- Top of page
- Materials and Methods
Long-term use of cyclosporine after renal transplantation results in nephrotoxicity and an increased cardiovascular risk profile. Tacrolimus may be more favorable in this respect. In this randomized controlled study in 124 renal transplant patients, the effects of conversion from cyclosporine to tacrolimus on renal function, cardiovascular risk factors, and perceived side-effects were investigated after a follow-up of 2 years. After conversion from cyclosporine to tacrolimus renal function remained stable, whereas continuation of cyclosporine was accompanied by a rise in serum creatinine from 142 ± 48 μmol/L to 157 ± 62 μmol/L (p < 0.05 comparing both groups). Conversion to tacrolimus resulted in a sustained reduction in systolic and diastolic blood pressure, and a sustained improvement in the serum lipid profile, leading to a reduction in the Framingham risk score from 5.7 ± 4.3 to 4.8 ± 5.3 (p < 0.05). Finally, conversion to tacrolimus resulted in decreased scores for occurrence of and distress due to side-effects. In conclusion, conversion from cyclosporine to tacrolimus in stable renal transplant patients is beneficial with respect to renal function, cardiovascular risk profile, and side-effects. Therefore, for most renal transplant patients tacrolimus will be the drug of choice when long-term treatment with a calcineurin inhibitor is indicated.
- Top of page
- Materials and Methods
Cyclosporine and tacrolimus are calcineurin inhibitors that are very effective in reducing the incidence of acute rejection after renal transplantation. However, cyclosporine-induced side-effects, including nephrotoxicity, hypertension, and hyperlipidemia (1) may contribute to the high cardiovascular morbidity in renal transplant patients (2), and can also lead to an accelerated loss of graft function (3,4). Tacrolimus has also been associated with nephrotoxicity, but several studies report an increased graft survival in patients using tacrolimus as initial immunosuppressive treatment (5,6). Moreover, tacrolimus is associated with less hypertension and hyperlipidemia, thereby improving the cardiovascular risk profile (1). However, a drawback of tacrolimus is the increased risk of developing diabetes mellitus (5,6), which could antagonize the beneficial effect of reduced blood pressure and improved hyperlipidemia on the cardiovascular morbidity and mortality.
Many of the renal transplant patients that have been transplanted in the past two decades are using cyclosporine as maintenance immunosuppressive therapy. The question arises whether conversion from cyclosporine to tacrolimus in these chronic renal transplant patients will lead to an improved cardiovascular risk profile and whether renal graft function still can be improved.
In this prospective randomized controlled multicenter trial, we compared the effects of conversion from cyclosporine to tacrolimus in stable chronic renal transplant patients on graft function and on the cardiovascular risk profile, expressed as the Framingham risk score. Also, we assessed the patients' perceived symptom experience associated with side-effects of the immunosuppressive therapy, using the ‘modified transplant symptom occurrence and symptom distress scale’ (7). The changes in several metabolic parameters during the first 6 months after conversion from cyclosporine to tacrolimus have been described in detail before (8). In this paper, data on renal function, Framingham risk score, and side-effects until the end of follow-up, at 2 years after randomization, are presented.
- Top of page
- Materials and Methods
A total of 124 patients were included in the study, 64 were converted to tacrolimus and 60 remained on cyclosporine. Mean time after transplantation was 6.6 ± 3.7 years for the tacrolimus group and 5.7 ± 2.7 years for the cyclosporine group. The demographic data have been described previously (9). At baseline, the cyclosporine level was 129 ± 42 ng/mL in the cyclosporine group and 130 ± 42 ng/mL in the tacrolimus group. During follow-up, the cyclosporine level remained unchanged in the cyclosporine group. In the tacrolimus group, the tacrolimus level was 7.2 ± 2.0 ng/mL at 3 months and 6.6 ± 1.8 ng/mL at 2 years after randomization. Baseline prednisone dose was equal in both groups (0.10 ± 0.04 mg/kg). Baseline azathioprine dose was 0.88 ± 0.40 mg/kg (n = 10) in the tacrolimus group and 0.95 ± 0.47 mg/kg (n = 15) in the cyclosporine group. In the cyclosporine group, two patients used mycophenolate mofetil 2000 mg/day. Prednisone was stopped in two patients, and they were both withdrawn from the study. Azathioprine was stopped in one patient in the tacrolimus group and in one patient in the cyclosporine group.
The study protocol was not completed until the end of follow-up in 17 patients in the tacrolimus group and in 15 patients in the cyclosporine group (Table 1). In the tacrolimus group, two patients reached end-stage renal disease (ESRD), due to biopsy-proven chronic allograft nephropathy. In the cyclosporine group one patient, known to have recurrent membranous glomerulonephritis, reached ESRD. In each group, one patient was withdrawn from the study due to deteriorating renal function associated with chronic allograft nephropathy (tacrolimus group) and cyclosporine nephrotoxicity (cyclosporine group), respectively.
Table 1. Patients that did not complete the study
|Death of patient||6 (10%)||2 (3.1%)|
| Pulmonary embolism||1||0|
|End-stage renal disease||1 (1.7%)||2 (3.1%)|
|Withdrawal of study medication|| || |
| Deterioration of renal function||1 (1.7%)||1 (1.6%)|
| Intractable hypertension||2 (3.3%)||0 (0%)|
| New-onset DM||0 (0%)||2 (3.1%)|
| Dysregulation of pre-existent DM||0 (0%)||2 (3.1%)|
| Other side-effects||5 (8.3%)||8 (12.5%)|
|Total||15 (25.0%)||17 (26.6%)|
Prior to inclusion, all patients had a stable serum creatinine concentration. During the 2-year follow-up period, serum creatinine increased in the cyclosporine group from 142 ± 48 to 157 ± 62 μmol/L (p < 0.001), while it remained stable in the tacrolimus group (139 ± 32 μmol/L at baseline and 145 ± 56 μmol/L at 24 months, p = 0.26, p < 0.05 for comparison between groups, Figure 1). In the cyclosporine group, the endogenous creatinine clearance decreased from 59 ± 26 mL/min to 49 ± 22 mL/min (n = 50; p < 0.001), whereas renal function remained stable in the tacrolimus group (60 ± 22 mL/min and 64 ± 33 mL/min, n = 58, p < 0.01 for comparison between groups).
In the cyclosporine group, 30 patients (50%) had a more than 10% rise in serum creatinine level during the 2 years of follow-up, as opposed to 15 patients (23.4%) in the tacrolimus group, while 6 patients (10%) in the cyclosporine group had a more than 10% improvement in serum creatinine level during follow-up, compared with 20 patients (31%) in the tacrolimus group (p < 0.001).
Proteinuria did not significantly change in either group during follow-up [median baseline value 0.2 (0–4.7) g/24 h in the cyclosporine group and 0.2(0–3.6) g/24 h in the tacrolimus group].
Cardiovascular risk factors
In the tacrolimus group, systolic and diastolic blood pressure (manually measured) were reduced from 144 ± 20 mmHg to 137 ± 21 mmHg (p < 0.001), and from 84 ± 11 mmHg to 80 ± 9 mmHg (p < 0.01), respectively (Figure 2). As a result, blood pressure in the tacrolimus group was significantly reduced when compared with the cyclosporine group (p < 0.01). The dynamap blood pressure readings essentially showed the same results (data not shown).
Figure 2. Systolic and diastolic blood pressure (mmHg, mean and SEM) on cyclosporine (——) and on tacrolimus (- - - -). p < 0.01 comparing the area under the curve (AUC) on cyclosporine and on tacrolimus for both systolic and diastolic blood pressure.
Download figure to PowerPoint
In the cyclosporine group, the systolic blood pressure was reduced by at least 12 mmHg in 15 patients (25%), as opposed to 28 patients (44%) in the tacrolimus group (p < 0.05). The diastolic blood pressure was reduced by at least 12 mmHg in 5 (8%) cyclosporine-treated patients and 17 (27%) tacrolimus-treated patients (p < 0.01).
The number of antihypertensive drugs per patient in the cyclosporine group was 1.9 ± 1.0 at baseline and 2.3 ± 1.1 at 2 years follow-up. In the tacrolimus group, the number of antihypertensive drugs was 2.0 ± 1.1 at baseline and 2.3 ± 1.1 at 2 years follow-up (no differences between groups).
Serum total and LDL cholesterol, and serum triglycerides were significantly reduced in the tacrolimus group, from 5.9 ± 0.8 mmol/L to 5.2 ± 0.8 mmol/L (p < 0.001), from 3.6 ± 0.8 mmol/L to 3.1 ± 0.8 mmol/L (p < 0.001), and from 2.1 ± 1.0 mmol/L to 1.7 ± 1.1 mmol/L (p < 0.05), respectively (p < 0.001 for comparison between groups for total and LDL cholesterol and p < 0.01 for triglycerides, Figure 3). Serum HDL cholesterol level remained unchanged in both groups. The use of a statin did not influence the changes in the serum lipid levels. Commencement of statin treatment or an increase in the dosage according to predefined guidelines was necessary in three patients in the tacrolimus group and in 10 patients in the cyclosporine group. After conversion to tacrolimus the statin was stopped in one patient, whereas this did not occur in the cyclosporine group.
Figure 3. Serum lipid levels and serum triglyceride levels (mmol/L, mean and SEM) on cyclosporine (——) and on tacrolimus (- - - -). p < 0.001 comparing the area under the curve (AUC) on cyclosporine and tacrolimus for total and low-density lipoprotein (LDL) cholesterol, and p < 0.01 for the serum triglyceride levels.
Download figure to PowerPoint
The fasting glucose level and the HbA1c level were analyzed for diabetic and nondiabetic patients separately, as in diabetic patients both glucose and HbA1c levels are influenced by dietary (non)-compliance and changes in medication. Fasting glucose and HbA1c levels did not change in either group during follow-up (Table 2). In the cyclosporine group, three patients developed new-onset diabetes mellitus vs. four patients in the tacrolimus group. In the diabetic patients of the cyclosporine group (n = 8), the anti-diabetic treatment remained unchanged in five patients. In two patients the dosage of anti-diabetic medication had to be increased and one patient was started on insulin. In the diabetic patients of the tacrolimus group (n = 8), the anti-diabetic medication remained unchanged in three patients. In two patients the dosage of anti-diabetic medication had to be increased and two patients were started on insulin. In one patient the dosage of insulin was reduced.
Table 2. Glucose regulation on cyclosporine vs. tacrolimus
| ||Cyclosporine||Tacrolimus|| |
|baseline||24 months||baseline||24 months|
|Glucose (mmol/L)||nondiabetics||5.0 (4.1–6.2)||5.2 (4.0–7.0)||5.0 (3.4–7.6)||5.2 (4.0–7.1)||NS|
| ||(n = 108)|
| ||diabetics||9.5 (6.5–11.8)||9.7 (5.5–4.0)||7.7 (5.8–11.6)||7.9 (6.0–14.1)||NS|
| ||(n = 16)|
|HbA1c (%)a||nondiabetics||5.5 ± 0.4||5.7 ± 0.3||5.5 ± 0.4||5.6 ± 0.4||NS|
| ||(n = 31)|
| ||diabetics||7.9 ± 0.6||8.0 ± 0.8||8.8 ± 1.7||7.5 ± 1.7||NS|
| ||(n = 16)|
The Framingham risk score was significantly reduced from 5.7 ± 4.3 at baseline to 4.8 ± 5.3 after 2 years in the tacrolimus group (p < 0.05), and remained unchanged in the cyclosporine group (6.0 ± 3.1 at baseline and 6.2 ± 3.6 at 24 months, p < 0.05 for comparison between groups, Figure 4). The use of a statin at baseline did not influence the change in Framingham risk score upon conversion to tacrolimus.
Figure 4. Framingham risk score (mean and SEM) on cyclosporine and on tacrolimus. p < 0.05 comparing the area under the curve (AUC) on cyclosporine and tacrolimus.
Download figure to PowerPoint
Symptom experience associated with side-effects of immunosuppressive therapy
In the tacrolimus group, eight patients were taken off tacrolimus due to the following side-effects: hair loss (n = 2), headache (n = 2), diarrhea (n = 1), palpitations (n = 1), tremor (n = 1) and toxic exanthema (n = 1). In the cyclosporine group, two patients were withdrawn from the study due to gingival hyperplasia (n = 1) and to severe osteoporosis for which prednisone was stopped (n = 1). In addition, two patients wanted to be taken off cyclosporine and one patient was converted from cyclosporine to azathioprine because of her wish to become pregnant.
The 10 most frequently occurring side-effects at baseline were in decreasing order of frequency: increased hair growth, fatigue, bruises, excessive appetite, poor concentration, changed bodily appearance, poor vision, fragile skin, swollen ankles, and insomnia. The 10 most distressing side-effects at baseline were: impotence for men and painful menstruation for women, fragile skin, insomnia, warts, poor concentration, back pain, muscle weakness, stomach complaints, fatigue and nightmares. The item hallucinations was excluded from the analysis of symptom distress as it occurred in only one patient.
The overall ridit for symptom occurrence evolved significantly after randomization, both in the cyclosporine and in the tacrolimus group. A significant increase (p < 0.0001) in symptom occurrence was observed in patients on cyclosporine at 3, 12 and 24 months after baseline measurement (Figure 5). Conversely, patients who were converted to tacrolimus experienced a significant decrease (p < 0.0001) in symptom occurrence compared with baseline level. The ridit of 0.458 for the tacrolimus group at 12 months means that a randomly selected patient from that group has a chance of 45.8% to score higher than a randomly selected patient at baseline (reference group). This chance of 45.8% is significantly lower than the chance of 50% (ridit 0.500) inherent to the reference group.
Figure 5. Evolution of overall symptom occurrence and symtom distress at 3, 12 and 24 months. p < 0.001 for all comparisons with baseline values.
Download figure to PowerPoint
For symptom distress, the overall ridit of the tacrolimus group at 3 months after conversion was significantly increased (p < 0.0001), while the ridit of the cyclosporine group was decreased. Nonetheless, at 12 and 24 months, patients of the tacrolimus group reported significantly less, and cyclosporine patients significantly more symptom distress compared with the baseline level.
- Top of page
- Materials and Methods
This randomized controlled study demonstrates that after conversion from cyclosporine to tacrolimus in stable renal transplant patients, graft function remains stable as opposed to a gradual deterioration of graft function when continuing on cyclosporine. Furthermore, replacement of cyclosporine by tacrolimus is followed by an improvement in cardiovascular risk profile, which is sustained during a 2-year follow-up period. In addition, the occurrence of and distress due to side-effects is reduced.
In patients using tacrolimus as part of the initial immunosuppressive treatment, the incidence of acute and steroid-resistant rejections is reduced, compared with patients using cyclosporine (5,6), without affecting 1-year graft and patient survival. Acute rejection is a major risk factor for developing chronic allograft nephropathy (CAN) (12), and several studies have suggested that tacrolimus may indeed improve long-term graft survival rates. Data from the UCLA-UNOS (United Network for Organ Sharing) kidney transplant registry showed a graft half-life of approximately 14 years on tacrolimus and 8–9 years on cyclosporine and other immunosuppressive agents (13). Since this first report several other studies have provided evidence that compared to an initial immunosuppressive protocol including cyclosporine, a protocol including tacrolimus may increase graft half-life (6,14). However, the influence of tacrolimus on graft survival is not unequivocal. In the analysis of Hariharan et al. the use of tacrolimus did not contribute to the recent improvements in graft survival (15).
The putative beneficial effect of tacrolimus on graft function and ultimately also graft survival may be related to several favourable effects of tacrolimus compared with cyclosporine. CAN is the leading cause of late graft failure and may result from chronic rejection as well as from nonimmunologic processes such as calcineurin inhibitor nephrotoxicity and graft vascular disease (16). The occurrence of an acute rejection is the strongest predictor for the development of chronic rejection (16), and tacrolimus reduces the incidence of acute rejections. Second, tacrolimus may be less nephrotoxic than cyclosporine, thus contributing to the prolonged graft survival. The induction of profibrotic cytokines such as osteopontin (17) and transforming growth factor (TGF)-β (18) by cyclosporine is one of the pathophysiologic mechanisms in chronic cyclosporine nephropathy. Although TGF-β is also involved in chronic tacrolimus nephropathy (19), several studies suggested a reduced renal mRNA-expression of TGF-β or other profibrotic genes on tacrolimus vs. cyclosporine (20,21). Finally, hyperlipidemia and hypertension are important nonimmunologic factors contributing to CAN (22–24). In this study, we observed reductions in serum LDL cholesterol and triglyceride levels and blood pressure, which were apparent at 3 months after conversion to tacrolimus and persisted until the end of follow-up. This sustained improvement in serum lipid levels as well as in blood pressure is likely to be of benefit in maintaining graft function.
Mayer et al. reported a reduced incidence of CAN after the use of tacrolimus as part of the initial immunosuppression (6). It is of interest to see that even in our study population, consisting of patients that were 6.1 ± 3.2 years post-transplantation, the course of the graft function can still be improved by conversion to tacrolimus. In this respect, the question can be raised whether equipotent dosages of tacrolimus and cyclosporine were used. In our population, consisting mainly of Caucasian patients, the dosage of tacrolimus was sufficient to prevent acute rejections in all patients. The mean tacrolimus trough level (approximately 7 ng/mL) as well as the mean cyclosporine trough level (approximately 130 ng/mL) was within the ranges that have been advised for maintenance immunosuppressive therapy after renal transplantation (25). The tacrolimus levels in our patients were similar to the levels found in the study of Mayer et al. (6). A second question is whether the method of cyclosporine monitoring could have influenced the results. When this study was designed, it was common practice to use the trough level (C0) for monitoring. However, it is now recognized that the area-under-the-curve for the first 4 h post-dose (AUC0-4) is superior to the C0 level of cyclosporine in predicting the risk for acute rejection and cyclosporine-induced nephrotoxicity (26), and the 2-h post-dose (C2) level of cyclosporine is an accurate surrogate marker of the AUC0-4 cyclosporine level (27). Prospective studies on the long-term effects of replacing C0 monitoring by C2 monitoring in maintenance renal transplant patients are limited. Available date suggest that dose reduction in overexposed patients leads to improvements in renal function and blood pressure (28). However, it is currently unclear whether C2 monitoring can result in comparable effects on graft function, blood pressure and serum lipids as we have observed after conversion from cyclosporine to tacrolimus.
We previously already showed that conversion from cyclosporine to tacrolimus resulted in a reduction in systolic and diastolic blood pressure, and in an improvement in several metabolic cardiovascular risk factors, including total and LDL cholesterol levels, triglyceride levels, fibrinogen levels, and the oxidizability of the LDL particles (8). We now demonstrate that this improvement in the cardiovascular risk profile is sustained. Although the reductions in systolic and diastolic blood pressure are modest, in patients with a high risk of developing cardiovascular disease, such as renal transplant patients, the numbers-needed-to-treat to prevent cardiovascular disease or cardiovascular death are relatively small. For example, in case of a reduction of the systolic blood pressure with 12 mmHg, the numbers-needed-to-treat to prevent cardiovascular disease is only 10 (29).
During the 2 years of follow-up, conversion from cyclosporine to tacrolimus did not affect plasma fasting glucose and HbA1c levels, and there was no significant difference in the occurrence of new-onset diabetes mellitus. The diabetogenic effect of tacrolimus is dose-related, and the risk of developing diabetes mellitus is highest in the initial period after transplantation when high trough levels are reached (30), whereas at standard maintenance trough levels, no difference between cyclosporine and tacrolimus could be determined with regard to their diabetogenic properties (31). The risk of developing diabetes mellitus after conversion to tacrolimus appears to be restricted to patients who already experience a disturbed glucose tolerance prior to therapy (5,32).
The improvements in lipid profile and blood pressure after conversion to tacrolimus resulted in a sustained reduction of the Framingham risk score. This cardiovascular risk score was developed in a predominantly Caucasian population without specific underlying diseases (10) and has been validated in renal transplant patients, where it was shown to predict ischemic cardiovascular events, though underestimating the absolute risk, especially related to diabetes mellitus (33). In the general population the observed reduction in Framingham risk score upon conversion to tacrolimus would imply a relative reduction of approximately 20% in the estimated risk for coronary heart disease over a period of 10 years. Although it is likely that renal transplant patients will also benefit from the reduced cardiovascular risk score, it is difficult to extrapolate the risk rates to this specific population.
After conversion from cyclosporine to tacrolimus, the occurrence of symptoms associated with side-effects of the immunosuppressive regimen decreased significantly. This was also observed for symptom distress, except for the measurement at 3 months after conversion. These findings demonstrate the beneficial effect of conversion to tacrolimus in terms of patients' symptom experience. Patients' subjective appraisal of symptoms associated with side-effects is very important for their quality-of-life (34), and is also of great importance for the motivation to comply with the medication regimen (35). Several studies have demonstrated a relation between distressing symptoms and subsequent noncompliance with immunosuppressive medication (36,37). Further analysis of symptom occurrence and symptom distress at item level will provide a more detailed picture of the evolution of individual symptoms after conversion from cyclosporine to tacrolimus.
In summary, conversion from cyclosporine to tacrolimus in stable chronic renal transplant patients has a beneficial effect on renal graft function, results in a sustained improvement in the cardiovascular risk profile and reduces the frequency and distress of side-effects. In our opinion, for most renal transplant patients tacrolimus will be the drug of choice when long-term treatment with a calcineurin inhibitor is indicated.