Abstract
Abstract. Most studies concerning the influence of tacrolimus on glucose metabolism have been performed either in animals or after organ transplantation. These clinical studies have largely been transversal with patients who were using steroids. Therefore, this prospective, longitudinal study investigated the influence of tacrolimus on glucose metabolism before and after transplantation. Eighteen Caucasian dialysis patients underwent an intravenous glucose tolerance test before and 5 d after the start of tacrolimus. Insulin sensitivity index (kG), insulin resistance (insulin/glucose ratio and homeostasis model assessment), and C-peptide and insulin secretion were calculated. Trough levels of tacrolimus were measured. After transplantation, the occurrence of posttransplantation diabetes mellitus (PTDM) was prospectively monitored. Statistical analysis was performed using the Wilcoxon signed ranks test and Spearman's rho for correlation. Before tacrolimus, kG was indeterminate in three patients. During tacrolimus, kG decreased in 16 of 18 patients, from a median of 1.74 mmol/L per min to 1.08 mmol/L per min (P < 0.0001). The correlation between C-peptide and insulin data was excellent. Insulin secretion decreased from 851.0 mU × min/L to 558.0 mU × min/L (P = 0.014), whereas insulin resistance did not change. Insulin sensitivity correlated negatively with tacrolimus trough level. After transplantation, three patients developed PTDM; before tacrolimus, two had an indeterminate and one a low normal kG. During tacrolimus administration, kG decreased in almost all patients as a result of a diminished insulin secretion response to a glucose load, whereas insulin resistance did not change. Patients with an abnormal or indeterminate kG seem to be at risk of developing PTDM while on tacrolimus.
Tacrolimus is a macrolide with potent immunosuppressive effects. It has been used successfully as a prophylactic immunosuppressant for solid organ transplantation. After transplantation, a high incidence of posttransplantation diabetes mellitus (PTDM) has been reported. However, systematic, prospective data regarding the incidence of PTDM in renal allograft recipients treated with tacrolimus are scarce. Studies report an incidence of 8 to 36% (1,2,3,4,5,6,7,8,9,10,11). This variation may be explained partly by different definitions of PTDM that are used. Moreover, it has been reported that the incidence is higher in older patients, that it is influenced by both steroid and tacrolimus dosage, and that it differs according to race. Thus, the incidence of PTDM depends in part on the demographics of the patient group studied. The development of PTDM has important sequelae. Miles et al. (12) studied the long-term outcome of 40 PTDM patients who were using cyclosporine. Although they failed to show a significant difference in mortality rate after 12 yr, they did find an increased frequency of sepsis as a cause of death and a greater risk of developing graft failure (relative risk = 3.72; P = 0.04).
The mechanism responsible for the development of PTDM has been examined both in animals and in clinical studies. An impaired insulin secretion during tacrolimus administration was found in the former. In clinical studies, an increase in insulin resistance was suggested as an additional mechanism underlying the development of PTDM (1,13). However, in the latter studies, all patients were taking steroids, which can cause an increase in insulin resistance.
In rats, Tamura et al. (14) found a defect in insulin synthesis as a result of an mRNA transcriptional defect dependent on duration of exposure to tacrolimus. They hypothesized that the defect in mRNA transcription induced by tacrolimus was mediated by binding to FK506-binding protein-12 and that the subsequent inhibition of calcineurin in β cells caused the diminished insulin secretion. Others have described a reduced insulin secretion in human pancreatic β cells transplanted into animals (15,16). In a transversal study in 14 liver transplant recipients, Fernandez et al. (17) found a reduced insulin secretion in patients who were using tacrolimus and cyclosporine compared with a healthy control group. This study was performed in liver transplant recipients who were being treated with steroids (as well as other medication) and who were examined after transplantation.
In the present study, insulin secretion and insulin resistance were assessed before and during the use of tacrolimus in patients who were awaiting renal transplantation. The patients did not receive steroids. The patients were also evaluated after transplantation for the development of PTDM. To our knowledge, this is the first prospective study performed before transplantation, with follow-up after transplantation, concerning the influence of tacrolimus on glucose metabolism. In an interim analysis of nine patients from this study (18), it was shown that both the insulin sensitivity index (kG) and insulin secretion decreased after tacrolimus. In this definitive report of 18 patients, these findings are confirmed and extended.
Materials and Methods
Patients
Patients were eligible to participate in the study if they met the following criteria: age 18 yr or older, on our waiting list for renal transplantation, no known history of clinical diabetes, and no use of steroids or other immunosuppressive agents. All patients had to give informed consent to participate in the study. Eighteen Caucasian, nondiabetic patients (10 men, 8 women) were included in the study. Their primary renal diseases were hypertension (n = 5), hemolytic uremic syndrome (n = 4), IgA glomerulonephritis (n = 3), focal intracapillary glomerulonephritis (n = 1), chronic pyelonephritis (n = 1), membranous glomerulonephritis (n = 1), reflux nephropathy (n = 1), and unknown (n = 2). Ten patients were on hemodialysis and eight were on peritoneal dialysis (PD). The median duration of dialysis was 19 mo (range, 8 to 48 mo). Their median age was 49 yr (range, 23 to 58 yr), and their median body mass index (BMI) was 23.3 kg/m 2 (range, 17.8 to 34.5 kg/m2).
Seventeen patients received a renal transplant. Three patients had no long-term follow-up (one died as a result of acute bleeding, one lost her graft as a result of shock, and one had primary nonfunction).
Glucose Metabolism
Glucose metabolism was studied using intravenous glucose tolerance tests (ivgtt). The tests were performed in the morning after an overnight fast both before an d 5 d after starting tacrolimus. In PD patients, the peritoneal cavity was empty from 23:00 h until after the test. Glucose (0.5 g/kg body wt) was administered intravenously for 2 to 3 min. Blood samples for measurement of whole blood glucose, C-peptide, and insulin were taken from the opposite arm at t = -15, 0, 5, 10, 15, 20, 30, 40, 50, and 60 min. The k G (glucose disappearance rate) was calculated by linear regression from the log-transformed glucose values of t = 10 to 30 min. A kG value below 0.8 mmol/L per min was considered as abnormal, between 0.8 and 1.2 mmol/L per min as indeterminate, and above 1.2 mmol/L per min as normal (19,20).
C-peptide and insulin secretion, i.e., the secretion response to a glucose load, were calculated as area under the curve using a linear trapezoidal technique from the serum value at each time point after subtraction of the t = 0 value (increment). Insulin resistance was calculated using the insulin/glucose ratio and the homeostasis model assessment (HOMA-R: fasting glucose [mmol/L] × fasting insulin [mU/L]/22.5) (21,22,23).
During hospitalization after transplantation, urine was examined daily for glucosuria; afterward, it was checked during every visit to the outpatient clinic. When glucosuria was detected, whole-blood glucose was examined. If no glucosuria was detected, glucose was measured initially at least once every week and later at least every 3 mo. When glucose values were abnormal (>6.1 mmol/L in the fasting state or >7.8 mmol/L in the nonfasting state), glycosylated hemoglobin (HbA1c) was measured.
A dipstick method was used for the detection of glucosuria. For the measurement of glucose in whole blood, the CX 7 (Beckman Instruments, Palo Alto, CA) was used, for C-peptide and insulin, the Autodelfia (Wallac, Turku, Finland) was used, and for HbA1c, an HPLC method, Variant 2 (Biorad, Hercules, CA), was used.
Immunosuppression
Before transplantation, patients were examined by ivgtt before and 5 d after oral exposure to tacrolimus (twice daily 0.15 mg/kg body wt). Before the second ivgtt, the tacrolimus whole-blood 12-h trough level was determined (IMx; Abbott, Hoofddorp, The Netherlands).
After transplantation, immunosuppression consisted of tacrolimus and steroids in all patients. Because of the prevailing immunosuppressive protocol at the time of transplantation, six patients also received azathioprine and three patients received mycophenolate mofetil beginning on day 3 after transplantation. Target tacrolimus trough levels for all patients were 15 to 20 ng/ml from day 1 until day 14 and then 10 to 15 ng/ml from day 15 until day 28; thereafter, they were adjusted gradually to 5 to 10 ng/ml. Steroid dosage consisted of 125 mg of methylprednisolone on day 0 or 50 mg of prednisolone on day 0 and 25 mg on day 1. Thereafter, all patients received 10 mg/d for the first 4 wk after transplantation. The dosage was tapered to 5 mg/d at 3 mo after transplantation and then to 0 mg in first transplant recipients without an episode of acute rejection. In five patients, tubulointerstitial rejection was treated with steroid pulse therapy. Because of very severe acute tubulointerstitial rejection with a possible vascular component and the development of HLA class 1 antibodies 3 wk after transplantation, one patient was treated with one bolus of steroids, antithymocyte globulin, and plasmapheresis; mycophenolate mofetil was also added to the immunosuppressive regimen.
At a median of 34 mo (range, 21 to 48 mo) after transplantation, eight patients were receiving tacrolimus monotherapy, two were receiving tacrolimus and azathioprine, two were receiving tacrolimus and mycophenolate mofetil, one was receiving tacrolimus and prednisolone, and one was receiving tacrolimus, azathioprine, and prednisolone.
Statistical Analyses
For statistical analysis, the SPSS version 9.0 for Windows (SPSS, Inc., Cary, NC) was used. To compare glucose metabolism before and during the use of tacrolimus, we performed the Wilcoxon matched-pairs signed rank sum test. For correlation between the different parameters, Spearman's rho was used. A P value below 0.05 was considered to be statistically significant.
Results
Glucose and Insulin Sensitivity
In Table 1, the results of glucose metabolism before transplantation (median and ranges) are shown. Before tacrolimus, fasting glucose was normal in all patients. The kG was normal in 15 patients and indeterminate in 3 (Figure 1). There was no correlation between kG and age or gender. There was a trend toward a lower kG with an increase in BMI (r = -0.423, P = 0.08).
Glucose metabolism before and during tacrolimusa
Pretransplant insulin sensitivity index (kG): before (pre) and during (post) tacrolimus. Patients who developed diabetes mellitus after transplantation (PTDM) are indicated with the open symbol. The light gray shade indicates an indeterminate value of kG, a higher value indicates a normal kG, and a lower value indicates an abnormal kG.
Five d after the start of tacrolimus, at a median trough level of 17.1 ng/ml (range, 7.6 to 38.0 ng/ml), there was a small but statistically significant increase in median fasting glucose level, from 5.1 mmol/L to 5.2 mmol/L (P = 0.013; Table 1). The fasting glucose level decreased in 2 patients (7 to 12%), did not change in 5 patients (±5%), and increased in 11 patients (7 to 23%). The median kG decreased 39%, from 1.74 mmol/L per min to 1.08 mmol/L per min (P < 0.0001; Table 1). In 16 of 18 patients, kG decreased during tacrolimus administration (Figure 1), whereas in 2 patients, there was a small increase in kG of less than 10% (from 1.63 to 1.77 mmol/L per min and from 2.02 to 2.11 mmol/L per min). Tacrolimus trough level correlated negatively (r = -0.478, P = 0.045) with kG during tacrolimus administration (Figure 2). There was no correlation between tacrolimus trough level and the relative reduction in kG.
Relationship between kG and tacrolimus trough level before transplantation. Linear regression analysis showed a significant negative correlation (r = -0.478, P = 0.045).
Insulin Secretion and Insulin Resistance
Insulin and C-peptide values correlated significantly (r = 0.673, P = 0.002), and results for C-peptide and insulin showed exactly the same trends. Therefore, only insulin data are shown. In Table 1, median values (and ranges) of insulin secretion are shown.
After the introduction of tacrolimus, basal insulin levels (at t = 0) did not change significantly. Insulin resistance did not change when HOMA-R was used. There was a small but significant decrease in insulin resistance when the insulin/glucose ratio was used (Table 1); the ratio decreased in 11 patients, increased in 5 patients, and remained the same in 1 patient. In response to a glucose load, insulin secretion decreased significantly (median 20%). Insulin secretion changed less than 5% in 4 patients and decreased 10 to 100% in 13 patients; in 1 patient, insulin values were not determined.
Correlations between Parameters of Glucose Metabolism
In Table 2, correlations between the different parameters of glucose metabolism are shown. Before and during tacrolimus administration, fasting glucose and fasting insulin correlated significantly (P = 0.007 and P = 0.044, respectively). Fasting glucose and insulin secretion did not correlate significantly. The kG and insulin secretion and their relative changes were highly correlated (P < 0.001 to P = 0.01) at all times. Insulin resistance (HOMA-R and insulin/glucose ratio) did not correlate with either kG or insulin secretion at any time (data not shown).
Correlations between different glucose metabolism measurements (17 patients)a
Posttransplantation
Seventeen of the 18 patients underwent transplantation. Fourteen patients were followed up for a median of 34 mo (range, 21 to 48 mo). Immediately after transplantation, all patients had hyperglycemia up to 1 d while receiving high doses of steroids (median glucose, 13.8 mmol/L; range, 9.7 to 30.2 mmol/L). Fourteen patients were normoglycemic thereafter. In three patients, hyperglycemia persisted (Table 3). Two of these patients used azathioprine and the other patient used mycophenolate mofetil in addition to tacrolimus and steroids for approximately 6 mo. One patient was treated with steroid pulse therapy on day 8 after transplantation because of tubulointerstitial rejection. Today, all three of them are receiving tacrolimus monotherapy with trough levels of 5.0, 7.2, and 7.3 ng/ml, respectively. Two of these patients are being treated with a sulfonylureum derivative and low-dose, long-acting insulin at bedtime. Their HbA1c is 6.8% and 7.7%. Before transplantation, their kG was 1.21 mmol/L per min and 1.05 mmol/L per min before tacrolimus and 0.82 mmol/L per min and 0.39 mmol/L per min during their use of tacrolimus (Figure 1). The third patient's nonfasting glucose levels are mainly below 8 mmol/L, with peaks up to 11.6 mmol/L and an HbA1c of 6.5%. He is being treated only with diet and uses no medication for his hyperglycemia. Before transplantation, his kG was 1.37 mmol/L per min before and 0.82 mmol/L per min during tacrolimus administration (Figure 1). All of the other patients have been normoglycemic without oral medication or insulin since transplantation.
Characteristics of the three patients with hyperglycemia posttransplantation during use of tacrolimusa
Discussion
In this first prospective study to evaluate glucose metabolism by ivgtt before transplantation in nondiabetic dialysis patients before and after tacrolimus administration, we investigated the mechanism that is responsible for PTDM. Furthermore, during follow-up of these patients after transplantation, we investigated factors that may be associated with the development of PTDM.
Before tacrolimus, all patients were normoglycemic, but 3 of 18 (17%) had an indeterminate kG. It is widely known that end-stage renal disease and dialysis are associated with impaired glycemic control caused, in part, by increased insulin resistance and, in part, by impaired insulin secretion (24).
Several investigations have shown that the incidence of PTDM depends on the population studied, e.g., age, BMI, genetic factors (1,2,3,4,5,6,7,8,9,10,11). In our Caucasian patients, no significant correlation between kG and age or gender was detected, and only a trend toward a negative correlation between kG and BMI was found.
Insulin is excreted by both the kidney and the liver. Notwithstanding the renal impairment, the median fasting insulin level (13.0 mU/L) was within the normal reference values from our laboratory (1 to 25 mU/L). Only one patient had a fasting insulin level above 25 mU/L (26.7 mU/L) and, therefore, a high insulin/glucose ratio before tacrolimus, indicating increased insulin resistance. Although an increased incidence of insulin resistance was reported in end-stage renal failure and dialysis by Mak (24), only 1 of our 18 patients (5.5%) had an increased basal insulin level before tacrolimus. A high BMI of 33.1 kg/m2 most likely accounted for the increased resistance in this patient. Four other patients with a BMI above 25 kg/m2 (25.5 to 29.6) had normal fasting insulin levels.
The results of median C-peptide levels were above the normal reference values from our laboratory. This is because C-peptide is almost completely cleared by the kidney. Because C-peptide data correlated well with insulin data and results showed exactly the same trends, only insulin data are shown.
During tacrolimus administration, glucose metabolism was significantly altered in most patients. The kG, in particular, was markedly reduced (-39%). A reverse linear correlation between kG and tacrolimus trough level was found. An abnormal kG during the use of tacrolimus was not seen in patients who had tacrolimus trough levels below 15 ng/ml and a normal kG before the use of tacrolimus. These findings are in accordance with several clinical reports in which an impaired glucose metabolism was observed, especially when tacrolimus trough levels are high, whereas after a reduction in the tacrolimus dose, the disorder disappeared in most patients (13,25,26,27,28).
A decrease in the kG indicates an impaired removal of a glucose load from the plasma. This can be caused either by reduced insulin secretion or by increased insulin resistance. We assessed secretion by calculating the increment from the values of the ivgtt and insulin resistance by calculating the insulin/glucose ratio and HOMA-R. Insulin secretion decreased significantly (-20%), whereas insulin resistance increased. There was no change when HOMA-R was used. Because HOMA-R has not been validated explicitly for patients with renal insufficiency, the insulin/glucose ratio was also used as a parameter for insulin resistance. Thus, if anything, a significant small reduction in insulin resistance was found. These data prove that the decrease in the kG in our population was caused by a decrease in insulin secretion and not by an increase in insulin resistance. These findings are a confirmation and extension of data from animal studies (14) and from a cross-sectional study in liver transplant recipients by Fernandez et al. (17). Moreover, we were able to refute suggestions of an increase in insulin resistance in humans during tacrolimus administration (1,13). In all other clinical reports, patients were using steroids, and it is widely known that the use of steroids can cause an increase in insulin resistance (1).
On days 0 and 1 after transplantation, hyperglycemia was observed in all patients. This illustrates the possible bias that can occur when the influence of tacrolimus on glucose metabolism is evaluated during the concomitant use of steroids, especially at high doses. Three patients (17%) developed PTDM. This incidence is within the range of 8 to 36% reported in the literature (1,2,3,4,5,6,7,8,9,10,11). Their trough levels were in accordance with target levels and did not differ from levels of patients without PTDM. In addition to tacrolimus and steroids, two of these patients were treated with azathioprine and one was treated with mycophenolate mofetil. One patient was treated with steroid pulse therapy because of rejection at day 8 after transplantation. In these patients, hyperglycemia was noticed immediately after transplantation and, contrary to those without PTDM, did not disappear after day 1. Because azathioprine or mycophenolate mofetil was added at day 3 and the steroid pulse therapy in the patient with rejection was administered at day 8, it is unlikely that the additional immunosuppression and rejection therapy played an important role in the development of PTDM.
Two of the three patients who developed PTDM already had an indeterminate kG before the start of tacrolimus, whereas the third patient who developed PTDM had a low normal kG before starting tacrolimus (1.21 mmol/L per min). In contrast, none of the patients with an initial kG above 1.21 mmol/L per min developed PTDM. Thus, it seems that by performing an ivgtt before transplantation, you can define a group of patients who are at risk of developing PTDM during tacrolimus use. Of course, other risk factors for PTDM may contribute to the actual development of PTDM, such as age and race (1,2,3,4,5,6,7,8,9,10,11), the use of steroids, and high tacrolimus trough levels (23,25,26,27,28). It is worth noting that all three patients who developed PTDM had a BMI above 25 kg/m2. One patient with an indeterminate kG (0.85 mmol/L per min) and a normal BMI (23.2 kg/m2) did not develop diabetes after transplantation. Three patients with a normal kG and a BMI above 25 kg/m2 (25.5, 27.7, and 29.6 kg/m2) also did not develop PTDM. Thus, in addition to a low kG, BMI could be an important factor in determining the actual development of PTDM.
In summary, this prospective study of 18 Caucasian dialysis patients shows that before transplantation, the kG decreases significantly after the introduction of tacrolimus in almost all patients. This decrease is caused by a reduction in insulin secretion, whereas insulin resistance does not increase. A high tacrolimus trough level correlates with a low kG.
After transplantation, 14 patients could be followed up long-term. Three patients developed PTDM, two with an indeterminate and one with a low normal kG before tacrolimus; none of the patients with a normal kG developed PTDM. This indicates that the kG, calculated from data before the use of tacrolimus, can define a population at risk for PTDM during the use of tacrolimus.
- © 2001 American Society of Nephrology
References
Jump to section
More in this TOC Section
Cited By...
- Metabolic Disorders with Kidney Transplant
- New-Onset Diabetes After Renal Transplantation: Risk assessment and management
- Transplant-Associated Hyperglycemia: A New Look at an Old Problem
- New-Onset Diabetes after Kidney Transplantation: Risk Factors
- The Immunosuppressant FK506 Uncovers a Positive Regulatory Cross-talk between the Hog1p and Gcn2p Pathways
- Improved Cardiovascular Risk Profile and Renal Function in Renal Transplant Patients after Randomized Conversion from Cyclosporine to Tacrolimus
- Pharmacological Agents That Directly Modulate Insulin Secretion
- Malonyl-CoA Signaling, Lipid Partitioning, and Glucolipotoxicity: Role in {beta}-Cell Adaptation and Failure in the Etiology of Diabetes
- Successful Islet Transplantation: Continued Insulin Reserve Provides Long-Term Glycemic Control
- Association of Hepatitis C with Posttransplant Diabetes in Renal Transplant Patients on Tacrolimus
- Glucose Metabolism in Renal Transplant Recipients on Tacrolimus: The Effect of Steroid Withdrawal and Tacrolimus Trough Level Reduction
- Glucose Metabolism in the First 3 Years after Renal Transplantation in Patients Receiving Tacrolimus versus Cyclosporine-Based Immunosuppression