Delayed Introduction of Reduced-Dose Tacrolimus, and Renal Function in Liver Transplantation: The ‘ReSpECT' Study


* Corresponding author: James M. Neuberger,


We report a multicenter, prospective, randomized, open-label trial investigating the effect of lower levels and delayed introduction of tacrolimus on renal function in liver transplant recipients. Adult patients with good renal function undergoing primary liver transplant were randomized to either: group A (standard-dose tacrolimus [target trough levels >10 ng/mL] and corticosteroids; n = 183); group B (mycophenolate mofetil [MMF] 2g/day, reduced-dose tacrolimus [target trough levels ≤8 ng/mL], and corticosteroids; n = 170); group C (daclizumab induction, MMF, reduced-dose tacrolimus delayed until the fifth day posttransplant and corticosteroids, n = 172). The primary endpoint was change from baseline in estimated glomerular filtration rate (eGFR) at 52 weeks. The eGFR decreased by 23.61, 21.22 and 13.63 mL/min in groups A, B and C, respectively (A vs C, p = 0.012; A vs B, p = 0.199). Renal dialysis was required less frequently in group C versus group A (4.2% vs. 9.9%; p = 0.037). Biopsy-proven acute rejection rates were 27.6%, 29.2% and 19.0%, respectively. Patient and graft survival was similar. In conclusion, daclizumab induction, MMF, corticosteroids and delayed reduced-dose tacrolimus was associated with less nephrotoxicity than therapy with standard-dose tacrolimus and corticosteroids without compromising efficacy or tolerability.


Although outcomes after liver transplantation are usually excellent, with current 5- and 10-year patient survival rates exceeding 70% and 60% (1,2), late complications remain of concern. Renal dysfunction remains a major cause of both morbidity and mortality (3). Significant renal impairment, occurring in up to 27% of liver allograft recipients at 5 years (4,5), results in end-stage renal disease in as many as 10% of patients at 10 years posttransplant (4) and late chronic renal dysfunction or renal failure is associated with the risk of premature death (3–6).

Late renal failure is associated with both pre- and postliver transplant factors, including higher concentrations of CNIs both early and late posttransplant (7,8) and can be predicted by creatinine levels in the first posttransplant year (9,10). Strategies to minimize the adverse renal effects of CNIs in patients include reducing the CNI dose (11–14) or complete withdrawal of the CNI (14,15), while adding other immunosuppressive agents such as mycophenolate mofetil (MMF) (12–14,16) or sirolimus (11,15). Alternatively, a preemptive strategy may be adopted by attempting to avoid CNI-induced renal impairment. Antibody induction therapy (such as interleukin-2 [IL2] receptor antagonists) may allow reduction in dose or delayed introduction of CNIs (17–19).

We conducted the ReSpECT study, a prospective, randomized trial in de novo adult liver transplant patients with good renal function pretransplant, to assess the effect on renal function, acute rejection, and graft and patient survival of three regimens: standard-dose tacrolimus and corticosteroids; MMF with reduced-dose tacrolimus and corticosteroids; and induction with daclizumab, MMF and delayed introduction of reduced-dose tacrolimus and corticosteroids.


Study design and patient population

This was a prospective, randomized, open-label, parallel group study. Inclusion criteria included age above 16 years, first orthotopic (whole or split) liver transplant, and expected survival and graft survival >7 days posttransplantation. The main exclusion criteria were: serum creatinine concentration above 200 μmol/L on the day of the transplant; need for renal replacement therapy within 30 days before transplantation; ABO incompatibility; HIV positivity or a previous history of malignancy other than adequately treated nonmelanoma skin cancer. Patients with hepatocellular carcinomas were included unless they had nodule(s) >5 cm in diameter, more than three nodules >3 cm in diameter, evidence of vascular invasion or evidence of metastases or local spread beyond the parenchyma of the liver.

Treatment regimens

Patients were randomized centrally within 12 h after surgery using a randomization list with blocks of six patients. Randomization was balanced within each center. Patients were randomized in a 1:1:1 ratio to receive:

Group A: the recommended dosage of tacrolimus (Prograf®) either orally in two divided doses approximately 12 h apart, or intravenously as a 24-h infusion if the patient was unable to tolerate oral medication. The initial doses were 0.10–0.15 mg/kg/day by mouth or 0.01–0.05 mg/kg/day intravenously, adjusted to achieve target whole blood trough levels of >10 ng/mL for the first month posttransplant. Thereafter, dosing was according to documented center practice. Corticosteroids were given according to the local practice.

Group B: triple therapy with tacrolimus; initial oral doses were 0.05–0.10 mg/kg/day and intravenous doses were 0.008–0.04 mg/kg/day, adjusted to achieve target trough levels of ≤8 ng/mL for the duration of the study. Corticosteroid therapy was administered according to local practice and MMF (CellCept®) 1g was given twice daily, intravenously for the first 5 days to ensure adequate therapeutic concentrations (20,21) and then orally.

Group C: as for group B but the introduction of tacrolimus was delayed until the fifth postoperative day, and additional immunosuppression provided by daclizumab (Zenapax®; 2 mg/kg within 12 h after transplantation and a second dose of 1 mg/kg given at 7 days).

Treatment with the following drugs was prohibited during the study: cholestyramine or colestipol, magnesium and aluminium hydroxide-containing antacids, any other immunosuppressive agents except for antilymphocyte antibodies or sirolimus, which were allowed for treatment of intractable rejection.

Study assessments

Hematology assessments and blood chemistry such as serum creatinine, bilirubin (total, direct and indirect), aspartate aminotransferase, alanine aminotransferase, gamma glutamyl transferase and total protein and albumin were assessed at baseline (day 0, defined as the day of liver transplantation), days 2, 7 and weeks 2, 6, 13, 26, 39 and 52 (or at premature discontinuation, if applicable). Whole blood tacrolimus trough levels were measured on days 2, 7 and weeks 2, 6, 13, 26 and 52 (or on premature discontinuation).


The primary endpoint was the change from baseline in calculated creatinine clearance (estimated using the Cockcroft–Gault formula (22)) at week 52. In addition, an assessment of renal function was made using an abbreviated Modification of Diet in Renal Disease (MDRD) formula based on estimations of glomerular filtration rate (GFR) (23). Baseline creatinine clearance was the one obtained closest and prior to transplant. A value of 15 mL/min was used in the calculation for patients on dialysis posttransplant, when creatinine clearance was to be assessed.

Secondary endpoints included: the change from baseline in serum creatinine at 52 weeks posttransplantation; change from baseline in calculated creatinine clearance at 26 weeks posttransplantation; the requirement for renal replacement therapy between 2 weeks and 52 weeks posttransplantation; biopsy-proven acute rejection (requiring pulse immunosuppression therapy) up to 26 weeks and 52 weeks posttransplantation; the composite endpoint of a 20% (or greater) decrease from baseline in calculated creatinine clearance, or acute rejection, or graft loss or death; acute rejection (biopsy-proven or presumed on clinical grounds only) requiring increased immunosuppressive therapy during 26 and 52 weeks; time to first biopsy-proven acute rejection requiring additional immunosuppression therapy up to 52 weeks; patient and graft survival at 52 weeks; time to graft loss or death up to 52 weeks; incidence of histologically determined recurrence of underlying liver disease posttransplantation. Patients were followed up to week 52 even if prematurely discontinuing study treatments.

Adverse events

All adverse events, including clinically significant abnormalities in clinical and laboratory parameters, were recorded throughout and at the safety follow-up visit. Adverse events were coded using the Medical Dictionary for Regulatory Activities (MedDRA version 7.0). The recurrence of hepatitis C was collected via adverse event reporting without central review of biopsies or a standardized approach to following such patients. Adverse events of interest identified prior to initiation of the study included hypertension, diabetes mellitus, hyperlipidemia, diarrhea, leukopenia and opportunistic infections (cytomegalovirus [CMV] infection/disease, Aspergillus, Candida, Pneumocystis carinii, Cryptococcus, Listeria monocytogenes, Herpes zoster virus, and Herpes simplex virus).

Ethics:  The study was conducted in compliance with the Declaration of Helsinki and Good Clinical Practice guidelines and in accordance with local and national regulatory requirements and laws. All relevant study documentation was approved by the Independent Ethics Committee or Institutional Review Board responsible for each study center. All patients gave signed informed consent and could withdraw from the study at any time.

Statistical analyses

Sample size: For a 90% power to detect a pair-wise treatment difference of 10 mL/min at a two-sided significance level of 2.5% (standard deviation in baseline estimated creatinine clearance of 25 mL/min), 157 patients were required per treatment group. To allow for a 10% drop out rate, the sample size was increased to 175 patients per treatment group, giving a total of 525 patients for the study.

The full analysis set (FAS) population (the primary analysis population) included all randomized patients who received at least one dose of any immunosuppressive medication according to the assigned study group and had at least one calculated creatinine clearance value at baseline and posttreatment. The per-protocol (PP) population included any patient in the FAS population who met the following criteria: no inclusion/exclusion criteria violation; at least one calculated creatinine clearance value beyond 6 months posttransplant; remained in the study beyond day 5; did not receive prohibited medication during the first 14 days after transplantation, and for less than 1 week at any time during the study; and was treated according to the protocol. The safety patient population included all patients who received at least one dose of any immunosuppressive medication as per the assigned study group and for whom at least one visit was recorded after the first dose.

Analysis of the primary endpoint was a pair-wise comparison of the change from baseline in calculated creatinine clearance at week 52 (FAS population) in group A versus groups B and C. For each comparison, statistical significance was set at 0.025, two-sided test. The same method was used to analyze the secondary endpoint of the change from baseline at week 52 in calculated creatinine clearance in the PP group. Analysis of covariance (ANCOVA) with treatment group, center and baseline values as factors in the model was used to analyze the change from baseline in serum creatinine and calculated creatinine clearance. The requirement for renal dialysis was analyzed using logistic regression methods, with treatment group, center and baseline calculated creatinine clearance as factors. The Cochran–Mantel–Haenszel test (stratified by center) was used to test for between-group differences for other secondary efficacy endpoints where appropriate. A sensitivity analysis was undertaken, in which no imputation of serum creatinine was performed in the cases of dialysis. In addition, the same analysis of serum creatinine (without imputation) was performed using LOCF (last observation carried forward) to account for missing values up to week 52, but only in cases where the final measurement was obtained between 40 weeks and 52 weeks.

Adverse events were summarized by treatment group using descriptive statistics, except for between-group comparisons for prospectively selected adverse events and opportunistic infections, which were analyzed using the chi-square test or Fisher's exact test. Statistical significance was declared for p < 0.05 and no correction for multiplicity was performed.



This study was conducted between February 2004 and February 2007 at 30 centers in 9 countries. Of 525 patients randomized to receive treatment, 517 were included in the FAS population—all patients were followed to 12 months unless they withdrew consent or withdrew from treatment according to the protocol for reasons other than death, graft loss or retransplantation. The flow of patients is outlined in Figure 1. Of the FAS population, 186 patients (36.0%) discontinued the study prematurely: 83 (44.6%) from group A, 52 (28.0%) from group B and 51 (27.4%) from group C. The main reasons for premature withdrawal were adverse events or intercurrent illness (Figure 1); 331 patients completed the study.

Figure 1.

Disposition of patients. Group A received standard-dose tacrolimus and corticosteroids, group B received MMF, reduced-dose tacrolimus and corticosteroids and group C received daclizumab induction, MMF, reduced-dose tacrolimus given 5 days after transplantation and corticosteroids. FAS = full analysis set.

Baseline patient demographics and disease/transplant characteristics for the FAS population were broadly similar between groups, including the mean model for end-stage liver disease (MELD) scores (Table 1). Pretransplant renal function was lowest in group C, both for the FAS (Table 1) and PP populations.

Table 1.  Baseline recipient and donor characteristics of the FAS population
 Standard-dose tacrolimus, corticosteroids (group A) (N = 181)MMF, reduced-dose tacrolimus corticosteroids (group B) (N = 168)Daclizumab induction, MMF, delayed reduced-dose tacrolimus and corticosteroids (group C) (N = 168)
  1. 1No. of evaluable pts: 170 in standard therapy group, 158 in reduced tacrolimus + MMF group and 159 in reduced, delayed tacrolimus + MMF group.

  2. 2Patients may have had more than one cause of liver disease. Some patients with liver disease designated ‘other’ initially were reassigned to more specific categories following medical review.

  3. CMV = cytomegalovirus; GFR = glomerular filtration rate; MELD = model for end-stage liver disease; MMF = mycophenolate mofetil; SD = standard deviation.

Recipient age, median years (min, max)53.0 (19, 69)54.0 (18, 73)55.0 (26, 70)
Male, N (%)127 (70.2) 109 (64.9) 116 (69.0) 
Caucasian, N (%)173 (95.6) 163 (97.0) 164 (97.6) 
Mean (± SD) calculated creatinine clearance based101.83 (44.15)   104.31 (37.15)  96.51 (34.94)  
 on Cockcroft–Gault (mL/min)101.83 (44.15)   104.31 (37.15)  96.51 (34.94)  
Mean (± SD) MELD score115.4 (6.57)  15.2 (5.59) 15.6 (6.26)  
Median (range) MELD score (unadjusted for HCC)    14 (6 − 49)  15 (5 − 33)  15 (5 − 41)
Cause of liver disease, N (%)2
 Alcoholic liver disease79 (43.6)72 (42.9)76 (45.2)
 Primary biliary cirrhosis18 (9.9)  14 (8.3) 12 (7.1) 
 Primary sclerosing cholangitis13 (7.2)  18 (10.7)13 (7.7) 
 Hepatitis B17 (9.4)  13 (7.7) 10 (6.0) 
 Hepatitis C36 (19.9)34 (20.2)39 (23.2)
 Hepatitis of unknown etiology2 (1.1)04 (2.4)
 Fulminant hepatic failure/toxic hepatitis004 (2.4)
 Hepatocellular carcinoma37 (20.4)38 (22.6)38 (22.6)
 Autoimmune hepatitis7 (3.9)5 (3.0)6 (3.6)
 Other30 (16.6)28 (16.7)28 (16.7)
Donor age, mean (± SD) years48.4 (15.58) 48.3 (16.86) 48.0 (17.61) 
Type of donor, N (%)
 Deceased, heart beating146 (80.7) 131 (78.0) 146 (86.9) 
 Deceased, nonheart beating26 (14.4)28 (16.7)19 (11.3)
 Living, related4 (2.2)3 (1.8)1 (0.6)
 Living, unrelated5 (2.8)6 (3.6)2 (1.2)
Donor/recipient CMV serological status, N (%)
 +/+77 (42.5)60 (35.7)63 (37.5)
 +/−36 (19.9)38 (22.5)38 (22.5)
 −/+37 (20.4)29 (17.3)31 (18.5)
 −/−21 (11.6)26 (15.5)25 (14.9)

Exposure to study medication

Most patients (89–95%) initially received oral tacrolimus. Mean tacrolimus trough levels throughout the study in all three treatment groups are shown in Figure 2. From month 4, mean tacrolimus trough levels between treatment groups tended to converge (Figure 2). The mean (± SD) tacrolimus concentrations for patients in the three groups (A, B and C) were 10.7 ± 3.7, 8.6 ± 3.2, 8.4 ± 2.8 ng/mL (at 2 weeks); 11.1 ± 3.2, 8.8 ± 2.8 and 7.7 ± 2.3 ng/mL (at 1 month); 10.4 ± 3.9, 9.1 ± 3.5 and 8.9 ± 3.1 ng/mL (at 3 months); 9.7 ± 3.1, 8.4 ± 2.8 and 8.4 ± 2.7 ng/mL (4–6 months) and 8.5 ± 3.7, 7.8 ± 2.7 and 7.4 ± 2.8 ng/mL (10–12 months).

Figure 2.

Mean (SD) whole blood tacrolimus trough levels in liver allograft recipients treated with standard-dose tacrolimus and corticosteroids (group A); MMF, reduced-dose tacrolimus and corticosteroids (group B); or daclizumab induction, MMF, reduced-dose tacrolimus given 5 days after transplantation and corticosteroids (group C). Patient numbers at month 1, 2, 3, 4–6, 7–9 and 10–12 were 181, 155, 133, 134, 118 and 115, respectively, in group A, 168, 151, 122, 134, 121 and 113, respectively, in group B and 161, 151, 127, 131, 123 and 120, respectively, in group C. The gray lines represent the minimum target trough level for the standard therapy group and the maximum target trough level for the groups receiving MMF.

Mean daily doses of MMF from week 1 to month 12 ranged from 1562 ± 587 to 1959 ± 382 mg/day in group B, and from 1555 ± 531 to 1886 ± 357 mg/day in group C. In group C, the median and mean first daclizumab doses were both 2.0 mg/kg.

Corticosteroid doses were similar for all three groups for all time points. During week 4, mean (± SD) daily maintenance doses of corticosteroids (dose calculated based on equivalence to prednisolone) for groups A, B and C were 25.5 ± 28.9, 22.3 ± 23.8, and 24.7 ± 44.0 mg; at 5–6 months 11.3 ± 11.5, 9.1 ± 10.1, and 22.5 ± 110.9 mg and 8.3 ± 11.4, 7.6 ± 7.3 and 24.8 ± 156.3 mg (9–12 months). Cumulative dose was 13.8, 13.5 and 13.1 mg/kg at 1 month; 40.3, 37.2 and 35.2 mg/kg at 5–6 months; and 55.0, 58.3 and 49.4 mg/kg at 9–12 months.

Concomitant medication

Almost all patients (≥99%) received concomitant medication. The use of nonsteroidal anti-inflammatory drugs (NSAIDs) and aminoglycoside antimicrobials was similar in each group.

Renal function

The change from baseline at week 52 in calculated creatinine clearance was significantly smaller (p = 0.012) in patients who received delayed introduction of reduced-dose tacrolimus (group C) than those in group A, whereas there was no statistically significant between-group difference for patients in groups A and B (p = 0.199) (Figure 3A).

Figure 3.

The mean change from baseline in calculated creatinine clearance based on the Cockcroft–Gault formula at 52 weeks in liver allograft recipients from the (A) FAS population and (B) PP population. Bars represent 95% confidence intervals for each mean. *p < 0.05 versus group A. FAS = full analysis set; PP = per protocol.

Sensitivity analyses were performed as described in the methods. The type of sensitivity analysis performed did not affect the primary efficacy comparison. PP population analysis of the change from baseline in calculated creatinine clearance is shown in Figure 3B. The group C regimen was also associated with less renal function impairment than the standard tacrolimus regimen at earlier time points. The decrease from baseline in calculated creatinine clearance was smaller (p < 0.001) in this group than in group A at week 13 (−13.23 vs. −25.96 mL/min) and 26 (−12.36 vs. −26.88 mL/min).

Changes in renal function assessed by estimating GFR based on the abbreviated MDRD formula (Table 2) were consistent with those based on the Cockcroft–Gault formula. There was a positive and significant correlation between estimated GFR by Cockcroft–Gault versus MDRD formula (R2= 0.7112, p < 0.0001).

Table 2.  Primary and selected secondary endpoints (FAS population)
 Standard-dose tacrolimus, corticosteroids (group A) (N = 181)MMF, reduced-dose tacrolimus corticosteroids (group B) (N = 168)Daclizumab induction, MMF, delayed reduced-dose tacrolimus and corticosteroids (group C) (N = 168)
  1. BPAR = biopsy-proven acute rejection; GFR = glomerular filtration rate; MDRD = modification of diet in renal disease.

  2. 1Post hoc analysis.

  3. 2 The composite secondary efficacy parameter was a 20% or greater decrease from baseline in calculated creatinine clearance, or acute rejection, or graft loss or death.

  4. *p < 0.05, **p < 0.01, ***p < 0.001 versus group A.

Mean (± SD) baseline GFR based on MDRD-4 (mL/min)193.46 (39.63)95.89 (36.81)86.71 (29.20)
 Mean (±SD) change from baseline in GFR based on MDRD-4 at week 52 (mL/min)1−24.89 (36.45)−22.60 (35.5)−14.62 (29.17)
 Mean (± SD) baseline GFR based on the Cockcroft–Gault formula at (mL/min)101.83 (44.16)104.31(± 37.15)96.51 (± 34.94)
 Mean (95%CI) change from baseline in GFR based on the Cockcroft–Gault formula at week 52 (mL/min)−23.61−21.22−13.63*
(−28.60, −18.62)(−26.12, −16.32)(−17.96, −9.30)
Mean (± SD) baseline serum creatinine levels (ng/mL)0.97 (0.35)0.92 (0.27)1.00 (0.36)
Mean (95%CI) change from baseline in serum creatinine levels (ng/mL)
(0.16, 0.29)(0.10, 0.21)*(0.04, 0.18)**
(0.16, 0.31)(0.12, 0.25)(0.01, 0.21)*
 Week 520.230.170.12
(0.17, 0.29)(0.12, 0.22)**(0.06, 0.19)*
Number of episodes of BPAR or presumed acute rejection during 52 weeks, N (%)
 0 episodes119 (65.7)106 (63.1) 124 (73.8)
 1 episode51 (28.2)52 (31.0)  39 (23.2)
 2 episodes10 (5.5) 8 (4.8)  5 (3.0)
 3 episodes1 (0.6)1 (0.6)0
 >3 episodes01 (0.6)0
BPAR requiring treatment, N (%)
 Up to week 2642 (23.2)42 (25.0)     25 (14.9)*
 Up to week 5244 (24.3)45 (26.8)   28 (16.7)
BPAR or presumed acute rejection requiring treatment, N (%)
 Up to week 2648 (26.5)45 (26.8)    29 (17.3)*
 Up to week 5250 (27.6)49 (29.2)  32 (19.0)
Patient survival up to week 52, N (%)164 (90.6) 149 (88.7) 157 (93.5) 
Graft survival up to week 52 (n, %)170 (93.9) 158 (94.0) 156 (92.9) 
Recurrence of liver disease at week 52, N (%)7 (9.7)10 (14.9)    15 (22.1)*
Composite endpoint of renal function and graft-related outcomes2 (n, %)167 (92.3) 151 (89.9)      130 (77.4)***

Increases from baseline in serum creatinine levels were significantly smaller in groups B and C than group A, except at week 26, where changes in group B were not statistically different (p = 0.081) from group A (Table 2). Significantly, fewer patients required dialysis between weeks 2 and 52 in group C (n = 7; 4.2%) than group A (n = 18; 9.9%) [p = 0.0367], but the between group difference between groups A and B (n = 7; 4.2%) was not statistically significant (p = 0.0558).

Other endpoints

Other endpoints are summarized in Table 2. There were no clinically significant differences between groups A and B for the occurrence of biopsy-proven or presumed acute rejection, first biopsy-proven acute rejection requiring treatment and graft loss or death during the 52-week treatment period. A similar percentage of patients across all three groups experienced graft loss (6–7%). The composite endpoint of renal function and graft-related outcomes was lower for group C compared with group A (p ≤ 0.001). Outcomes related to rejection requiring treatment at week 26 were also significantly lower in group C versus group A (p < 0.05; Table 2).

Adverse events

A summary of the overall adverse event profile is presented in Table 3. Most treatment-emergent adverse events were of mild or moderate severity as defined by the investigator. Adverse events leading to withdrawal or death: there were more withdrawals because of adverse events in group A than in the other two groups (Table 3). With the exception of renal insufficiency, there were no notable differences between the three groups with respect to the nature of adverse event that led to study withdrawal. Infections were the most frequent adverse events leading to death (Table 3).

Table 3.  Overall adverse event summary for the safety population, N (%)
 Standard-dose tacrolimus, corticosteroids (group A) (N = 182)MMF, reduced-dose tacrolimus corticosteroids (group B) (N = 169)Daclizumab induction, MMF, delayed reduced-dose tacrolimus and corticosteroids (group C) (N = 169)
  1. 1Treatment-related was defined as a highly probable, probable, possible, unlikely or missing relationship.

  2. 2The most common adverse events leading to death were infections, occurring in 1.1%, 3.0% and 2.4% in groups A, B and C, respectively.

  3. 3The most frequent adverse events leading to study discontinuation were graft dysfunction (3.3% in group A, 2.4% in group B, and 3.0% in group C) and renal insufficiency (6.0%, 1.2%, 0%, respectively).

  4. AE = adverse event; SAE = serious adverse event.

AE182 (100) 167 (98.8)166 (98.2)
SAE113 (62.1)107 (63.3)110 (65.1)
Treatment-related AEs1161 (88.5)149 (88.2)146 (86.4)
Treatment-related SAEs 58 (31.9) 65 (38.5) 63 (37.3)
AEs leading to death216 (8.8)16 (9.5) 7 (4.1)
AEs leading to withdrawal3 55 (30.2) 35 (20.7) 34 (20.1)
Opportunistic infections 40 (22.0) 45 (26.6) 43 (25.4)
Malignancies 4 (2.2) 1 (0.6) 3 (1.8)

Adverse events of special interest:  Treatment-emergent diarrhea was not more common in the groups receiving MMF (groups B and C) than in group A (Table 4). Indeed, most treatment-emergent adverse events occurred with a broadly similar frequency between treatment groups, with some notable exceptions. Hypertension occurred less frequently in groups B and C versus group A, and leukopenia occurred more often in group C than group A (p = 0.0062) (Table 4).

Table 4.  Incidence (N, (%)) of protocol-specified select treatment-emergent adverse events and adverse events occurring in ≥ 10% of patients in any group and with a difference in incidence of at least 5% between any two groups
 Standard-dose tacrolimus, corticosteroids (group A) (N = 182)MMF, reduced-dose tacrolimus corticosteroids (group B) (N = 169)Daclizumab induction, MMF, delayed reduced-dose tacrolimus and corticosteroids (group C) (N = 169)
  1. Statistical analysis was performed on pre-specified parameters (hypertension and leukopenia), with those showing a significant difference from standard therapy shown in the table: *p = 0.0252; **p = 0.0095, ***p = 0.0062 versus standard therapy group.

Hypertension74 (40.7) 49 (29.0)*  46 (27.2)**
Diabetes mellitus87 (47.8)72 (42.6)66 (39.1)
Hyperlipidemia1 (6.0)10 (5.9) 8 (4.7)
Diarrhea39 (21.4)52 (30.8)41 (24.3)
Leukopenia12 (6.6) 18 (10.7)  27 (16.0)***
Thrombocytopenia11 (6.0) 21 (12.4)16 (9.5) 
Hepatitis C infection4 (2.2)9 (5.3)17 (10.1)
Back pain17 (9.3) 24 (14.2)24 (14.2)
Insomnia30 (16.5)36 (21.3)27 (16.0)
Pruritis13 (7.1) 8 (4.7)17 (10.1)
Renal insufficiency39 (21.4)20 (11.8)13 (7.7) 

The overall incidence of opportunistic infections was similar in all groups (Table 3). Serious adverse events occurred in about two-thirds of patients in each group (Table 3), the most common being graft dysfunction (4.4%, 4.7% and 3.6% in groups A, B and C) and CMV syndrome (3.3%, 3.0%, 5.3%, respectively). Malignancies were not common (Table 3) and specific malignancies did not occur in more than one patient in any treatment group.

There were no notable differences in mean blood chemistry or hematology values or body mass index between the three treatment groups at day 0, week 26 or 52.


This study demonstrated that delayed introduction of reduced-dose tacrolimus, under the protection of MMF and daclizumab (group C), is associated with less impairment of renal function compared with standard therapy with tacrolimus and corticosteroid (group A) without an increased frequency of rejection, graft loss or death. However, there was no such benefit with a triple therapy regimen of MMF, reduced-dose tacrolimus and corticosteroids (group B) initiated immediately after transplant relative to the standard therapy.

We recognize that there are limitations in this study: with increasing experience, regimes of immunosuppression have been modified and, as this was an open-label study conducted in experienced high-volume centers, clinicians had a low threshold to modify treatment regimes. In the registration studies, and according to the license for its use, initial tacrolimus target trough levels are 5–20 ng/mL (24), which are then typically tapered to maintenance target trough levels of 5–10 ng/mL (4,25). Target tacrolimus trough levels and the use of daclizumab were determined by the Steering Group: levels in Group A were in line with recommended levels. However, we recognize that both target and achieved levels were greater than those used in most centers currently. Daclizumab is not currently indicated for use in liver transplant patients, although it is effective and safe in this patient population (26,27); the dose used in our study was agreed based on experience from renal allograft recipients (28). Although the combination of tacrolimus and MMF is not currently licensed for use in liver transplant recipients, they are commonly coprescribed (2) and are an effective combination (29–33). In addition, there was a relatively high rate of withdrawal, especially in group A where many patients were withdrawn so mycophenolate could be given. To address these concerns, we defined, post hoc, a ‘per protocol’ set of patients where levels of tacrolimus achieved were within 10% of target; in this smaller group of patients, similar conclusions were reached.

There was no significant adverse impact from the two reduced-dose tacrolimus treatment regimes (B and C), as patient and graft loss were similar and there was no increased risk of infection, or biopsy-proven or treated rejection. Our study was not designed to evaluate the effect of induction therapy, so it is not possible to exclude an effect of IL2-receptor on the overall rate of rejection in group C. Moreover, the mechanism underlying the benefits in renal function conferred by delayed introduction of tacrolimus remains to be fully elucidated. It is possible that the kidney, put at risk by reperfusion and/or in combination with other transplant-related events, is more susceptible to the effects of CNI in the immediate posttransplant period, so delaying administration by even a few days provides benefits in terms of renal function.

Renal function was estimated by calculated creatinine clearance using the Cockcroft–Gault formula (22). Although direct measurement by inulin clearance is the gold standard (34), it is not practical in such a large-scale clinical trial setting. We were able to show, however, that GFR by Cockcroft–Gault (22) was correlated with GFR by MDRD (23). Moreover, the changes from baseline in serum creatinine levels in each group were consistent with the changes in calculated creatinine clearance.

Our study confirms results from others (17–19). In one study only 148 patients were enrolled, including those with pretransplant renal dysfunction, and withdrawal rates were 16–22% (19). Withdrawal rates in this study were about 30% in the two groups receiving MMF (with or without induction), corticosteroids plus tacrolimus, which were lower than those in the standard therapy group (46%), primarily due to a reduction in withdrawal for poor renal function.

The safety findings of this study were consistent with the known safety profiles of the drugs involved (29,35). Leukopenia and diarrhea were more common in those receiving MMF, although it should be noted that there were inconsistencies with the increases in the risk of these side-effects between the two groups receiving MMF. It is likely that group C patients had adequate immunosuppression: rejection rates were slightly lower in this group and the incidence of CMV disease was highest.

Although we have shown that delayed reduced-dose tacrolimus is associated with a lower reduction in renal function at 1 year, it remains to be shown whether this will translate into a reduced risk for late-onset end-stage renal disease, need for renal replacement therapy and lower mortality risk. However, renal function at 9 and at 12 months (10,36) is predictive of development of end-stage renal disease and chronic renal dysfunction in liver transplant recipients up to 4 (36), 6 (10) or 10 years (37) posttransplant. We also do not know whether the use of IL2-receptor blockade is needed. Finally, we intentionally excluded those with significant renal disease pretransplant, to assess the effectiveness of the tacrolimus-sparing regimens used in groups B and C in maintaining renal function; thus, we acknowledge that the patients included in the study may have had better renal function and a lower MELD score (unadjusted for cancer) than the typical liver transplant patient. However, it seems reasonable to assume that the benefits of a regimen of delayed low-dose tacrolimus seen in this study may be even more marked in those with a high MELD score and worse renal function. Further studies in patients with high MELD score and poor renal function are required to confirm this hypothesis.

The differences between groups in HCV recurrence rates following transplantation could not be interpreted because there was no formal or consistent approach to evaluate this, no requirement for graft liver histology and the study was not designed or powered for such an analysis.

In conclusion, in liver transplant recipients, a regimen of daclizumab induction, corticosteroids, MMF and reduced-dose tacrolimus given 5 days after transplantation was associated with less renal function impairment at 52 weeks after transplantation than the control regimen that included only standard-dose tacrolimus plus corticosteroids. The importance of the delayed administration of tacrolimus in the improvement of renal function seen in this study is underlined by the immediate introduction of reduced-dose tacrolimus regimen in combination with MMF and corticosteroids offering no advantage in terms of renal function versus the standard therapy group.


Editorial assistance was provided by Richard Glover, Tracy Harrison and Sheridan Henness under the guidance of James Neuberger.

The ReSpECT Study Group comprises the following Investigators (in order of recruitment): P. Neuhaus (Charite Universitätsmedizin Berlin, Germany); J. Neuberger and D Mayer (Queen Elizabeth Medical Centre, Birmingham, UK); J. Pirenne (UZ Gasthuisberg, Leuven, Belgium); D. Samuel (Hopital Paul Brousse, Villejuif, France); H. Isoniemi Helsinki (University Hospital, Transplantation Unit, Helsinki, Findland); L. Rostaing (CHU Rangueil, Toulouse, France); A. Rimola (Hospital Clinic, Barcelona, Spain); J. Fabregat-Prous [previously Dr Figueras] (Cirugía General y Digestiva Cuitat Sanitaria i Universitaria de Bellvitge, Barcelona, Spain); F. Durand (Service d’Hepatologie Hospital, Beaujon, Clichy, France); S. Friman (Sahlgrenska University Hospital, Goteburg, Sweden); W. Bechstein (Goethe-Universität, Frankfurt am Main, Germany); J. Schmidt and P. Schemmer (Universitätsklinikum Heidelberg, Germany); J.P. Hauss (Universitätsklinikum Leipzig, AöR, Leipzig, Germany); P.A. Clavien (Universitätsspital Zurich, Switzerland); C.E. Broelsch (Universitätsklinikum Essen, Germany); A. Bernardos (Hospital Univeritario Virgin del Rocio, Sevilla, Spain); O. Detry (C.H.U. Sart-Tilman, Liege, Belgium); O. Bouillot (Hopital Edouard Herriot, Lyon, France); G. Soderdahl (Huddinge, Stockholm, Sweden); I.B. Brekke and O. Bentdal (Rikshospitalet, Oslo, Norway); A. Mueller and B. Dreske (Universitätsklinikum Kiel, Germany); I. Colle (UZ Ghent, Belgium); V. Cuervas-Mons (Hepatologia Hospital Puerta de Hierro, Madrid, Spain); M Salcedo (Digestivo Hospital Gregorio Marañón, Madrid, Spain); C. Zuelke and H. Schlitt (Klinikum der Universität Regensberg, Regensberg, Germany); MF. Suárez (Hospital Juan Canalejo, Dr La Coruña, Spain); D. Patch (Royal Free Hospital, London, UK); N. Declerck (Hospital ClaudeHuriez, Lille, France); E. Varo (Hospital Clinico de Santiago, Santiago de Compostela, Spain) and N Boon (Hopital Erasme, Brussels, Belgium).

Funding Sources: Funding for the study was provided by F. Hoffmann-La Roche Ltd. (Basel, Switzerland) and the University Hospital Birmingham National Health Service (NHS) Trust (Birmingham, United Kingdom).

Conflict of Interest Statement: JMN has received speaker support from Astellas, F. Hoffmann-La Roche, and Novartis and is a consultant for Bristol Myers. RDM has received consulting fees from F. Hoffmann-La Roche. PN, JP, DS, HI, LR, AR and DM do not have any financial conflict of interest related to this manuscript. SM is a consultant for F. Hoffmann-La Roche.