Secondary Logo

Journal Logo

Lower Variability in 24-Hour Exposure During Once-Daily Compared to Twice-Daily Tacrolimus Formulation in Kidney Transplantation

Stifft, Frank1; Stolk, Leo M.L.2; Undre, Nasrullah3; van Hooff, Johannes P.1; Christiaans, Maarten H.L.1,4

doi: 10.1097/01.TP.0000437561.31212.0e
Clinical and Translational Research

Introduction Tacrolimus has originally been registered as a twice-daily formulation (Prograf, Tac BID), although a once-daily formulation (Advagraf, Tac QD) is also available. A reduced intrapatient variability of Tac Cmin, a surrogate marker for 24-hour drug exposure (AUC0–24), has been suggested. The variability of AUC0–24 has never been studied prospectively yet. The purpose of this study was to investigate the change in intrapatient variability of Tac AUC0–24 after converting from Tac BID to Tac QD.

Methods Forty renal transplant patients on Tac BID were converted on a 1:1 (mg/mg) basis to Tac QD in an investigator-driven comparative pharmacokinetic (PK) study. AUC0–24 was determined five times before and after conversion. Duplicate samples were collected by the patients themselves using the dried blood spot method. The main outcome measure is the change in intrapatient variability of AUC0–24 expressed as coefficient of variation (CV). Moreover, the influence of Cyp3A5 genotype polymorphism on the change in CV was studied.

Results In total, 400 AUC0–24 profiles were available for analysis. Conversion to Tac QD resulted in a significant improvement in intra-patient CV from 14.1% to 10.9% (P=0.012). Patients with the Cyp3A5*1/*3 genotype (n=11) had a numerically larger improvement in CV than patients with the CYP3A5*3/*3 genotype.

Conclusion Intrapatient CV of Tac AUC0–24 improved after converting from Tac BID to Tac QD in stable renal transplant patients, especially in patients with the CYP3A5*1/3 genotype. Given the very strict protocol of this PK study, this improvement is most likely due to the different intrinsic PK properties of Tac QD and Tac BID.

1 Department of Internal Medicine, Division of Nephrology, Maastricht University Medical Centre, Maastricht, The Netherlands.

2 Department of Clinical Pharmacology and Toxicology, Maastricht University Medical Centre, Maastricht, The Netherlands.

3 Astellas Pharma Europe Ltd., Chertsey, United Kingdom.

4 Address correspondence to: Maarten H.L. Christiaans, Department of Internal Medicine, Maastricht University Medical Centre, P. Debyelaan 25, 6229 HX Maastricht, The Netherlands.

F.S. and L.S. declare no conflicts of interests. N.U. is an employee of Astellas. J.v.H. and M.H.L.C. participated in trials sponsored by Astellas and have received lecture fees from Astellas.

This investigator-driven study was performed with an unrestricted grant from Astellas Pharma Europe Ltd.


F.S. participated in performing the research, analyzing data, and writing of the article. L.S. contributed analytical tools and participated in the revision of the article. N.U. contributed to the research design and participated in the revision of the article. J.v.H. participated in the research design and writing of the article. M.H.L.C. is the principal investigator and participated in the research design and writing of the article.

Received 24 June 2013. Revision requested 9 July 2013.

Accepted 10 October 2013.

Accepted December 16, 2013

Tacrolimus (Tac) is the cornerstone immunosuppressive drug in solid organ transplantation and has been available as an immediate-release formulation administered twice daily (Prograf, Tac BID) for almost 20 years. More recently, a once-daily prolonged-release formulation (Advagraf, Tac QD) has been made available. The efficacy and safety profile of Tac QD is comparable to that of Tac BID after converting to Tac QD in stable kidney transplant recipients (1, 2).

Tacrolimus has a narrow therapeutic window and high interpatient variability in its bioavailability. Therefore, therapeutic drug monitoring (TDM) of exposure is mandatory to ensure that patients are maintained in a narrow predefined therapeutic exposure window (3). The area under the 24-hour blood concentration-time curve (AUC0–24) is presumed to better reflect exposure to the drug than predose concentrations or other single-time measurements. In clinical practice, this is not practically feasible, and therefore, the trough level (Cmin) that correlates well with AUC0–24 is used as a surrogate marker for exposure. This relationship between Cmin and the AUC0–24 is similar for Tac QD and Tac BID (1, 2, 4).

There are indications that high intrapatient (within-patient) coefficient of variation (CV) in Tac BID Cmin is a risk factor for poor long-term outcome after kidney transplantation (5). Earlier safety and tolerability or pharmacokinetic (PK) studies have suggested that CV for Cmin was lower during Tac QD when compared to Tac BID in stable kidney transplant patients (1, 2). An explanation for this lower CV is that once-daily dosing of tacrolimus leads to an improved adherence (6). However, improvement in the intrinsic PK properties of Tac QD might result in a lower CV as well.

Alloway et al. (1) reported a greater decrease in the CV of tacrolimus exposure with Tac QD versus that with Tac BID for African American patients compared to Caucasian patients. Similarly, Wu et al. (7) reported a decrease in the CV in Cmin after converting to Tac QD in a Chinese population. This decrease in CV might be linked to the higher prevalence of cytochrome P450 3A5*1 (Cyp3A5*1) allele in African Americans and Chinese. It is expressed in more than 70% among African Americans and 30% in Chinese in contrast to only 10% in whites (8–10).

To our knowledge, intrapatient CV of the Tac exposure (AUC0–24)—both of Tac BID and Tac QD—has not yet been investigated prospectively in stable renal transplant recipients. Therefore, a prospective PK study was designed, in which CV of the AUC0–24 in stable renal transplant patients was compared before and after converting from Tac BID to Tac QD. Moreover, the influence of the CYP3A5 polymorphism on the change in CV after conversion was analyzed.

Back to Top | Article Outline


Forty stable renal transplant patients on Tac BID (25 men and 15 women) were included in the study after giving informed consent. Mean age was 61.2 years (range, 36–78 y) with an average time since transplantation of 4.5 years (range, 1.3–13.5 y). They had an average estimated glomerular filtration rate of 40 mL/min/1.73 m2 (range, 20–66 mL/min/1.73 m2). Thirty-one patients were on tacrolimus monotherapy. Concomitant immunosuppression consisted of mycophenolic acid in four patients, sirolimus (1 mg QD) in three patients, and prednisolone (5 mg) in 2 patients. None of the patients received more than one concomitant immunosuppressive drug. Of the 40 patients, 11 had the Cyp3A5*1/*3 genotype and the other 29 had the Cyp3A5*3/*3 genotype. There were no patients with the Cyp3A5*1/*1 genotype.

Each patient performed 10 PK profiles (n=400 profiles), consisting of 8 duplicate samples per profile. For each profile, the duplicate samples were analyzed in separate analytical runs. In 6.3% of these runs, the quality control of one of the runs was insufficient. In these instances, AUC was computed on one value per time point. In 3% of the profiles, one noncrucial value was missed. This missing value could be estimated by interpolation. Hence, only 12 patients had to repeat one profile.

Back to Top | Article Outline


Relevant PK parameters and the corresponding average 24-hour profiles for both formulations before and after conversion are given in Table 1 and Figure 1. One week after conversion, in three patients. the tacrolimus trough levels were less than 3.5 μg/L. In these patients, according to the protocol, the Tac doses were increased by 1, 1.5, and 2 mg to reach a Cmin greater than 4.0 μg/L, and the profiles were made only after establishing stable Cmin. No further dose adjustments were made. Therefore, in Table 1, AUC0–24 and Cmin are also listed as dose-normalized values. There was good correlation between AUC0–24 and Cmin for both formulations, with virtually identical slopes of the Tac BID and Tac QD regression curves (Fig. 2).







Back to Top | Article Outline

Intrasubject Variability

Because all five time-concentration profiles (AUC0–24) for each formulation in each patient were performed with a fixed Tac dose, all 40 patients could be included for the analyses. After conversion, intrapatient CV of the AUC0–24 decreased significantly from 14.1% to 10.9% (P=0.012). The intrapatient CV for Cmin decreased only marginally from 15.3% to 13.7% (P=0.2).

Back to Top | Article Outline

Cyp3A5 Polymorphisms

Table 2 and Figure 3 depict the influence of the Cyp3A5 polymorphism on the intrapatient variability during Tac BID and Tac QD. In the Cyp3A5 *1/*3 group, the CV decreased from 18.2% to 12.8% (P=0.06), whereas in the Cyp3A5 *3/*3 group. the CV decreased from 12.6% to 10.2% (P=0.16).





The statistically significant difference in CV between both CYP3A5 groups during Tac BID became numerically much smaller after conversion and was not statistically significant anymore (P=0.22). Changes in the CV for Cmin were in the same direction, although to a lesser extent.

Back to Top | Article Outline


Our study is the first designed to prospectively study the change in intrapatient variability in systemic tacrolimus exposure after converting from Tac BID to Tac QD in stable renal transplant patients. Five sequential PK profiles both during Tac BID and Tac QD were collected under steady-state conditions by means of the dried blood spot (DBS) method (11). This method is routinely used for measuring Tac levels in our outpatient clinical practice for more than 5 years.

We have proven that this method is very reliable for determining Tac levels, especially when performed by well-instructed and motivated patients (11). The low percentage of missing crucial samples and samples of low quality in this study is another proof that the DBS method is robust. Only 12 of the 400 profiles had to be repeated.

A significant improvement in the intrapatient variability in 24-hour exposure after converting from Tac BID to Tac QD was found, confirming previous post hoc analyses (1, 2). This improvement in CV can be caused either by improved adherence of the patients or by intrinsic differences in the PK properties of the two Tac formulations. Although nonadherence can never completely be excluded, it is very unlikely that it has occurred with these selected patients and with this strict protocol. Hence, the improvement in intrapatient CV after converting from Tac BID to Tac QD is most likely due to differences in the PK properties of Tac QD compared to those of Tac BID.

Although the AUC differed only by −2.7% between Tac BID and Tac QD, Cmin did so by −10.8%. Earlier phase 2 trials showed a 5% to 10% lower AUC for Tac QD in stable kidney transplant recipients compared to Tac BID (1, 2, 12), with a larger patient population compared to our study (60 and 67 vs. 40, respectively). It could be that so that our study was underpowered to detect a significant difference with respect to the AUC, but values and the corresponding confidential interval of the current study comply with those earlier reports. Our current study was not designed to compare the relative bioavailability of Tac BID and Tac QD. On the other hand, we also found an identical relationship between Cmin and AUC given the virtually identical slopes of the regression lines (Fig. 2).

Especially in the Cyp3A5*1/*3 group, the intrapatient CV decreased after conversion, although this did not reach statistical significance (P=0.062). It has been hypothesized that Cyp3A activities are generally decreased in the Cyp3A5*3/*3 genotype, and the chances of being influenced by environmental factors would then increase through induction of Cyp3A4 and Cyp3A5 by the pregnane X receptor (13). Cyp3A4 is postulated to be more prone to induction than Cyp3A5, and since, in Cyp3A5 expressors, the Cyp3A5 activity would be rather stable, different amounts of Cyp3A4 could cause substantial intra-individual variability.

On the contrary, Pashaee et al. (14) could not establish a relationship between the Cyp3A5 genotype and intrapatient variability. We found in our study that Cyp3A5*1/*3 carriers had a higher CV under Tac BID, which declined after converting to the once-daily formulation. This would be in line with the findings of Alloway et al. (1) and Wu et al. (7), at least if we assume that the African Americans and Chinese represent a high prevalence of the Cyp3A5*1 genotype (8, 9).

In our white population, we included only a limited number of patients with the CYP3a5*1 allele, and none of the patients were homozygous for this allele. Also in patients homozygous for the Cyp3A5*1 allele, a larger decrease in CV can be expected. Nevertheless, in our study, the CYP3A5*1/*3 group had a significantly higher intrapatient CV compared to the CYP3A5*3/*3 group under Tac BID, a difference that was not significant anymore after converting to Tac QD.

In conclusion, we confirmed the reduced intrapatient CV for Tac QD compared to Tac BID. This improvement does not merely result from improved adherence but reflects the intrinsic PK properties of Tac QD. The long-term clinical significance of a decreased variability in Tac drug exposure has yet to be determined. Thereafter, patients might profit from a drug with a reduced variability in systemic exposure. The reduction in intrapatient CV is numerically larger in patients with the Cyp3a5*1 allele. A limitation of our study is the relatively small number of patients carrying the Cyp3a5*1 allele and especially the absence of Cyp3a5*1 homozygotes. This should be explored in a future study.

Back to Top | Article Outline



All patients 18 years or older with a stable renal allograft function 6 months before inclusion and on stable Tac BID dosages were eligible for participation in our study. Because we were particularly interested in the Cyp3A5*1 genotype, patients with a daily dose at least twice as high as the average daily dose (i.e., ≥6 mg) were regarded as potential carriers of the cyp3A5*1 genotype and were preferentially selected. Patients were excluded if they had any organ other than kidney transplanted, had active malignancy, had renal replacement therapy, had signs of infection shortly before inclusion, or have chronic diarrhea. Patients were enrolled after they had given their informed consent. The study had been approved by the local medical ethics committee (METC 09-2-040) and was registered in the EudraCT trial register (registration no. 2009-010400-28).

Back to Top | Article Outline

Study Design

The study was an investigator-driven, single-center, open-label, prospective, two-sequence conversion comparative study, in which PK parameters were evaluated with respect to their variability before and after converting from Tac BID to Tac QD. The study period started with a run-in period of 2 weeks during which the patients were instructed to take the Tac BID under fasting conditions, that is, 1 hour before or 2 hours after a meal. From the third week on, five weekly 24-hour blood concentration-time profiles to determine AUC0–24 were taken (Fig. 4), and if one of the profiles was not evaluable (see below), an additional profile was taken. After five evaluable profiles had been taken, patients were converted to Tac QD on a 1:1 (mg/mg) basis. After a stabilization period of 2 weeks, five additional evaluable 24-hour AUC profiles at weekly intervals were taken in the second phase of the study (Fig. 4). One week after conversion, a venous tacrolimus trough level was taken and Tac QD dose was adjusted if the trough level was less than 3.5 μg/L. Sampling of the first blood concentration-time profile was postponed up to 2 weeks after the last Tac QD adjustment.



Back to Top | Article Outline

Sample Size

Because no numerical data on the change in intrapatient variability in 24-hour drug exposure were available, no formal statistical power analysis could be performed. The study population was set at 40 patients based on common practice in other PK studies (12).

Back to Top | Article Outline

Study Parameters/Endpoints

The primary objective of this study was to evaluate and compare the intrapatient variability of AUC0–24 of Tac BID and Tac QD in stable renal transplant patients. Secondary endpoints were correlation of the PK profiles with the CYP3A5 genotype and comparison of the intrapatient variability of the trough levels (Cmin).

Back to Top | Article Outline

Assessments and Assay

A total of five AUCs were arbitrarily considered adequate for the assessment of the intrapatient variability for each formulation. Each PK profile consisted of eight blood samples that were collected before dosing and at 2, 4, 8, 12, 14, 16, and 24 hours after dosing by means of the DBS method. This eight-point 24-hour PK profile was assumed to give a valid AUC for both Tac BID and Tac QD.

As for Tac BID, earlier reports guarantee prediction of the AUC0–12 when samples at t=0 hour, t=2 hours, and t=4 hours are included (15). For the AUC0–24, we added the corresponding time points after the evening dose, namely, t=12 hours, t=14 hours, and t=16 hours. We extended this six-point strategy to eight points by adding t=8 hours and t=24 hours to get more information about the elimination phase. For the patient’s convenience, we did not add t=20 hours.

As for the Tac QD profiles, our center has previously participated in a local board-approved phase 3 trial (METC code 03-102; EudraCT no. 2005-005714-20) in which 27 stable renal transplant patients provided a venous 24-hour PK profile under fasting conditions. This profile consisted of 10 time points at t=0, 1, 2, 3, 4, 6, 8, 10, 12, and 24 hours. The eight-point AUC was then validated by means of linear regression: this model revealed a maximum R2 of 0.995 when the samples at t=2, 4, 8, 12, and 24 hours were included. An AUC was regarded as adequate in case at least six valid blood spot samples were available and if none of the above-mentioned crucial sample or the trough level were missing.

If a profile did not fulfill these criteria, it had to be repeated 1 week later. In addition, patients visited our outpatient clinics before the start of the study, 1 week after conversion, and at the end of the study (Fig. 4). During these outpatient visits, regular venous blood and urine samples were drawn for routine posttransplantation medical care and safety reasons, including determination of a venous tacrolimus Cmin. These data were not included in the PK analysis. To rule out potential noncompliance, medication intake had to be registered and was controlled by pill count for both formulations.

The patients themselves collected the required blood samples by means of the DBS method at home, with strict recording of dosing and sampling time points. This method has been described and validated before (11). In brief, fingerprick blood samples were collected using a contact-activated lancet (Becton, Dickinson and Company, Franklin Lakes, NJ). The first two drops were collected from the lateral part of the fingertip to fill two 8-mm premarked circles, for duplicate sampling, on the sampling paper (Item No. 10535097; Whatman, Schleicher and Schuell, Dassel, Germany). The blood spots were allowed to dry at room temperature for at least 10 min before packaging and were stored in a sealed plastic bag containing silicone granules. After completion of the 24-hour profile, all samples were collectively sent by regular post to our laboratory where they were stored at 4°C until further analysis. It has been proven that tacrolimus in dry blood spot is stable during postal transport (16, 17). Before the start of the study, patients were trained in blood spot sampling, and the study was initiated only after all patients had repeatedly sent in a blood spot sample that had passed the quality control.

Each of the two blood spots on the sampling paper was analyzed on separate analytical runs, and the average value of both analyses was the reported value of the tacrolimus concentration. If the two results diverged by more than 30%, this value was reported as missing and was not considered. If one analytical run was invalid because of a poor quality control, we only disposed of one instead of two values. In case two runs had a missing noncrucial value at the same single time point, the average concentration at that particular time point was revealed by interpolation after natural logarithmic transformation. If more samples were missing because of analytical problems, the run was redone by analyzing the diluted extraction solution. In case a crucial sample was missing at a time point as defined above, the PK profile had to be repeated.

Analysis of the concentration of tacrolimus started with punching out the dried blood spots and adding an extraction solution. The prepared samples were analyzed in the same way as the routine tacrolimus assay in venous blood. The tacrolimus assay is validated according to, and complied with, the ‘“Bioanalytical Methods Validation, Guidance for Industry”’ (18), with a lower limit of quantification of 1.0 μg/L (11). The laboratory participated in the international tacrolimus proficiency testing scheme ( The concentration of tacrolimus was determined by means of tandem mass spectrometry. Chromatography was performed using an Alliance 2795 system (Waters Ltd., Watford, UK). The mass spectrometer used in our study was a Quattro Micro (Micromass, Manchester, UK).

Back to Top | Article Outline


Each patient provided a venous blood sample at the screening visit of this study for genetic analysis. Genomic DNA was extracted according to the manufacturer’s instructions (Qiagen, Leusden, The Netherlands). Real-time polymerase chain reaction fluorescence resonance transfer assays were used for genotyping with the LightCycler (Roche Diagnostics, Almere, The Netherlands).

Back to Top | Article Outline


Demographic data were presented as mean and range and PK parameters were presented as geometric mean and 90% confidence interval (CI). AUC was calculated using the linear trapezoidal rule. Before statistical analysis, PK data were natural log-transformed and transformed back to the original scale for display of the data.

Intrapatient variability was calculated through the CV as follows:

where SWR is the within-subject standard deviation of the natural log-transformed values.

PK parameters were compared with the paired-measurements t test; Cyp3A5 polymorphism was analyzed with nonparametric tests (Wilcoxon signed rank test for paired measurements and Mann-Whitney U test for nonpaired measurements).

P<0.05 was considered statistically significant. Statistical analysis was performed using SPSS 16.0 for Windows (release 16.0.1).

Back to Top | Article Outline


The authors thank their research nurse, Mrs. Monique Mullens, for her support during the study. The study would not have been feasible otherwise.

Back to Top | Article Outline


1. Alloway R, Steinberg S, Khalil K, et al. Conversion of stable kidney transplant recipients from a twice daily Prograf-based regimen to a once daily modified release tacrolimus-based regimen. Transplant Proc 2005; 37: 867.
2. van Hooff J, Van der Walt I, Kallmeyer J, et al. Pharmacokinetics in stable kidney transplant recipients after conversion from twice-daily to once-daily tacrolimus formulations. Ther Drug Monit 2012; 34: 46.
3. Budde K, Glander P. Pharmacokinetic principles of immunosuppressive drugs. Ann Transplant 2008; 13: 5.
4. Florman S, Alloway R, Kalayoglu M, et al. Conversion of stable liver transplant recipients from a twice-daily Prograf-based regimen to a once-daily modified release tacrolimus-based regimen. Transplant Proc 2005; 37: 1211.
5. Borra LC, Roodnat JI, Kal JA, et al. High within-patient variability in the clearance of tacrolimus is a risk factor for poor long-term outcome after kidney transplantation. Nephrol Dial Transplant 2010; 25: 2757.
6. Kuypers DR, Peeters PC, Sennesael JJ, et al. Improved adherence to tacrolimus once-daily formulation in renal recipients: a randomized controlled trial using electronic monitoring. Transplantation 2013; 95: 333.
7. Wu MJ, Cheng CY, Chen CH, et al. Lower variability of tacrolimus trough concentration after conversion from Prograf to Advagraf in stable kidney transplant recipients. Transplantation 2011; 92: 648.
8. Cheung CY, Op den Buijsch RA, Wong KM, et al. Influence of different allelic variants of the CYP3A and ABCB1 genes on the tacrolimus pharmacokinetic profile of Chinese renal transplant recipients. Pharmacogenomics 2006; 7: 563.
9. Lee SJ, Usmani KA, Chanas B, et al. Genetic findings and functional studies of human CYP3A5 single nucleotide polymorphisms in different ethnic groups. Pharmacogenetics 2003; 13: 461.
10. van Schaik RH, van der Heiden IP, van den Anker JN, et al. CYP3A5 variant allele frequencies in Dutch Caucasians. Clin Chem 2002; 48: 1668.
11. Hoogtanders K, van der Heijden J, Christiaans M, et al. Therapeutic drug monitoring of tacrolimus with the dried blood spot method. J Pharm Biomed Anal 2007; 44: 658.
12. Wlodarczyk Z, Squifflet JP, Ostrowski M, et al. Pharmacokinetics for once- versus twice-daily tacrolimus formulations in de novo kidney transplantation: a randomized, open-label trial. Am J Transplant 2009; 9: 2505.
13. Yong Chung J, Jung Lee Y, Bok Jang S, et al. CYP3A5*3 genotype associated with intrasubject pharmacokinetic variation toward tacrolimus in bioequivalence study. Ther Drug Monit 2010; 32: 67.
14. Pashaee N, Bouamar R, Hesselink DA, et al. CYP3A5 genotype is not related to the intrapatient variability of tacrolimus clearance. Ther Drug Monit 2011; 33: 369.
15. Op den Buijsch RA, van de Plas A, Stolk LM, et al. Evaluation of limited sampling strategies for tacrolimus. Eur J Clin Pharmacol 2007; 63: 1039.
16. Cheung CY, van der Heijden J, Hoogtanders K, et al. Dried blood spot measurement: application in tacrolimus monitoring using limited sampling strategy and abbreviated AUC estimation. Transpl Int 2008; 21: 140.
17. Hoogtanders K, van der Heijden J, Christiaans M, et al. Dried blood spot measurement of tacrolimus is promising for patient monitoring. Transplantation 2007; 83: 237.
18. In: Agency EM, ed. Note for Guidance on the Investigation of Bioavailability and Bioequivalence (NfG on BA/ BE). London; 2010.

Tacrolimus; Pharmacokinetics; Therapeutic drug monitoring; Intra-patient coefficient of variability; Drug exposure; Formulation

© 2014 by Lippincott Williams & Wilkins