Glucose Metabolism before and after Conversion from Cyclosporine Microemulsion to Tacrolimus in Stable Renal Recipients

Background. Tacrolimus is more diabetogenic than cyclosporine. However, this difference is only discernible in the first few months after renal transplantation. In randomized trials, investigating the effects of immunosuppression after renal transplantation, no increase in diabetes mellitus has been reported. However, no sensitive technique was used in these trials, so subclinical alteration of glucose metabolism cannot be excluded. Methods. We, therefore, decided to use an intravenous glucose tolerance test (IV-GTT), to investigate whether conversion from cyclosporine-based immunosuppres-sion, with a median trough level of 120 mg/l, to tacrolimus-based immunosuppression with a median trough level of 6.5 mg/l influences glucose metabolism and whether patients on steroids behave differently from those not on steroids. Results. Thirty stable, non-diabetic patients, transplanted 10 or more years earlier, were converted from cyclosporine to tacrolimus without changing their concomitant medication. IV-GTT's were performed before and 2.5 months after the conversion. Before conversion, 40% of the patients had an abnormal glucose disappearance rate (kG): in 7%, kG was below 0.8 (abnormal range) and in 34%, kG was between 0.8 and 1.2 (indeterminate range). After conversion, stimulated insulin production, kG, HbA1C and fasting glucose did not change significantly. Insulin resistance (HOMA-R) of the whole group increased significantly, mainly due to a rise in HOMA-R in patients on steroids (n ¼ 18). None of these patients developed overt diabetes mellitus. Conclusions. Some 40% of long-term cyclosporine-treated patients had an abnormal glucose metabolism. Conversion from cyclosporine to tacrolimus does not negatively influence stimulated glucose metabolism or insulin resistance in stable, steroid-free renal transplant recipients. However, in patients receiving steroids, conversion leads to an increase in insulin resistance while insulin output remains the same.


Introduction
Tacrolimus is more diabetogenic than cyclosporine.However, this difference is only discernible in the first few months after renal transplantation, when the tacrolimus dose is high and combined with corticosteroids.Thereafter, and in steroid-free regimens, there is no difference [1,2].
Many patients are converted from cyclosporine to tacrolimus either for cosmetic reasons, lack of efficacy, cardiovascular side effects or decreased renal function [3][4][5][6].
Our centre's policy is to convert renal recipient patients to tacrolimus because of their more stable renal function and cardiovascular profile after conversion [6].
To date, in randomized trials investigating the effects of immunosuppression several months after renal transplantation, no increase in diabetes mellitus has been reported [5,6].However, no sensitive technique has been used in any of these trials, so subclinical alteration of glucose metabolism cannot be excluded.Intravenous glucose tolerance testing (IV-GTT) is a sensitive technique that can detect subclinical glucose metabolism disorders and discern between increased resistance and impaired insulin production.Using this technique, we have shown that tacrolimus at high trough levels reduces insulin production, whereas steroids increase insulin resistance [1,[7][8][9].In a comparative cohort study of patients on cyclosporine or tacrolimus, a difference in insulin production was seen in the first weeks after transplantation.However this difference disappeared after lowering the trough levels of tacrolimus [8].Because there are no data on changes in insulin resistance and production by conversion from cyclosporine to tacrolimus, we investigated by using IV-GTT whether conversion from cyclosporine-based to tacrolimus-based immunosuppression influences glucose metabolism and whether patients on steroids behave differently from those who are not on steroids.

Study design
In a single-centre crossover study, stable renal transplant patients on cyclosporine microemulsion were converted to tacrolimus without changing their concomitant medication.
Before and after conversion, glucose metabolism measurements were taken in a fasting state.After the first IV-GTT, cyclosporine was stopped; the next day, the patients were started on tacrolimus, 0.2 mg/kg per day orally in two doses.Twice a week the trough level was measured and the doses were adjusted to reach a target trough level between 5 and 7 mg/l.After the level had been in the target range for 1 month, a second IV-GTT was done.
The study was approved by the local Ethics Committee and all patients gave written informed consent.

Patients
Patients were eligible to participate in the study if they met the following criteria: age 18 years or older, transplanted more than 1 year ago, stable renal function for more than 6 months, no other organ transplantation, and on cyclosporine microemulsion-based immunosuppression.Patients receiving any drug known to interfere with tacrolimus or cyclosporine or patients with a rejection episode requiring antibody therapy in the previous 6 months were excluded.All recipients of renal grafts who were using oral antidiabetic tablets or insulin were also excluded.

Glucose metabolism
Glucose metabolism was studied using IV-GTT.The tests were performed in the morning, after a 12-h overnight fast.Glucose 50% (0.5 g/kg body weight, maximum 75 g glucose) was administered intravenously for 2-3 min.Venous blood samples for measuring fasting glucose, C-peptide and insulin were taken from the opposite arm at t ¼ À15, 0, 5, 10, 15, 20, 30, 40, 50 and 60 min.The insulin sensitivity index (glucose disappearance rate, kG) was calculated as described earlier [7].A kG value below 0.8% per minute was considered to be abnormal, between 0.8 and 1.2% per minute as indeterminate, and above 1.2% per minute as normal.C-peptide and insulin secretion, i.e. the secretion response to a glucose load, were calculated as the area under the curve using a linear trapezoidal technique from the serum value at each time point (total secretion) and also after subtraction of the t ¼ 0 value (increment in secretion).Insulin resistance (HOMA-R) was calculated using the homoeostasis model assessment [10].Glycosylated haemoglobin (HbA1c) was measured at the time of the IV-GTTs.

Renal function, lipids and blood pressure
Renal function was calculated from age, race, gender, serum creatinine, albumin and BUN according to the MDRD7 formula [11].Total cholesterol, HDL-cholesterol, LDL-cholesterol and triglycerides were measured in plasma during a fasting state.The 24-h blood pressure measurements were registered with an A&D medical UA-767 digital blood pressure monitor (A&D Company Ltd, Tokyo).

Statistics
Statistical analysis was performed using the SPSS package, version 12.0.To compare basal and stimulated glucose metabolism, renal function, blood pressure and lipids, the paired Wilcoxon signed rank test was used.Unless indicated otherwise, data are given as median values.To compare waist/hip ratio with HOMA-R, logistic regression was used.

Patients
Thirty-three patients were initially included in the study.However, three withdrew their consent after the first IV-GTT, leaving 30 to complete the study.The median age of the patients was 50.4 years and 21 patients were male.Twenty-nine patients were Caucasian and one was Asian.Their median body mass index (BMI) was 25.1 and their waist/hip ratio was 0.90.In contrast to expectation, the waist/hip ratio was lower in steroid-treated patients.Thirteen percent of the patients were smokers.
Eighteen patients (60%) used steroids: 15 used 5 mg daily and 3 patients 10 mg.In the other 12 patients, steroids were withdrawn during the first year after renal transplantation according to centre policy; in none of these patients were steroids withdrawn because of glucose abnormalities.The median time after transplantation was 10.1 years.The median time between conversion and the second stimulated glucose metabolism was 72 days (Table 1).Before conversion, the cyclosporine trough level was 120 mg/l (range 50-160 mg/l).After conversion, the median tacrolimus trough level was 6.4 mg/l (range 4.2-9.5 mg/l).After conversion, none of the patients experienced a rejection episode and there were no deaths.

All patients
Basal parameters.After conversion, in the total group of 30 patients, there was a tendency towards a 6% increase in median fasting glucose, from 5.0 to 5.3 mmol/l (P ¼ 0.18), while HbA1c remained stable (from 5.5 to 5.6%, P ¼ 0.48).There was a significant increase (23%) in median fasting insulin from 7.65 to 9.90 mU/l (P ¼ 0.04) as well as a significant increase (23%) in HOMA-R, from 1.66 to 2.16 mol/lÃmU/l (P ¼ 0.04; Table 2, Panel A).Glucose metabolism and conversion from cyclosporine to tacrolimus Stimulated parameters.Neither the insulin sensitivity index (kG) nor the insulin output (total and increment) changed significantly after conversion (Table 2, Panel A).
Before conversion, 12 of the clinically non-diabetic patients had a kG below 1.20% per minute (40%; Table 3): 10 had an indeterminate kG between 0.8 and 1.2% per minute and 2 had an abnormal kG below 0.8% per minute.After conversion, 14 patients had a kG below 1.20 (10 an indeterminate kG and 4 an abnormal kG).
The patients without steroids had a higher fasting glucose level at the beginning of the study.Logistic regression analysis with fasting glucose as the dependent parameter and steroid use, age, gender and waist/ hip ratio as independent risk factors showed that only waist/hip ratio was related to a higher fasting glucose.Stimulated parameters.Both groups were comparable for baseline kG and insulin output (total and increment) during cyclosporine use.After conversion, there was no change in these parameters in either group (Table 2, Panel B).

Renal function, lipids and blood pressure
After conversion, renal function and lipids improved significantly: creatinine clearance (18% improvement) went from 38 to 45 ml/min (P ¼ 0.001), total cholesterol (13% improvement) from 5.3 to 4.7 (P 0.001), LDL-cholesterol (10% improvement) from 3.3 to 2.9 (P 0.001) and triglycerides (12% improvement) from 1.62 to 1.45 (P ¼ 0.009).Lipids improved independent of the use of statins.When the groups with and without steroids were analysed separately, total cholesterol and LDL-cholesterol improved significantly in both groups.Triglycerides improved significantly only in the group without steroids (P ¼ 0.02).Renal function improved significantly in both groups.There was no significant change in 24-h blood pressure.Total, day, and night mean arterial pressure (MAP); as well as systole and diastole remained unaltered before and after conversion (Table 4).

Discussion
In this study, we compared the stimulated glucose metabolism during cyclosporine and tacrolimus immunosuppression in renal transplant recipients.The fact that none of the patients experienced a rejection episode after conversion from cyclosporine to tacrolimus shows that the conversion was safe.
The significant improvement in lipids and in renal function after conversion is a confirmation of our previous findings.Notwithstanding the use of cyclosporine for a prolonged period of time (median > 10 years), an increase in renal function of almost 20% was observed, the same magnitude as the difference in renal function at 2 years between cyclosporine and sirolimus in the study of Kreis et al. [12].However, in contrast to previous studies, blood pressure did not change 3 months after conversion.The main difference between the present and the previous study was that patients in the present study were using cyclosporine for a much longer period of time (10 years vs 6 years).It would appear that after such a long period, blood pressure is fixed.
A remarkable finding in our study was the high incidence (40%) of unknown glucose metabolism disorders revealed using a sensitive test in transplant recipients treated with cyclosporine.Undetected glucose metabolism disorders are a frequent event.For example, in the Hoorn study, a study in a normal population of 50-75 years old without previously known diabetes [13], 5% had diabetes and 12% had impaired fasting glucose according to ADA-criteria.In the Transcend study, an incidence of previously unknown glucose disorder of up to 30% was found with an oral glucose tolerance test (OGTT) in patients older than 55 years with an increased cardiovascular risk [14] and in a study of patients undergoing an elective coronary angiography without a history of DM, impaired glucose regulation was diagnosed in 40.4%, and DM in 22.7% [15].Thus, in the general  population, but especially in elderly patients with a cardiovascular risk, there is a high incidence of abnormal glucose metabolism.For this population, screening for a pre-diabetic state is advocated because there is a high risk of cardiovascular disease.Lifestyle intervention, lipid-lowering therapy, and optimization of blood pressure can improve prognosis considerably [1].
Although unproven, it is very likely that both screening and intervention are also very useful for transplant patients.Renal transplant patients have a very elevated risk of developing cardiovascular disease.In fact, one of the major causes of graft loss is the death of the recipient due to cardiovascular disease.
Before conversion, the patients had a median cyclosporine level of 120 mg/l.In 10 of these patients we measured the C2 level of cyclosporine.The median C2 level was 670 mg/l.
After conversion, with the tacrolimus trough level of 6.5 mg/l, there is no reduction in insulin secretion.This is a confirmation of a previous finding in a smaller group of patients earlier after transplantation [8].After conversion, our total study group showed a small, but not significant, increase in fasting glucose and a significant increase in insulin resistance (HOMA-R), without differences in insulin sensitivity (kG) or insulin output.
As a consequence of increases in insulin resistance and fasting glucose, and stable insulin secretion, one might expect that the insulin sensitivity index decreases.This does not happen, leading to the conclusion that there is probably a surplus in insulin production during IV-GTT also in patients with low trough level of tacrolimus.
When the groups with and without steroids were analysed separately, fasting glucose and HOMA-R increased significantly after conversion, but only in the patients on steroids.Thus, conversion from cyclosporine to tacrolimus in patients not on steroids can be done to obtain a better cardiovascular risk profile and without an additional risk for diabetes.For patients taking steroids, there is a significant rise in insulin resistance.These findings are compatible with the clinical observation that tacrolimus taken concomitantly with steroids is diabetogenic while tacrolimus without steroids is not [1].The increase in HOMA-R in the tacrolimusþsteroids group after transplantation might be explained in two ways.First, it can be due to the steroid mimetic effect of tacrolimus.Tacrolimus binds to FK506-binding protein (FKBP), predominantly FKBP-12.Another immunophilin, FKBP-52 is associated with the cytoplasmic glucocorticoid (GC) receptor complex.When cells are exposed to glucocorticoids, the steroid binds to the GC receptor and liberates it from the complex.By binding to FKBP-52 in the GC receptor complex, tacrolimus may alter the affinity of interactions and either cause a release of the GC receptor at lower steroid concentrations, a steroidsparing effect, or it may free the GC receptor I absence of steroids [16].Second, tacrolimus increases the bioavailability of steroids.
The groups of patients in this study with and without maintenance steroids differed in various aspects.Our centre has a policy of tailor-made immunosuppression [17].In low-and intermediaterisk patients without an acute rejection, steroids are stopped within the first year after transplantation, while all other patients are kept on 5 mg of prednisone.Therefore, it is understandable that the group of steroid-free patients is older and contains more males than the group of patients using steroids.
In contrast to what one would expect, the waist/hip ratio was higher in steroid-free patients.This paradox is probably due to the higher age of the patients without the steroids.It is well known that body fat (and waist/hip ratio) increases with age.
There was also a significant difference in the number of patients with an increased fasting glucose between the steroid group and the steroid-free group ( 2 -test, P ¼ 0.004).Both groups were comparable for baseline parameters, except for a higher fasting glucose level in patients without corticosteroids.Logistic regression analysis with fasting glucose as the dependent Glucose metabolism and conversion from cyclosporine to tacrolimus parameter and steroid use, age, gender and waist/hip ratio as independent risk factors showed that only waist/hip ratio was related to a higher fasting glucose.Patients without steroids had a significantly higher waist/hip ratio.

Conclusions
Some 40% of long-term cyclosporine-treated patients not known with diabetes mellitus have an abnormal glucose metabolism.Conversion from cyclosporine to tacrolimus does not negatively influence stimulated glucose metabolism or insulin resistance in stable, steroid-free renal transplant recipients.However, in patients receiving steroids, conversion leads to an increase in insulin resistance while insulin output remains the same.

Fig. 1 .
Fig. 1.Change in HOMA-R after conversion; with and without steroids.

Table 2 .
Basal and stimulated parameters glucose metabolism before and after conversion

Table 4 .
Lipids before and after conversion a a All results are in median values.