following outcomes were also evaluated: incidence, overall frequency and time to first biopsy-proven acute rejection (BPAR); severity of BPARs; patient and graft survival; delayed graft function (DGF); renal function; absolute change in serum lipids (total cholesterol, high-density lipoprotein, low-density lipoprotein, triglycerides); incidence of AEs; incidence of new onset diabetes after transplant (NODAT); incidence of hypertension.

A renal biopsy was performed if indicated by clinical signs or symptoms or an increase of 10% in serum creatinine above posttransplant baseline. Graft loss was defined as death, graft nephrectomy or dialysis ongoing at study end or study withdrawal. DGF was defined as dialysis for more than 1 day during the first week (days 0–7) after transplantation. Renal function was evaluated using the Schwartz formula (26) to calculate glomerular filtration rate (GFR). NODAT was defined as at least 2 fasting blood glucose levels (FPG) of >7 mmol/L or treatment with antidiabetic agents for >30 consecutive days (WHO Consultation Report WHO/NCD/NCS/99 2). Hypertension was defined as the need of treatment with antihypertensive medicines. Major protocol violations were defined as follows: steroid administration >14 days after day 7 in TAC/MMF/DAC group, skeletal age larger than permitted, administration of recombinant human growth hormone.

Study design

Based on a two-sided t-test for independent samples and assuming an SD of 0.65, 75 patients per treatment group were needed to detect a 0.3 change in height SDS from baseline with a power of at least 80%, significance level α= 5%. A postulated difference in mean change in height SDS of 0.3 was established by consensus of the senior investigators. Under the additional assumption that 25% of patients would not provide evaluable data, the sample size was determined to be 100 patients per treatment group.

Analyses presented here are from the intention-to-treat (ITT) population and include all randomized and transplanted patients who received at least one dose of study medication (tacrolimus, MMF, daclizumab or corticosteroids).

Allocation to treatment groups was performed using sealed sequentially numbered randomization envelopes provided by the study sponsor. RANCODE (version 3.6) was used to generate the randomization sequence. Randomization was 1:1 with stratification by center and pubertal status, which was based on the categorization by Tanner (stage 1 defined prepubertal and stage 2–5 pubertal status) (27).

Statistical methods

The primary endpoint was analyzed by ANCOVA with the factors treatment group, pubertal status and the covariate baseline height SDS. ANCOVA analysis with factor treatment group and covariate baseline height SDS was performed separately on the subgroups of prepubertal and pubertal patients. A missing height measurement at the last study visit was replaced by the last available height, and a missing baseline height was replaced by the next available height. Missing laboratory values were not imputed.

Incidence of and time to first BPAR and patient and graft survival were analyzed by Kaplan–Meier cumulative survival probability estimates, differences between survival estimates were assessed using the Wilcoxon–Gehan test (two-sided). Adverse events were coded using version 8.0 of the Medical Dictionary for Regulatory Activities (MedDRA), and the overall incidence was compared pairwise using chi-square test or Fisher's exact test. Investigators verified data prior to full data analysis.


Between September 2005 and February 2008, 200 patients from 30 international centers across 13 countries were randomized to the study. Randomization resulted in an equal number of patients in each treatment group (Figure 1). The study was completed by 169 patients (86.2% of the ITT), with similar patient disposition between groups. Fewer patients in the TAC/MMF/DAC group (66/98, 67.3%) than in the TAC/MMF/STR group (81/98, 82.7%) completed the study per protocol. Steroid administration for >14 days (administered for treatment of rejection) after day 7 accounted for the majority of protocol violations in the TAC/MMF/DAC group. One patient in that group was withdrawn due to administration of growth hormone during the study.

Figure 1.

Flow of patients through the study. A similar percentage of patients in each group completed the study. The most common reason for premature study withdrawal was an adverse event. ITT, intention to treat. Randomized and transplanted patients who received at least one dose of tacrolimus, MMF, daclizumab or steroids. 1MMF severely modified or stopped for >21 cumulative days or more than 2 episodes of any length during the scheduled medication period unless medically necessary.

Patient demographics, viral status at baseline and HLA organ match were largely comparable with one demographic parameter (race) presenting a notable exception; there were more Caucasians in the TAC/MMF/STR group (Table 1). This difference was not considered to bias study results.

Table 1.  Baseline characteristics of study population
 TAC/MMF/DAC (n = 98)TAC/MMF/STR (n = 98)
Male, n (%)66 (67.3)59 (60.2)
Age, n (%), years
 2–516 (16.3)11 (11.2)
 6–1127 (27.6)35 (35.7)
 ≥1255 (56.1)52 (53.1)
Age, mean (SD)10.8 (± 4.2)11.3 (± 4.1)
Pubertal status, n (%)
 Prepubertal48 (49.0)52 (53.1)
 Pubertal50 (51.0)46 (46.9)
Height (cm)
 Mean (SD)134.4 (± 24.9)136.8 (± 23.9)
 Median (range)139.1 (77–177)140.0 (84–185)
 Prepubertal, mean (SD)114.8 (± 18.8)119.2 (± 17.0)
 Pubertal, mean (SD)153.6 (± 11.6)156.7 (± 11.7)
Skeletal age, n/mean (SD), years
 All patients, male63/10.6 (± 4.3)54/10.3 (± 4.3)
 All patients, female32/9.4 (± 4.5)36/10.3 (± 3.5)
 Prepubertal, male29/7.1 (± 3.5)28/6.9 (± 2.9)
 Prepubertal, female17/5.6 (± 2.7)18/7.8 (± 3.1)
 Pubertal, male34/13.5 (± 2.3)26/13.9 (± 1.9)
 Pubertal, female15/13.6 (± 1.2)18/12.7 (± 1.5)
PRA (%)
 Not recorded 1.0 1.0
Previous transplant (%)
 1 5.1
Race (%)
 Black 6.1 4.1
 Asian 9.8 4.1
 All other 3.1
EBV positive (%)56.356.7
HBV positive (%) 3.1 2.2
HCV positive (%) 1.0
Cytomegalovirus status
 Recipient-negative/donor-positive (%)32.726.3
ABO identical (%)84.793.9
 Mean total HLA mismatch 2.5 2.8
Donor type (%)

Prior to the study, recombinant human growth hormone (rhGH) was being taken by 24.5% of patients (24 of 98) in the TAC/MMF/DAC group and by a similar percentage, 25.5% (25 of 98), in the TAC/MMF/STR group.

The mean tacrolimus daily dose was similar for both treatment groups during and at the end of the study (Table 2). The mean tacrolimus trough levels at all assessment time points were well within the targeted range of exposure, and the mean trough levels were similar at study end.

Table 2.  Immunosuppressive medications and doses at the end of 6 months
 TAC/MMF/DAC (n = 98)TAC/MMF/STR (n = 98)
  1. 1Includes patients completing the study and patients withdrawn from the study.

  2. 2Includes all corticosteroids administered for routine maintenance immunosuppression, rejection treatment and perioperative bolus.

  3. 3p < 0.001, Wilcoxon test comparing cumulative doses up to the end of month 6.

Patients receiving randomized regimen,1 n (%)
   Tacrolimus + MMF75 (76.5)
   Tacrolimus + MMF + steroids10 (10.2)85 (86.7)
   All other combinations13 (13.3)13 (13.3)
 Mean (SD) daily dose (mg/kg), all patients0.18 (0.15)0.15 (0.09)
     Prepubertal0.24 (0.18)0.18 (0.1)
     Pubertal0.12 (0.007)0.13 (0.07)
 Mean (SD) trough level (ng/mL), all patients6.85 (2.76)7.15 (2.96)
     Prepubertal6.72 (3.05)7.25 (3.57)
     Pubertal6.99 (2.46)7.05 (2.21)
 Mean (SD) daily dose (mg/m2), all patients580 (19)570 (17)
     Prepubertal580 (19)540 (17)
     Pubertal570 (18)600 (18)
 Mean (SD) cumulative dose (mg/m2), all patients1149 (976)32658 (1405)
     Prepubertal1316 (1199)32831 (1768)
     Pubertal927 (513)32492 (951)

The protocol-defined fixed dosing of MMF was followed, and mean daily doses at study end were slightly below targeted doses and similar in the two groups (Table 2).

In the TAC/MMF/DAC group, adherence to protocol-specified corticosteroid discontinuation at day 4 was good with 88 of 98 patients steroid-free at week 2. At 6 months, maintenance immunosuppressive treatment with corticosteroids was required by 12 patients; mean daily dose for these patients was 5.6 ± 3.2 mg/m2. The most common reason for the reintroduction of corticosteroids was for treatment of rejection.

In the TAC/MMF/STR group, adherence to protocol-specified corticosteroid tapering was excellent (Table 2). All patients completing the study were on maintenance corticosteroids. The mean daily dose at study end was 4.9 ± 3.2 mg/m2.

The mean cumulative dose of corticosteroids was significantly lower in the TAC/MMF/DAC group than the dose in the TAC/MMF/STR for all patients as well as for the subgroups of patients stratified by pubertal status (p < 0.001, Wilcoxon test) (Table 2).

Primary outcome

Analysis of the primary endpoint showed a statistically significant superiority in favor of the treatment administered in the TAC/MMF/DAC group. The mean (SD) change in height SDS (adjusted for pubertal status and baseline value) for this group was 0.16 ± 0.32 compared with 0.03 ± 0.32 in the TAC/MMF/STR group (Table 3). The treatment group difference in adjusted mean change, 0.13, was significant (p < 0.005 [95% CI 0.04–0.22]).

Table 3.  Primary endpoint: mean change in height SDS
All patients, n9898
Change in height SDS adjusted for pubertal status and baseline value, mean (SD)0.16 (0.32)0.03 (0.32)
Treatment group difference in adjusted mean change0.13, p < 0.005 [95% CI 0.04–0.22]
Prepubertal patients, n4749
Change in height SDS adjusted for baseline value, mean (SD)0.34 (0.38)0.13 (0.38)
Treatment group difference in adjusted mean change0.21, p = 0.009 [95% CI 0.05–0.36]
Pubertal patients, n5046
Change in height SDS adjusted for baseline value, mean (SD)−0.01 (0.24)−0.06 (0.24)
Treatment group difference in adjusted mean change0.05, p = 0.33 [95% CI −0.05–0.14]

The results of a repeat analysis performed for the subgroups of prepubertal and pubertal patients reflected the overall analysis; however, greater growth was found in prepubertal patients (Table 3). In prepubertal children, significant differences in adjusted mean change in growth between the treatment groups favored the TAC/MMF/DAC regimen. In pubertal children, differences in the adjusted mean change in growth between the treatment groups were not significant.

Of clinical interest was the greater absolute linear growth recorded in patients in the TAC/MMF/DAC group, especially in prepubertal patients (Figure 2).

Figure 2.

Absolute change in mean height at 6 months. Absolute change in height (centimeter) was greater in patients treated with TAC/MMF/DAC than in those treated with TAC/MMF/STR and greater in prepubertal than in pubertal children in both treatment groups (p = ns for comparisons between groups).

Sensitivity analyses were performed on variables which might have affected the primary endpoint results. Neither the type of organ donation nor size of transplant center affected results. Skeletal age, however, had a significant impact on the change in height SDS. The difference between the groups in mean change in growth adjusted for skeletal age at baseline was 0.14 (p = 0.002 [95% CI 0.05–0.22]). Approximately, one-quarter of patients in each group took rhGH prior to study enrollment. To exclude the influence of hormone treatment, an analysis of the primary endpoint was performed on patients who had not taken rhGH. These results demonstrated statistically significant superiority of the TAC/MMF/DAC regimen with a change in mean height SDS of 0.18 ± 0.41 in the TAC/MMF/DAC group (n = 70) compared with a change of 0.06 ± 0.32 in the TAC/MMF/STR group (n = 67); the 0.12 difference in growth between the groups was significant (p = 0.04 [95% CI 0.004–0.232]).

Secondary outcomes

Biopsy-proven acute rejection:  The frequency of BPAR was low in both groups, and the severity of rejection was similar across groups (Table 4). No rejections in either group showed spontaneous resolution, and more rejections in the TAC/MMF/DAC than in the TAC/MMF/STR group were sensitive to treatment with corticosteroids.

Table 4.  Secondary efficacy data
 TAC/MMF/DAC (n = 98)TAC/MMF/STR (n = 98)
  1. 1Two patients in the TAC/MMF/STR group underwent two biopsies each: one patient had one mild (Banff I) and one moderate (Banff II) rejection, both sensitive to corticosteroids; and another patient had one mild (Banff I) and one moderate (Banff II) rejection of which one was sensitive and one resistant to steroid treatment.

Biopsy-proven acute rejection, n (%)10 (10.2)7 (7.1)
Biopsy severity; patients, n (%)/biopsies, n
   Mild (Banff I)7 (7.1)/105 (5.1)/61
   Moderate (Banff II)2 (2.0)/23 (3.1)/31
   Severe (Banff III)1 (1.0)/11 (1.0)/1
Outcome of rejection treatment; patients, n (%)
 Corticosteroid-resistant/resolved with other treatment1 (1.0)3 (3.1)
 Sensitive to corticosteroids9 (9.2)6 (6.1)
Patient survival (Kaplan–Meier estimate) (%)99.0100.0
Graft survival (Kaplan–Meier estimate) (%)96.996.9
Delayed graft function, n (%)9 (9.2)11 (11.2)

Estimated rates of freedom from BPAR were 89.2% in the TAC/MMF/DAC group and 92.3% in the TAC/MMF/STR group (Kaplan-Meier method) (Figure 3).

Figure 3.

Estimated rate of patients free from biopsy-proven acute rejection. The estimated rate of patients free from BPAR was 89.2% in the TAC/MMF/DAC group and 92.3% in the TAC/MMF/STR group (Kaplan–Meier method). The difference between the groups at the end of 6 months was −0.02 (p = 0.46, Wilcoxon–Gehan test [95% CI −0.107–0.063]).

Patient and graft survival:  Kaplan–Meier estimates of patient and graft survival were excellent for both groups (Table 4). Graft failure occurred in three patients in each group including one patient in the TAC/MMF/DAC group who died with a functioning graft of sepsis 1 month after study withdrawal. The incidence of never functioning graft was 2% (two patients) in each treatment group.

Delayed graft function:  The incidence of DGF was comparable (Table 4). The mean duration of dialysis in the presence of DGF was similar at 9.9 ± 5.6 days versus 7.4 ± 7.3 days in the TAC/MMF/DAC and TAC/MMF/STR groups, respectively.

Renal function:  Calculated mean GFR was similar in the two groups throughout the study and at study end. At study completion, mean GFR was 97.1 ± 32.3 mL/min per 1.73 m2 with TAC/MMF/DAC and 98.7 ± 27.2 mL/min per 1.73 m2 with TAC/MMF/STR. Mean serum creatinine was also similar at 86.8 ± 40.9 μmol/L with TAC/MMF/DAC and 81.8 ± 30.4 μmol/L with TAC/MMF/STR.

Serum lipids:  Mean serum lipid values decreased over time in the TAC/MMF/DAC group. The mean total cholesterol value at baseline was 4.8 ± 1.6 mmol/L and decreased by 0.9 ± 1.7 mmol/L at study end. In the TAC/MMF/STR group, mean baseline total cholesterol was 4.8 ± 1.4 mmol/L and decreased by 0.4 ± 1.2 mmol/L (p = 0.001 for difference in change over time). The mean baseline triglyceride was 2.3 ± 1.8 mmol/L in the TAC/MMF/DAC group and decreased by 1.1 ± 1.8 mmol/L at study end. In comparison, mean baseline triglyceride in the TAC/MMF/STR group was 2.0 ± 1.0 mmol/L and decreased by 0.4 ± 1.2 mmol/L (p = 0.02 for difference in change over time; F-test/ANCOVA for all comparisons). The use of lipid-lowering medications was similar at 3.1% in the TAC/MMF/DAC and 2.0% in the TAC/MMF/STR group.


The overall incidence of AEs in the TAC/MMF/DAC and TAC/MMF/STR groups was similar (Table 5) as was the incidence of causally related AEs (77.6 and 80.6%), serious AEs (64.3 and 61.2%), and causally related serious AEs (43.9 and 37.8%), respectively. Study discontinuation due to an AE was reported in 8.2% of patients (8 of 98) in the TAC/MMF/DAC group and in 7.1% of patients (7 of 98) in the TAC/MMF/STR group. The types of AEs leading to discontinuation were similar in the two groups and no single AE accounted for the majority of patient discontinuations.

Table 5.  Overall incidence of adverse events1 regardless of relationship to study medication
System organ class2 preferred termTAC/MMF/DAC (N = 98)TAC/MMF/STR (N = 98)
  1. 1Presented are adverse events with an occurrence of ≥10% of patients in either treatment group or with a significant difference between groups or of clinical importance. Excluded are adverse events related to the transplanted organ, abnormal laboratory values and complications of surgery.

  2. 2Coded using MedDRA version 8.0.

  3. 3p < 0.05, Fisher's exact test comparing the number of patients.

Any adverse event, n (%)95 (96.9)93 (94.9)
Infections62 (63.3)57 (58.2)
 Bacterial, urinary tract17 (17.3)13 (13.3)
 Bacterial, respiratory tract36 (6.1)
 Cytomegalovirus12 (12.2)12 (12.2)
 Febrile infection37 (7.1)
Gastrointestinal disorders39 (39.8)41 (41.8)
 Diarrhea24 (24.5)21 (21.4)
 Vomiting14 (14.3)8 (8.2)
Blood/lymphatic system disorders30 (30.6)18 (18.4)
  Anemia322 (22.4)8 (8.2)
Metabolism disorders23 (23.5)38 (38.8)
 Hyperglycemia33 (3.1)12 (12.2)
Vascular disorders23 (23.5)33 (33.7)
 Hypertension21 (21.4)25 (25.5)
Respiratory/thoracic disorders15 (15.3)10 (10.2)
Nervous system disorders14 (14.3)16 (16.3)
Skin disorders14 (14.3)11 (11.2)
Musculoskeletal disorders11 (11.2)12 (12.2)
Posttransplant lymphoproliferative disease2 (2.0)

Other important AEs assessed during the study were two cases of posttransplant lymphoproliferative disorder (PTLD) reported in the TAC/MMF/DAC group.

Bacterial respiratory tract infection, febrile infection and blood system disorders (recorded as AEs) occurred with a significantly higher frequency in patients in the TAC/MMF/DAC group (Table 5). By contrast, hyperglycemia, reported as an AE, occurred significantly more frequently in the TAC/MMF/STR group (p < 0.05 for all comparisons, Fisher's exact test comparing group differences).

A notable difference between treatment groups in the incidence of NODAT was found. No patient in the TAC/MMF/DAC group required antidiabetic treatment for >30 days compared with 5 of 98 patients (5.4%) in the TAC/MMF/STR group. At least two FPG values >7 mmol/L were reported in no patient in the TAC/MMF/DAC group and in 3 of 98 patients (3.3%) in the TAC/MMF/STR group.

At study completion, ≥3 antihypertensive medications were being taken by 7 of 91 patients (7.7%) in the TAC/MMF/DAC group compared with 16 of 87 patients (18.3%) in the TAC/MMF/STR group.


The results show that early corticosteroid withdrawal aids growth, expressed as change in height SDS from baseline to end of 6 months, in pediatric renal allograft recipients. Importantly, favorable linear growth was achieved without increasing the risk of acute rejection or graft loss and early steroid withdrawal resulted in secondary advantages in safety in terms of a reduction in metabolic disorders.

We found that prepubertal children showed greater benefit from minimizing steroids as evidenced by a greater increase in height than their older counterparts. These results are corroborated by data from NAPRTCS, which showed better improvement in mean growth deficit as well as immediate increase in height even as early as 6 months after renal transplantation in younger patients (1). And linear growth was significantly greater at 6 months in children <5 years using a steroid-free regimen in a prospective study using a historical cohort (5). Even the discontinuation of steroids at 1 year after transplantation in children <5 years provided significant and sustained improvement in height at 3 years (28).

In addition to favorable growth development, other positive outcomes were found with (or despite) early steroid withdrawal. Of note was the low incidence of BPAR. In small retrospective and/or uncontrolled studies using different induction and maintenance immunosuppressive protocols, incidences of acute rejection ranged from 5 to 25% using steroid avoidance regimens (5,8,16,29) and 0 to 21% with early steroid withdrawal regimens (9,30,31). Higher rates of rejection were usually reported in the comparator groups of these trials. Other benefits of early steroid withdrawal over conventional steroid administration we observed were lower incidences of NODAT, improvement in total cholesterol and triglyceride levels, and the need for fewer antihypertensive medications.

Our results support the work previously conducted on the feasibility of facilitating minimization of steroids using short-term IL2-R induction and the benefits attained by early steroid withdrawal. We are cautious, however, in making recommendations to alter practice based on these study results for several reasons. First, the 6-month duration of this study is too short to allow us to draw conclusions on longer term outcomes including late-onset acute rejection, renal graft fibrosis or maintenance of a steroid-free immunosuppressive regimen and growth. A landmark double-blinded, randomized study in adults, however, showed no increased risk of graft loss or BPAR in the longer term with early steroid withdrawal (32). We plan an extended follow-up of patients.

Second, we were well aware that our European pediatric transplant population would be low to intermediate immunologic risk, and we therefore designed the study using two doses of 1 mg/kg daclizumab. For this reason, we cannot speculate on the efficacy of the regimen in higher risk patients (i.e. those with a PRA >50%). The rate of acute rejection at month 6 in this study was low at 10.2 and 7.1%, and a majority of rejections occurred during the first 3 months. In our opinion, these results justify using a shorter course (two doses) of daclizumab to achieve a balance between efficacy and sufficient immunosuppressive coverage.

Bone marrow suppression leading to leukopenia, infection and anemia (all recorded as AEs but not predefined study endpoints) occurred more commonly with the TAC/MMF/DAC regimen and are attributed to a lack of steroid-driven cellular proliferation. We observed two cases of PTLD which are within the expected range for this patient population. The PTLD incidence in our previous trial was 2% at 6 months (10), and an incidence of 1.8% was reported at 1 year in a European randomized trial using basiliximab and cyclosporine microemulsion with MMF and steroids (33).

Although a substantial increase in the use of MMF has been reported (34), the benefit of combining tacrolimus with MMF in pediatrics has not been well defined to date (35,36), and our study represents the first time that tacrolimus combined with fixed-dose MMF was tested in a prospective randomized, controlled trial in a pediatric population. Patient tolerability of MMF was excellent with 87% of patients in both groups maintained on protocol-defined doses of MMF. And calculated GFR was better in both groups than GFR in our previous study in which tacrolimus was combined with azathioprine (10).

In designing this study, we estimated that a change in height of 0.3 SDS from baseline might be attained following early steroid withdrawal. This estimate was in fact greater than the actual height changes measured in either treatment group. Nevertheless, a clinically significant difference between the two groups was discernable at the end of the study.

In conclusion, in our short-term study we found a positive effect on growth at 6 months following early steroid withdrawal in pediatric renal transplant patients, this observation being mainly attributable to prepubertal children. Importantly, early steroid withdrawal had no detrimental effect on 6-month acute rejection, patient and graft survival rates, and both lipid and glucose metabolism profiles were significantly better with this regimen.


All named authors participated in the design, implementation and interpretation of the study and contributed to and approved the final version of the manuscript. R.G. had access to the study data and had final responsibility for the decision to submit the manuscript for publication.


The following investigators also participated in the TWIST study: Claude Guyot, Mihai Lucan, Bernd Hoppe, Nadeem Moghal, Moin Saleem, Caroline Jones, Luca Dello Strologo, Styrbjörn Friman, Mignon McCulloch, Nathalie Godefroid, Lars Pape, Po-Huang Lee, Alberto Edefonti, Francesco Perfumo, Dierdra Hahn. We acknowledge writing and editorial assistance provided by C. Krcmar who received payment from Astellas Pharma Europe Ltd for this work. The study sponsor, Astellas Pharma Europe Ltd, supplied funding and monitored the progress and conduct of the study.

Conflict of Interest

R.G. has received study and travel grants from Astellas Pharma Europe; B.T. has received grant support and lecture fees from Astellas Pharma Europe, Novartis, Roche and Wyeth; K.V. and D.V.M. have received lecture fees and travel grants from Astellas Pharma Europe; M.B. is an employee of Astellas Pharma Europe Ltd; N.J.A.W. has received consulting, advisory board, lecture fees, and any combination of the three from Astellas Pharma Europe, Merck, Novartis, Roche and Wyeth and has received grant support from Astellas Pharma Europe and Novartis.