• CsA;
  • cyclosporine;
  • diabetes;
  • hyperglycemia;
  • new-onset diabetes;
  • tacrolimus


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendix

DIRECT (Diabetes Incidence after Renal Transplantation: Neoral C2 Monitoring Versus Tacrolimus) was a 6-month, open-label, randomized, multicenter study which used American Diabetes Association/World Health Organization criteria to define glucose abnormalities. De novo renal transplant patients were randomized to cyclosporine microemulsion (CsA-ME, using C2 monitoring) or tacrolimus, with mycophenolic acid, steroids and basiliximab. The intent-to-treat population comprised 682 patients (336 CsA-ME, 346 tacrolimus): 567 were nondiabetic at baseline. Demographics, diabetes risk factors and steroid doses were similar between treatment groups. The primary safety endpoint, new-onset diabetes after transplant (NODAT) or impaired fasting glucose (IFG) at 6 months, occurred in 73 CsA-ME patients (26.0%) and 96 tacrolimus patients (33.6%, p = 0.046). The primary efficacy endpoint, biopsy-proven acute rejection, graft loss or death at 6 months, occurred in 43 CsA-ME patients (12.8%) and 34 tacrolimus patients (9.8%, p = 0.211). Mean glomerular filtration rate (Cockcroft–Gault) was 63.6 ± 20.7 mL/min/1.73 m2 in the CsA-ME cohort and 65.9 ± 23.1 mL/min/1.73 m2 with tacrolimus (p = 0.285); mean serum creatinine was 139 ± 58 and 133 ± 57 μmol/L, respectively (p = 0.005). Blood pressure was similar between treatment groups at month 6, but total cholesterol, LDL-cholesterol and triglyceride levels were significantly higher with CsA than with tacrolimus (total cholesterol:HDL remained unchanged). The profile and incidence of adverse events were similar between treatments. The incidence of NODAT or IFG at 6 months post-transplant is significantly lower with CsA-ME than with tacrolimus without a significant difference in short-term outcome.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendix

Development of new-onset diabetes after transplant (NODAT) or impaired fasting glucose (IFG) after renal transplantation increases the risk of cardiac events, peripheral vascular disease, graft failure and death (1–3). Indeed, diabetes confers a higher risk of ischemic heart disease than hypertension, smoking or severe hyperlipidemia in renal transplant patients (4), and onset of diabetes after transplantation is associated with more than a 60% increase in risk of graft failure and an almost a 90% increase in risk of death (1). Less severe hyperglycemic abnormalities are also associated with increased cardiovascular risk. In the general population, the DECODE study in 25 413 individuals has shown conclusively that fasting glucose level is significantly related to 5- and 10-year cardiovascular mortality (5), while in the renal transplant setting Cosio et al. have demonstrated that IFG (5.6–6.9 mmol/L) is associated with a significantly higher incidence of posttransplant cardiac events and peripheral vascular disease compared to normoglycemic recipients (6). However, since few studies have used the American Diabetes Association (ADA) (7) or World Health Organization (WHO) (8) definitions for NODAT or IFG, and fewer have employed oral glucose tolerance testing (OGTT), the incidence of glucose metabolism abnormalities in the renal transplant population is underestimated (6,9).

Established risk factors for NODAT following renal transplantation include increasing age, obesity, black or Hispanic ethnicity, family history, hepatitis C and cytomegalovirus (CMV) infection, steroid use and type of calcineurin inhibitor (1,2,10). NODAT is caused by insufficient insulin release in response to an increase in insulin resistance. Since calcineurin inhibitors are commonly used with steroids, it is difficult to identify whether diabetes in a posttransplant setting is due to steroids, calcineurin inhibitors or both. Steroids are known to increase peripheral insulin resistance. Tacrolimus also appears to inhibit insulin production. In vitro, tacrolimus depletes pancreatic beta-cell insulin mRNA and protein in a dose-dependent manner (11) and, consistent with this, tacrolimus-based regimens are associated with more marked beta-cell morphological changes (12,13) and inhibit insulin secretion to a greater extent than CsA-based regimens during the first 3 months post-transplant (13). Meta- (14,15) and registry analyses (1,16) have indicated that tacrolimus is associated with increased incidence of NODAT. To date, however, no randomized trial of the two agents has been designed with NODAT or other glycemic abnormalities as a primary endpoint.

The DIRECT (Diabetes Incidence after Renal Transplantation: Neoral C2 Monitoring Versus Tacrolimus) study was designed to address this question by using a combined primary endpoint of NODAT or IFG, selected because both conditions are predictive for poor patient outcomes (1,2,6). A co-primary endpoint was the prevention of biopsy-proven acute rejection (BPAR) or graft loss or death, since any benefit in diabetogenic risk must be balanced against immunosuppressive efficacy.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendix

Study design and conduct

DIRECT was a 6-month, open-label, randomized multicenter study undertaken at transplant centers in which de novo renal transplant patients were randomized in a 1:1 ratio to receive either cyclosporine microemulsion (CsA-ME) or tacrolimus as previously described (17). Written informed consent was obtained from all patients and the study was conducted in full compliance with the amended Declaration of Helsinki following approval from the Institutional Review Committee at each center.

Patient population

Adult recipients (18–70 years) of a first or second renal transplant from a deceased, living-related or living-unrelated donor were eligible for recruitment. Main exclusion criteria included multiple organ transplant, donation after cardiac death or from a donor who tested positive for hepatitis B or C virus.

Randomization and immunosuppression

Patients were stratified at the time of transplant into three groups: (i) existing Type 1 or 2 diabetes; (ii) nondiabetic, white or (iii) nondiabetic, nonwhite. They were then randomized 1:1 within each group to CsA-ME C2 monitoring or tacrolimus using an automated system that informs investigators of treatment allocation via an interactive voice response system. Patients were classified as having diabetes if they met the ADA/WHO criteria (7,8) or were receiving hypoglycemic treatment at time of transplant or prior to dialysis.

CsA-ME (Neoral®, Novartis Pharma AG, Basel, Switzerland) was initiated within 24 h of transplantation at 10 mg/kg/day, adjusted to achieve a C2 target of 1600 ng/mL (range 1400–1800 ng/mL) during month 1, 1400 ng/mL (range 1200–1600 ng/mL) during months 2–3 and 1000 ng/mL (range 800–1200 ng/mL) during months 4–6. Tacrolimus (Prograf®, Astellas Pharma, Tokyo, Japan) was initiated at a dose of 0.2 mg/kg/day within 24 h post-transplant, adjusted to achieve C0 within the range 10–15 ng/mL during months 1–3 and 5–10 ng/mL during months 4–6. An initial dose of methylprednisolone 500 mg i.v. was given, with oral prednisone tapered from 100–200 mg/day on day 1 to 5–10 mg/day from month 3 onward. All patients received mycophenolate mofetil (MMF, Cellcept®, Roche Pharmaceuticals, Basel, Switzerland) or enteric-coated mycophenolate sodium (EC-MPS, myfortic®, Novartis Pharma AG), administered according to local practice but initiated within 24 h of graft reperfusion. Different MMF or EC-MPS dosages were permitted in the two treatment groups in accordance with the differences in pharmacokinetic interaction observed with tacrolimus and CsA-ME (18–20). However, once the dosage was defined for each group within a center, all patients within each group at that center had to receive the same dosage. Basiliximab 20 mg (Simulect®, Novartis Pharma AG) was administered intravenously on day 0 and day 4. In suspected cases of acute rejection an allograft biopsy was to be performed prior to or within 24 h of the start of antirejection therapy and graded according to Banff 1997 criteria (21).

Study endpoints

The primary safety endpoint was a composite of NODAT or IFG within the first 6 months post-transplant among patients classified as nondiabetic at time of transplant. The primary efficacy endpoint was a composite of BPAR, graft loss or death at 6 months posttransplant in all patients.


In addition to standard posttransplant laboratory measurements, HbA1c and C-peptide levels were measured at baseline, day 90 and day 180, and insulin levels were measured at days 90 and 180. Diagnosis of NODAT and IFG was based on criteria specified by the ADA (7) and WHO (8). Treated diabetes was defined as receipt of any oral hypoglycemic medication or insulin for >14 days between day 15 and month 6. OGTT (75 g anhydrous glucose or 82.5 g glucose monohydrate) was performed at day 90 and days 180 following WHO guidelines (8) and results were used to calculate insulin release and sensitivity (22,23). GFR (glomerular filtration rate) was calculated by the Cockcroft–Gault formula (24).

Statistical analysis

Analyses were based on the intent-to-treat (ITT) population, which comprised all patients who were transplanted, randomized, received at least one dose of the randomized study medication and underwent at least one efficacy assessment, e.g. known survival status, after the first dose of study medication. Patients who discontinued the randomized study medication for whatever reason were censored at the time of discontinuation as defined in the analysis plan prior to data analysis. A sequentially ordered testing strategy was a priori defined such that the primary safety endpoint (NODAT or IFG) was tested first and as the result was statistically significant, the primary efficacy endpoint was also tested in a confirmative way without further adjustment of the significance level (25,26). The primary safety analysis was based on a superiority null hypothesis and the primary efficacy analysis on a noninferiority null hypothesis whereby CsA-ME was to be inferred noninferior to tacrolimus if the upper limit of the 95% confidence interval for the difference observed was less than the noninferiority margin of 10%. For the primary safety analysis, the Cochran–Mantel–Haenszel (CMH) test accounting for the two nondiabetic strata was used. Standard statistical methods were applied for further description of the data and comparisons of treatment groups, including chi-square test for categorical data and Wilcoxon rank sum test for numerical data. Efficacy and safety endpoints in study subpopulations are presented descriptively.

Choosing a power of 80% and a significance level (α) of 5% led to a necessary sample size of 270 nondiabetic patients per arm to test the safety endpoint and, with a noninferiority margin of 10%, to a sample size of 252 patients per arm to test the efficacy endpoint. With the enrollment of 80% non-diabetic patients, the adjusted number of patients per arm was 338. Hence the overall sample size required was 676.

An interim analysis was undertaken based on data from patients who had completed the day 90 visit by July 2004, in which data were blinded and results from the two groups pooled so that the integrity of the trial was not compromised.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendix

A total of 726 patients at 59 transplant centers in 16 countries were screened, out of which 682 patients represented the ITT population (CsA-ME 336, tacrolimus 346) (Figure 1). Of these, 115 were diabetic at baseline, 483 were non-diabetic/white and 84 were nondiabetic/nonwhite. There were 32 major protocol violations in the CsA-ME group and 38 in the tacrolimus group, of which the most frequent were patients with 0 HLA mismatches (n = 30), current or past malignancy against protocol (n = 12) and administration of CsA-ME or tacrolimus >24 h after graft reperfusion (n = 11). Two patients in the tacrolimus arm had an interruption of study medication >14 consecutive days. There were no significant differences in any demographic or baseline characteristic between the treatment groups overall (Table 1) or between treatment groups in the three predefined subpopulations (data not shown). Risk factors for NODAT were similar between the CsA-ME and tacrolimus groups in non-diabetic patients in each of the subpopulations (Table 2). Recruitment and first patient visits took place during October 2003 to March 2005, with the last 6-month visit occurring in September 2005.


Figure 1. Patient disposition. CsA-ME = cyclosporine microemulsion; ITT = intent-to-treat.

Download figure to PowerPoint

Table 1.  Demographics and baseline characteristics
 CsA-ME (n = 336)Tacrolimus (n = 346)
  1. All differences were nonsignificant.

  2. 1Delayed graft function was defined as the need for >1 dialysis session between graft reperfusion and day 7.

Recipient factors
 Mean age ± SD (years)47.0 ± 11.946.3 ± 13.6
Age group (years)
  <4098 (29.1%)115 (33.2%)
  40–54134 (39.8%)123 (35.5%)
  55—6482 (24.3%)76 (22.0%)
  >6523 (6.8%)32 (9.2%)
 Male gender241 (71.5%)225 (65.0%)
 Mean body mass25.6 ± 4.925.4 ± 4.7
 index ± SD (kg/m2)
Ethnic group
  White282 (83.7%)288 (83.2%)
  Black14 (4.2%)19 (5.5%)
  Asian12 (3.6%)13 (3.8%)
  Other29 (8.6%)26 (7.5%)
Cause of end-stage renal disease
  Glomerular disease102 (30.3%)101 (29.2%)
  Polycystic disease48 (14.2%)46 (13.3%)
  Diabetes mellitus46 (13.6%)40 (11.6%)
  Hypertension/nephrosclerosis30 (8.9%)34 (9.8%)
  Other/unknown111 (32.9%)125 (36.1%)
Previous transplant (%)
  0317 (94.1%)329 (95.1%)
  119 (5.6%)17 (4.9%)
  21 (0.3%)0
Dialysis at time of transplant
 None37 (11.0%)29 (8.4%)
 Hemodialysis (%)234 (69.4%)262 (75.7%)
 Peritoneal dialysis (%)66 (19.6%)55 (15.9%)
 Median duration of dialysis [range] (months)26.7 [12.6–51.9]27.2 [12.3–56.4]
Panel reactive antibodies
 Median (interquartile range)9.5 (3.5–26.5)6.5 (3–24)
Donor factors
 Mean age ± SD (years)45.2 ± 14.444.9 ± 14.6
Type of donor
  Deceased donor (%)234 (69.4%)230 (66.5%)
  Living-unrelated donor (%)36 (10.7%)28 (8.1%)
  Living-related donor (%)67 (19.9%)88 (25.4%)
Ethnic group
  White273 (81.0%)282 (81.5%)
  Black7 (2.1%)10 (2.9%)
  Asian8 (2.4%)7 (2.0%)
  Other13 (3.9%)13 (3.8%)
  Unknown36 (10.7%)34 (3.8%)
Transplant factors
 Mean cold ischemic time ± SD (h)11.8 ± 8.111.4 ± 8.2
 Mean number of HLA mismatches ± SD3.4 ± 1.443.4 ± 1.49
 Delayed graft function1 (%)65 (19.3%)69 (19.9%)
Table 2.  Risk factors for diabetes in the nondiabetic population
Nondiabetic, white (n = 483)CsA-ME (n = 238)Tacrolimus (n = 245)
  1. All differences were nonsignificant.

Mean age ± SD (years)46.7 ± 11.845.7 ± 13.5
Mean body mass index ± SD (kg/m2)25.1 ± 4.624.9 ± 4.2
White ethnicity238 (100%)245 (100%)
Hepatitis C positive1 (0.4%)2 (0.8%)
Family history of diabetes34 (14.3%)37 (15.2%)
Mean fasting plasma glucose ± SD (mmol/L)5.3 ± 1.15.1 ± 0.8
Fasting plasma glucose (mmol/L)
Mean C-peptide level ± SD (nmol/L)2.8 ± 1.72.8 ± 1.7
Mean HbA1c level ± SD (%)5.3 ± 0.535.3 ± 0.63
HbA1c >7%1 (0.5%)2 (0.9%)
Nondiabetic, nonwhiteCsA-METacrolimus
 (n = 84)(n = 43) (n = 41)
Mean age ± SD (years)42.5 ± 11.342.0 ± 13.5
Mean body mass index ± SD (kg/m2)25.0 ± 4.824.2 ± 5.1
Black10 (23.3%)15 (36.6%)
Asian11 (25.6%)9 (22%)
Other22 (51.2%)17 (41.5%)
Hepatitis C positive3 (7.0%)0
Family history of diabetes9 (21.4%)12 (30.0%)
Mean fasting plasma glucose ± SD (mmol/L)5.6 ± 0.75.3 ± 1.2
Fasting plasma glucose (mmol/L) 
Mean C-peptide level ± SD (nmol/L)2.6 ± 1.62.7 ± 2.5
Mean HbA1c level ± SD (%)5.3 ± 0.55.2 ± 0.55
HbA1c >7%00

Median CsA-ME C2 level reached target range at day 14 (1470 ng/mL, interquartile range 1112–1850 ng/mL), and remained at the lower end or below target range for the remainder of the study [month 1, 1496 ng/mL (1176–1788 ng/mL); month 3, 1100 ng/mL (832–1416 ng/mL); month 6, 885 ng/mL (666–1197 ng/mL)]. Median CsA-ME dose was 9.5 mg/kg/day (interquartile range 6.9–10.0 mg/kg/day) on day 2 posttransplant and gradually decreased to reach 3.6 mg/kg/day (3.0–4.6 mg/kg/day) during months 4–6. Median tacrolimus C0 reached target range by day 2 [median 11.4 ng/mL (6.5–18.1 ng/mL)] and remained within target range throughout the study [10.4 ng/mL (8.8–12.5 ng/mL) at month 3 and 8.5 ng/mL (7.0–11.0 ng/mL) at month 6]. Median tacrolimus dose on day 2 and during months 4–6 was 0.16 mg/kg/day (interquartile range 0.10–0.19 mg/kg/day), and 0.08 mg/kg/day (0.05–0.13 mg/kg/day), respectively. The median daily dose of MPA (with 720 mg EC-MPS considered bioequivalent to 1000 mg MMF) over the study period was 2000 mg/day (interquartile range 1456–2000 mg/day) in the CsA-ME group and 1389 mg/day (1000–2000 mg/day) in the tacrolimus arm. The median daily dose of steroids and interquartile ranges were similar in CsA-ME and tacrolimus patients throughout the study period overall (Figure 2): during weeks 3–4, the median daily dose was 20 mg/day (interquartile range 20–25 mg/day) in both cohorts; during month 3 it was 11 mg/day (10–15 mg/day) in the CsA-ME group and 10 mg/day (10–15 mg/day) in the tacrolimus group; and during months 4–6 it was 10 mg/day in both groups (CsA-ME, interquartile range 10–11 mg/day; tacrolimus, 10–14 mg/day). Steroid doses were also similar between treatment groups within each patient strata (data not shown).


Figure 2. Median daily dose of corticosteroids administered to month 6 posttransplant. Q1 and Q3 represent interquartile ranges.

Download figure to PowerPoint

The primary safety endpoint, NODAT or IFG within the first 6 months posttransplant among patients without pre-existing diabetes, occurred in 73/281 CsA-ME patients (26.0%) and 96/286 tacrolimus patients [33.6%, p = 0.046 (CMH test)]. Time to onset of hypoglycemic treatment is shown in Figure 3; the incidence of treated diabetes at 6 months was significantly lower in the CsA-ME group (25/281, 8.9%) than in the tacrolimus arm (48/286, 16.8%; p = 0.005) and among the patients who required treatment, only 1 patient in the CsA-ME group required a combination of insulin and oral treatment versus 10 patients in the tacrolimus group. For patients who had untreated diabetes at 3 months, 64.0% (16/25) in the CsA-ME group but only 35.3% (6/17) in the tacrolimus group showed an improvement in glycemic control by 6 months (with improvement defined as IFG, IGT or normal glucose level) (p = 0.067, chi-square test). IFG occurred in 6.8% (19/281) of the CsA-ME-treated patients versus 11.9% (34/286) in the tacrolimus group (p = 0.036). The difference observed in the whole population originated from differences observed in both subpopulations: Among nondiabetic white patients, NODAT or IFG developed in 59 CsA-ME patients (24.8%) versus 79 tacrolimus patients (32.2%); among nonwhites NODAT/IFG occurred in 14 patients (32.6%) and 16 patients (39.0%).


Figure 3. Kaplan–Meier plot of time to onset of first hypoglycemic treatment in the nondiabetic population.

Download figure to PowerPoint

Although the median cumulative dose of steroids received to 6 months did not differ between the CsA-ME group and the tacrolimus group (CsA-ME 3071 mg, tacrolimus 2815 mg; p = 0.070), it was significantly higher among CsA-ME patients who developed NODAT or IFG at 6 months versus those who did not (3320 mg vs. 2994 mg, p = 0.050). This difference was not, however, significant in the tacrolimus cohort (2955 mg with NODAT/IFG vs. 2690 mg without, p = 0.583). Patients who developed NODAT or IFG received significantly higher doses of steroids if they were on CsA-ME than tacrolimus (CsA-ME 3320 mg, tacrolimus 2690 mg; p = 0.006). Insulin secretion, sensitivity and disposition index in patients who developed NODAT but did not receive anti-diabetic medication are shown in Table 3. Compared to normoglycemic patients, insulin secretion was reduced in both treatment arms with a more pronounced reduction in the tacrolimus arm. Insulin sensitivity was also reduced in both arms but no obvious difference was observed between groups.

Table 3.  Insulin secretion, sensitivity and disposition index at 6 months among normoglycemic patients and patients with NODAT who were not receiving hypoglycemic medication
 Normoglycemic patients (n = 300)Untreated NODAT: CsA-ME (n = 29)Untreated NODAT: tacrolimus (n = 14)
  1. Values are shown as medians (interquartile range).

  2. 1Disposition index = (first and second phase insulin release) × insulin sensitivity.

Insulin secretion (pmol/L)
 Phase 11146 (899–1359)625 (279–920)314 (124–581)
 Phase 2304 (244–357)187 (103–246)116 (72–186)
Insulin sensitivity0.076 (0.065–0.089)0.043 (0.034–0.057)0.049 (0.016–0.064)
Disposition index1100 (82–127)26 (8–45)9 (−1, 22)

The primary efficacy endpoint, incidence of BPAR, graft loss or death at month 6, occurred in 43 CsA-ME patients (12.8%) and 34 tacrolimus patients (9.8%, p = 0.211) (Table 4). The Kaplan–Meier estimate of the incidence of the primary endpoint at month 6 was 13.8% with CsA-ME and 10.0% with tacrolimus. Noninferiority testing indicated that the 95% CI for this difference (3.8%, 95% CI −1.2% to 8.8%) did not exceed the pre-defined criterion for noninferiority (10%), i.e. CsA-ME was noninferior to tacrolimus for the primary efficacy endpoint. There were no significant differences in any efficacy endpoint (Table 4). In the diabetic, nondiabetic/white and non-diabetic/nonwhite subpopulations, the 6-month incidence of BPAR was 14.5% (8 patients), 9.2% (22 patients) and 9.3% (4 patients) with CsA-ME and 8.3% (5 patients), 6.9% (17 patients) and 4.9% (2 patients) with tacrolimus, respectively. In the CsA-ME arm, 20/34 episodes of rejection (59%) were graded mild (Grade IA or IB), 13/34 were graded moderate (Grade IIA) and 1/34 (3%) severe (Grade >2B) compared to 11/24 (46%), 7/24 (29%) and 6/24 (25%) in the tacrolimus group, respectively. The rate of recurrent BPAR was similar with CsA-ME and tacrolimus. Graft losses due to rejection were observed in three tacrolimus-treated patients (Table 4). One CsA-ME patient died due to respiratory insufficiency and subsequent failure and three died in the tacrolimus group due to cardiac arrest, myocardial infarction and invasive aspergillosis while they were on study medication. Seven additional deaths occurred in the CsA-ME group and five in the tacrolimus group after the patients had discontinued study medication, such that eight deaths in total occurred in each group.

Table 4.  Efficacy endpoints at month 6 posttransplant
 CsA-ME (n = 336)Tacrolimus (n = 346)p-Value
  1. 1Banff classification (21).

  2. 2An additional seven CsA-ME patients and five tacrolimus patients died after discontinuation of study medication.

  3. BPAR = biopsy-proven acute rejection.

BPAR, graft loss or death43 (12.8%)34 (9.8%)0.211
BPAR34 (10.1%)24 (6.9%)0.132
Severity of rejection1
 Grade IA159Not done
 Grade IB52 
 Grade IIA137 
 Grade IIB15 
 Grade III01 
Treated rejection50 (14.9%)38 (11%)0.128
Antibody-treated rejection11 (3.3%)12 (3.5%)0.888
Recurrent rejection4 (1.2%)3 (0.9%)Not done
Graft loss8 (2.4%)10 (2.9%)0.693
Graft loss due to rejection03 (0.9%)0.087
Death or graft loss9 (2.7%)13 (3.8%)0.439
Death1 (0.3%)23 (0.9%)20.338

At 6 months, mean GFR (Cockcroft–Gault) was not different at 6 months: tacrolimus 65.9 ± 23.1, CsA-ME 63.6 ± 20.7 (p = 0.285). Mean serum creatinine was 139 ± 58 μmol/L in the CsA-ME group and 133 ± 57 μmol/L among tacrolimus-treated patients (p = 0.005). The incidence of adverse events, serious adverse events, infections and serious infections is presented in Table 5. Mean blood pressure was similar between treatment groups at 6 months (Wilcoxon rank sum test) (Table 6). Median total cholesterol, LDL-cholesterol and triglycerides were higher at 6 months in the CsA-ME arm than in the tacrolimus arm. Between baseline and month 6, patients in the CsA-ME cohort showed an increase in both median LDL-cholesterol (LDL + 0.42 mmol/L, p<0.001) and HDL-cholesterol (+ 0.23 mmol/L, p < 0.001), but the ratio of total cholesterol/HDL remained unchanged (3.9 at month 6  vs. 4.0 at baseline) (Table 6).

Table 5.  Adverse events occurring in >15% of patients, or for which there was a significant difference in incidence between treatment groups, at month 6 posttransplant
 CsA-ME (n = 336)Tacrolimus (n = 346)p-Value
Any adverse event320 (95.2%)328 (94.8%)0.792
Any serious adverse event164 (48.8%)154 (44.5%)0.260
Any infection202 (60.1%)195 (56.4%)0.319
Any serious infection82 (24.4%)67 (19.4%)0.111
Any neoplasm7 (2.1%)7 (2.0%)0.956
Urinary tract infection102 (30.4%)93 (26.9%)0.315
Anemia81 (24.0%)79 (22.8%)0.695
Diarrhea55 (16.3%)95 (27.5%)<0.001
Nausea68 (20.2%)61 (17.6%)0.385
Constipation67 (19.9%)62 (17.9%)0.500
Peripheral edema78 (23.1%)50 (14.5%)0.003
Tremor50 (14.8%)75 (21.7%)0.022
Hypertension47 (13.9%)52 (15.0%)0.700
Any renal impairment78 (23.1%)82 (23.7%)0.881
CMV infection41 (12.2%)20 (5.8%)0.003
Hirsutism or hypertrichosis30 (8.9%)2 (0.6%)<0.001
Alopecia4 (1.2%)11 (3.2%)0.077
Table 6.  Blood pressure, cholesterol parameters and newly administered medication at baseline and at month 6 posttransplant
 CsA-ME (n = 336)Tacrolimus (n = 346)p-Value
  1. Blood pressure and lipid parameters are shown as median values and interquartile ranges; use of medication is shown as number of patients (%).

  2. 1Includes treatment administered at any point during the first 6 months posttransplant.

Blood pressure (mmHg)
  Systolic140 (128–160)140 (121–150)0.040
  Diastolic85 (77–90)80 (70–90)0.004
 Antihypertensive medication40 (11.9%)47 (13.6%)0.511
 Total cholesterol (mmol/L)4.6 (3.8–5.4)4.5 (3.8–5.4)0.930
 LDL-cholesterol (mmol/L)2.5 (1.9–3.2)2.4 (1.8–3.1)0.467
 HDL-cholesterol (mmol/L)1.1 (0.9–1.5)1.1 (0.9–1.4)0.274
 Total/HDL-cholesterol4.04.0Not done
 Triglycerides (mmol/L)1.7 (1.2–2.6)1.7 (1.1–2.6)0.769
 Lipid-lowering therapy68 (20.2%)84 (24.3%)0.205
6 months
Blood pressure (mmHg)
  Systolic130 (120–140)130 (120–140)0.101
  Diastolic80 (70–88)80 (70–85)0.153
 Antihypertensive medication1145 (43.2%)125 (36.1%)0.061
 Total cholesterol (mmol/L)5.3 (4.6–6.3)4.8 (4.0–5.5)<0.010
 LDL-cholesterol (mmol/L)3.0 (2.4–3.7)2.6 (2.1–3.2)<0.010
 HDL-cholesterol (mmol/L)1.4 (1.1–1.7)1.3 (1.1–1.6)0.067
 Total/HDL-cholesterol3.93.7Not done
 Triglycerides (mmol/L)1.9 (1.4–2.6)1.7 (1.2–2.3)0.003
 Lipid-lowering therapy1170 (50.6%)164 (47.4%)0.404

Seventy-one patients (21%) discontinued CsA-ME and 39 patients (11%) discontinued tacrolimus. The most frequent reason for study drug discontinuation was adverse events (CsA-ME 37, tacrolimus 18), including renal impairment (8 CsA-ME, 4 tacrolimus) and infections (5 CsA-ME, 4 tacrolimus). Eight patients in the CsA-ME group (2.4%) discontinued the drug due to rejection compared to three patients (0.9%) in the tacrolimus group.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendix

These results demonstrate that CsA-ME is associated with a significantly lower incidence of NODAT and IFG, and NODAT requiring treatment, than tacrolimus at 6 months posttransplant in renal transplant patients. This advantage was achieved with no loss of efficacy for CsA-ME versus tacrolimus, as measured by a composite endpoint of BPAR, graft loss or death.

The severity of glucose abnormalities may differ between CsA-ME and tacrolimus. A lower proportion of CsA-ME patients who became diabetic during the study required hypoglycemic medication and fewer of these required dual therapy with insulin and oral medication compared to those receiving tacrolimus. This is consistent with the fact that patients who developed diabetes under CsA-ME had lower HbA1c levels than those who developed diabetes while receiving tacrolimus, and also with the fact that insulin secretion was reduced to a greater extent with tacrolimus than CsA-ME in patients with NODAT. These findings could not be attributed to differences in diabetic risk factors at the time of transplantation. Neither could differences be accounted for by variations in steroid exposure, since the dose of steroids throughout the study was comparable in both treatment groups. Greater steroid exposure was associated with the development of glucose metabolism abnormalities in the CsA-ME group but not among tacrolimus-treated patients. This suggests that steroids played a major role in the development of diabetes occurring in the CsA-treated patients. With tacrolimus, a reduction in steroid dose does not appear to affect glycemic status, probably because tacrolimus impacts on insulin secretion (27).

Diabetes increases risk of ischemic heart disease after transplantation and related mortality by approximately three-fold (4,28). This disadvantage, however, must of course be balanced against any heightened cardiovascular risk associated with hypertension or dyslipidemia in patients receiving CsA-ME (29). In our population we found no difference in blood pressure or the incidence of reported hypertension at 6 months between the CsA-ME and tacrolimus arms. However, cholesterol, LDL-cholesterol and triglycerides were significantly higher at 6 months in the CsA-ME arm than in the tacrolimus arm.

CsA-ME showed noninferiority to tacrolimus in terms of the primary efficacy endpoint, BPAR, graft loss or death at 6 months. The incidence of BPAR was low in both arms (<10%) with no significant difference between agents. Numerically there were more mild episodes of BPAR in the CsA-ME group than the tacrolimus cohort, possibly due to the fact that exposure to CsA-ME was slightly lower than the recommended target levels. The observation that more rejections were graded severe (Grade >2B) in the tacrolimus group, and that graft losses due to rejection were only observed in the tacrolimus arm, is interesting. No obvious reason relating to patients' risk status was found to explain this observation. These findings are in contrast to the findings of a recent meta-analysis which found that CsA was associated with a higher incidence of acute rejection (15). This discrepancy may be due to several factors: C2 monitoring of CsA-ME in our trial compared to conventional trough monitoring, use of the microemulsion formulation of CsA, adoption of current target ranges for tacrolimus exposure (as opposed to higher ranges in early trials), and inclusion of MPA therapy and basiliximab in the regimen. There were few graft losses or deaths, and these were evenly distributed between treatment groups. Both groups achieved good renal function at 6 months. Serum creatinine was significantly higher in the CsA-ME cohort. GFR (Cockcroft–Gault) showed no significant difference between treatment groups although it was numerically higher in the tacrolimus group.

Despite a similar incidence and spectrum of adverse events, more patients discontinued CsA-ME than tacrolimus due to adverse events. The reason for this is unclear, but may be partly accounted for by an increased tendency among investigators to switch patients from CsA to tacrolimus in this open-label trial. The higher rate of CMV infection in the CsA-ME group was related neither to a difference in donor–recipient status nor to a difference in prophylaxis and the cause remains unclear; previous randomized trials of CsA-ME versus tacrolimus in renal (30) and liver (31) transplants have reported similar rates of CMV infection with each agent. It is interesting to note that although CMV infection is a risk factor for the development of diabetes posttransplantation (10) and was more frequent under CsA-ME, diabetes incidence was still higher in the tacrolimus arm. Numerically more cases of NODAT or IFG were observed among patients who experienced CMV infection post-transplant (39% vs. 29%) but the difference was not significant (p = 0.126).

The study had a number of strengths. It was adequately powered to detect differences between treatments, and involved a large number of centers such that center bias was minimized. We systematically evaluated glucose metabolic status according to WHO/ADA guidelines, as has been recommended previously in transplant recipients (32). Immunosuppression other than the type of calcineurin inhibitor, including steroid exposure, was consistent between groups and the difference in MPA dose was consistent with known pharmacokinetic variation in MPA exposure with concomitant administration of CsA-ME or tacrolimus (33). While mean CsA Clevel was below or toward the lower end of target ranges, possibly due to investigator concerns about over-exposure, this does not appear to have impaired efficacy. The study did, however, include some limitations. We were not able to collect OGTT results pretransplant due to practical reasons, raising the possibility that some patients may have been misclassified as nondiabetic at baseline. This was partially addressed by measurement of baseline fasting blood glucose and HbA1c. While some nondiabetic patients had FPG >7 mmol/L at baseline, only three had an HbA1c >7%. It is likely that these high FPG values in the nondiabetic group may not in fact have true FPG but random plasma glucose values, which is not uncommon in the hours immediately prior to transplantation. None of the three patients with HbA1c >7% developed NODAT or IFG. However, using a more sensitive marker, HbA1c >6% (9) (i.e. outside the normal range), 14 (6%) patients classified as nondiabetic had HbA1c >6.1% in the CsA-ME group compared to 17 (7%) in the tacrolimus group (n.s.). None of these patients in the CsA-ME group progressed to NODAT or IFG but in the tacrolimus cohort five developed NODAT and three developed IFG. The study used an open-label design with the consequent potential for bias due to investigators' preferences, for example, toward more discontinuation of CsA-ME following acute rejection or adverse events. Finally, follow-up was relatively short (6 months), although the majority of cases of NODAT are known to develop within the first 3–6 months post-transplant (16).

In conclusion, the results of the DIRECT study show that the risk of NODAT or IFG is common in calcineurin inhibitors based regimens but is significantly lower with CsA-ME versus tacrolimus in the first 6 months posttransplant.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendix

The DIRECT Study was funded by Novartis Pharma AG. We have the following conflicts of interest to declare with regards to Novartis concerning activity within the last 3 years: SF, ES, LR, MDP, PM and SC are members of Novartis Advisory Boards; FV, ES, LR, TJ, MDP, PM, AW, SC and ME-S have received travel grants and/or honoraria.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendix
  • 1
    Kasiske BL, Snyder JJ, Gilbertson D, Matas AJ. Diabetes mellitus after kidney transplantation in the United States. Am J Transplant 2003; 3: 178185.
  • 2
    Davidson J, Wilkinson A, Dantal J et al. New-onset diabetes after transplantation: 2003 international consensus guidelines. Transplantation 2003; 75(10 Suppl): SS3SS24.
  • 3
    Miles AM, Sumrani N, Horowitz R et al. Diabetes mellitus after renal transplantation: As deleterious as non-transplant associated diabetes? Transplantation 1998; 65: 380384.
  • 4
    Kasiske BL, Chakkera HA, Roel J. Explained and unexplained ischemic heart disease risk after renal transplantation. J Am Soc Nephrol 2000; 11: 17351743.
  • 5
    Balkau B, Hu G, Qiao Q, Tuomilehto J, Borch-Johnsen K, Pyorala K, DECODE Study Group; European Diabetes Epidemiology Group. Prediction of the risk of cardiovascular mortality using a score that includes glucose as a risk factor. The DECODE Study. Diabetologia 2004; 47: 2118–2128.
  • 6
    Cosio FG, Kudva Y, Van Der Velde M et al. New onset hyperglycemia and diabetes are associated with increased cardiovascular risk after kidney transplantation. Kidney Int 2005; 67: 24152421.
  • 7
    American Diabetes Association. Report of the Expert Committee on the diagnosis and classification of diabetes mellitus. Diabetes Care 2003; 26: 520.
  • 8
    World Health Organization. Diabetes. Available at: http://www.who./int/ncd/dia/databases4.htm. Accessed 31 January 2006.
  • 9
    Hoban R, Gielda B, Temkit M et al. Utility of HbA1 c in the detection of subclinical post renal transplant diabetes. Transplantation 2006; 81: 379383.
  • 10
    Hjelmesaeth J, Sagedal S, Hartmann A et al. Asymptomatic cytomegalovirus infection is associated with increased risk of new-onset diabetes mellitus and impaired insulin release after renal transplantation. Diabetologia 2004; 47: 15501556.
  • 11
    Redmon JB, Olson LK, Armstrong MB, Greene MJ, Robertson MP. Effects of tacrolimus (FK506) on human insulin gene expression, insulin mRNA levels, and insulin secretion in HIT-T15 cells. J Clin Invest 1996; 98: 27862793.
  • 12
    Drachenberg CB, Klassen DK, Weir MR et al. Islet cell damage associated with tacrolimus and cyclosporine: Morphological features in pancreas allograft biopsies and clinical correlation. Transplantation 1999; 68: 396402.
  • 13
    Marchetti P, Navalesi R. The metabolic effects of cyclosporin and tacrolimus. J Endocrinol Invest 2000; 23: 482490.
  • 14
    Heisel O, Heisel R, Balshaw R, Keown P. New onset diabetes mellitus in patients receiving calcineurin inhibitors: A systematic review and meta-analysis. Am J Transplant 2004; 4: 583595.
  • 15
    Webster A, Woodroffe RC, Taylor RS, Chapman JR, Craig JC. Tacrolimus versus cyclosporin as primary immunosuppression for kidney transplant recipients. Br Med J 2005; 331: 810814.
  • 16
    Woodward RS, Schnitzler MA, Baty J et al. Incidence and cost of new onset diabetes mellitus among U.S. wait-listed and transplanted renal allograft recipients. Am J Transplant 2003; 3: 590598.
  • 17
    Vincenti F, Rostaing L. Rationale and design of the DIRECT study: A comparative assessment of the hyperglycemic effects of tacrolimus and cyclosporine following renal transplantation. Contemp Clin Trials 2005; 26: 1724.
  • 18
    Zucker K, Rosen A, Tsaroucha A et al. Unexpected augmentation of mycophenolic acid pharmacokinetics in renal transplant patients receiving tacrolimus and mycophenolate mofetil in combination therapy, and analogous in vitro findings. Transplant Immunol 1997; 5: 225232.
  • 19
    Van Gelder T, Klupp J, Barten MJ, Christians U, Morris RE. Comparison of the effects of tacrolimus and cyclosporine on the pharmacokinetics of mycophenolic acid. Ther Drug Monit 2001; 23: 119128.
  • 20
    Pou L, Brunet M, Cantarell C et al. Mycophenolic acid plasma concentrations: Influence of comedication. Ther Drug Monit 2001; 23: 3538.
  • 21
    Racusen LC, Solez K, Colvin RB et al. The Banff 97 working classification of renal allograft pathology. Kidney Int 1999; 55: 713723.
  • 22
    Stumvoll M, Mitrakou A, Pimenta W et al. Use of the oral glucose tolerance test to assess insulin release and insulin sensitivity. Diabetes Care 2000; 23: 295301.
  • 23
    Hjelmesaeth J, Midtvedt K, Jenssen T, Hartmann A. Insulin resistance after renal transplantation: Impact of immunosuppressive and antihypertensive therapy. Diabetes Care 2001; 24: 21212126.
  • 24
    Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976; 16: 3141.
  • 25
    Bauer P, Rohmel J, Maurer W, Hothorn L. Testing strategies in multi-dose experiments including active control. Stat Med 1998; 17: 21332146.
  • 26
    The European Agency for the Evaluation of Medicinal Products (EMEA), Committee for Proprietary Medicinal Products (CPMP). Points to consider on multiplicity issues in clinical trials, London (2002)
  • 27
    Van Duijnhoven EM, Boots JM, Christiaans MH, Van Hooff JP. Metabolic aspects of tacrolimus in renal transplantation. Consequences for the choice of an immunosuppressive regimen and for the management of post-transplant diabetes mellitus. Minerva Urol Nefrol 2003; 55: 3342.
  • 28
    Lindholm A, Albrechtsen D, Frodin L, Tufveson G, Persson NH, Lundgren G. Ischemic heart disease—Major cause of death and graft loss after renal transplantation in Scandinavia. Transplantation 1995; 60: 451457.
  • 29
    Jardine A. Assessing the relative risk of cardiovascular disease among renal transplant patients receiving tacrolimus or cyclosporine. Transplant Int 2005; 18: 379384.
  • 30
    Margreiter R for the European Tacrolimus vs Ciclosporin Microemulsion Renal Transplantation Study Group. Efficacy and safety of tacrolimus compared with ciclosporin microemulsion in renal transplantation: A randomised multicentre study. Lancet 2002; 359: 741746.
  • 31
    Levy G, Villamil F, Samuel D et al. Results of LIS2 T, a multicenter, randomized study comparing cyclosporine microemulsion with C2 monitoring and tacrolimus with C0 monitoring in de novo liver transplantation. Transplantation 2004; 77: 16321638.
  • 32
    Wilkinson A, Davison J, Dotta F et al. Guidelines for the treatment and management of new-onset diabetes after transplantation. Clin Transplant 2005; 19: 291298.
  • 33
    Kaplan B, Meier-Kriesche H-U, Minnik P et al. Randomized calcineurin inhibitor cross over study to measure the pharmacokinetics of co-administered enteric-coated mycophenolate sodium. Clin Transplant 2005; 19: 551558.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendix

Appendix The DIRECT (Diabetes incidence after renal transplantation: Neoral® Cmonitoring versus tacrolimus) study group

Juan Jose Almenabar, Hospital de Cruces, Barakaldo, Spain; Amado Andres, Hospital 12 de Octubre, Madrid, Spain; Lászió Asztaios, University of Debrecen, Debrecen, Hungary; William Bennett, Good Samaritan Hospital, Portland, USA; François Berthoux, Hôpital Nord, Saint Priest en Jarez, France; Roy Bloom, University of Pennsylvania Medical Center, Philadelphia, USA; Kenneth L. Brayman, University of Virginia Health System, Charlottesville, USA; Laura Buist, Glasgow Western Infirmary, Glasgow, UK; Jesus Bustamante, Hospital Universitario de Valladolid, Valladolid, Spain; Josep Campistol, Hospital Clinic de Barcelona, Barcelona, Spain; Carl Cardella, Toronto General Hospital, Toronto, Canada; M. Castagneto, Istituto di Clinica Chirurgica Centro Trapianti Policlinico Gemelli, Rome, Italy; Domingo del Castillo, Hospital Reina Sofia de Cordoba, Cordoba, Spain; Arun Chandrakantra, University of Alabama at Birmingham, Birmingham, USA; Dr Cotterell, VCU Health Systems, Medical College of Virginia, Richmond, USA; Mohammed El-Shahawy, University of Southern California Kidney Transplant Program, Los Angeles, USA; Josette Eris/Steve Chadban, Royal Prince Alfred Hospital, Camperdown, Australia; Pedro Errasti, Clinica Universitari de Navarra, Pamplona, Spain; A. Famulari, Centro Trapianti di Rene-Ospedale Civile S. Salvatore, Aquila, Italy; Styrbjörn Friman, Enheten för transplantation och leverkirurgi, Gothenburg, Sweden; Jorge Garces, Ochsner Clinic Foundation Kidney Transplant Program, New Orleans, USA; Reginald Gohh, Rhode Island Hospital, Providence, USA; Peter Gross, Medizinische Klinik III/Nephrologie, Dresden, Germany; Marcus Hart, University of California-San Diego, San Diego, USA; Andrew House, London Health Sciences Center, London, ON, Canada; Ashley Irish, Renal Unit Royal Perth Hospital, Perth, Australia; Jeno Jarai, Semmelweis University, Budapest, Hungary; Trond Jenssen, Rikshospitalet, Oslo, Norway; Dr Johnston, University of Kentucky Transplant Center, Lexington, USA; Anil Kapoor, McMaster University, St Joseph's Healthcare, Urology Institute, Hamilton, Canada; Marian Klinger, Klinika Nefrologii, Wroclaw, Poland; Gregory Knoll, The Ottowa Hospital, Ottowa, Canada; Ricardo Lauzurica, Hospital Universitari Trias I Pujol, Badalona, Spain; Christophe Legendre, Hôpital Necker, Paris, France; Jimmy Light, Washington Hospital Center, Washington, USA; Arnost Martinek, Faculty Hospital Ostrava, Ostrava, Czech Republic; Robert Mendez, St Vincents MC/NIT, Los Angeles, USA; Hans-H Neumayer, Medizinische Klinik mit Schwerpunkt Nephrologie, Berlin, Germany; Barbara Nonnast-Daniel, Medizinische Klinik IV, Erlangen, Germany; Leszek Paczek, Klinika Immunologii, Transplantologii I Chorob Wewnetrznych, Warsaw, Poland; Ravi Parasuraman, Henry Ford Hospital, Detroit, USA; Mark D. Pescovitz, Indiana University, Indianapolis, USA; Thomas Pearson, Emory University Hospital, Atlanta, USA; Lionel Rostaing, Hôpital Rangueil, Toulouse, France; Graeme R. Russ, The Queen Elizabeth Hospital, Woodville, Australia; Ernst Scheuermann, Medizinische Klinik IV, Frankfurt, Germany; G. Segoloni, Divisione di Nefrologia Dialisi e Trapianti Presidio Molinette, Torino, Italy; Craig Shadur, Iowa Methodist Medical Center, Iowa, USA; Jean Paul Soulilou, Hôpital Hôtel Dieu, Nantes, France; V. Sparacino, Unita Medica di Trapianto, Ospedale Civico G. di Cristina M. Ascoli, Palermo, Italy; Gunter Stein, Klinik für Innere Medizin III, Universitätsklinik Jena, Jena, Germany; Pal Szenohradszy, University of Szeged, Szeged, Hungary; Jean Tchervenkov, Royal Victoria Hospital, Montreal, Canada; Richard Thistlethwaite, University of Chicago, Chicago, USA; Murat Tuncer, Akdeniz U. Medical School, Antalya, Turkey; Aydin Turkmen, Istanbul Medical School, Istanbul, Turkey; Kazuharu Uchida, Nagoya Daini Red Cross Hospital, Nagoya City, Japan; Flavio Vincenti, UCSF Kidney Transplant Service, San Francisco, USA; Andrzej Wiecek, Klinika Nefrologii Immunologii Transplantologii I Chorob Wewnetrznych, Warsaw, Poland.