Mycophenolic acid (MPA), the active constituent of mycophenolate mofetil (MMF) and enteric-coated mycophenolate sodium (EC-MPS), plays an important role in the triple immunosuppressant regimen in renal transplant recipients.1 MMF has an excellent short- and long-term efficacy, which has been clinically proven for years.2 However, for some patients, MMF would cause unfavorable side effects like gastrointestinal complications, which might lead to the early dose reduction and even discontinuation of MMF. As a result, the very low MPA exposure would increase the risk of rejection and graft loss.3 It was demonstrated that 720 mg EC-MPS delivered bioequivalent mean MPA exposure compared with 1000 mg MMF.4,5 Due to the specially designed dosage form, EC-MPS delays the release of MPA in the small intestine, reducing the gastrointestinal burden and alleviating the gastrointestinal complications, and thus, it might lower the risk of rejection and graft loss because of the insufficient MPA exposure.6–9
The 12-hour area under the concentration–time curve (AUC0–12 h) of MPA is regarded as the best parameter to reflect the exposure of the drug.10 The AUC0–12 h is tightly related to the clinical outcomes after the transplantation and the risk of side effects, and that is why, therapeutic drug monitoring (TDM) is needed.11 A range between 30 and 60 mg·h·L−1 of AUC0–12 h has been recommended for renal transplant recipients, and the drug dosage should be adjusted through TDM to achieve the best clinical outcomes.12–14 However, more than 10 samples at different time points over 12 hours are needed to calculate the accurate AUC0–12 h with the regular method and that would be impractical for both patients and staff of the hospital.15 Therefore, limited sampling strategy (LSS) is a relatively feasible and practicable method for TDM of MPA.
In the past decades, it has been proven that LSS could provide a good estimation of the MPA AUC0–12 h for MMF, and this technique has been applied in many transplant centers.16 But when it comes to the EC-MPS, its special dosage form causes a more unpredictable absorption profile than MMF, and the huge variability makes the difficulty for drug concentration monitoring.17 Although several studies investigated the application of LSS for EC-MPS, there was still a paucity of data, especially for the Chinese population.9,14–16,18–22
The development of a model that balances precision and practicability to estimate the MPA exposure of EC-MPS is highly clinically appreciated. This study is aimed to search for the best equation with the least time points based on the profiles of Chinese renal transplant recipients receiving EC-MPS and tacrolimus.
METHODS AND MATERIALS
A total of 36 renal transplant patients were recruited for this study from 3 transplant centers: 19 patients from Zhongshan Hospital, Fudan University, Shanghai; 9 patients from Changzheng Hospital, Second Military Medical University, Shanghai; and 8 patients from Shanghai First People's Hospital, Shanghai Jiao Tong University, Shanghai. All the patients were adults (>18 years) receiving the first-time living-related renal transplantation, and an informed consent was obtained before participation. All donors had a documented linear blood relationship with the recipients, and the transplantation was carried out in accordance with the Declaration of Helsinki and with the 2007 Chinese “Regulation on Human Organ Transplantation.” Exclusion criteria included retransplantation, combined transplantation, ABO blood incompatibility transplantation, pregnancy, allergy to EC-MPS, severe liver/lung/heart diseases, active infection, tumor, or mental diseases. The immunosuppressant regimen was tacrolimus + EC-MPS + prednisone. The initial dose of EC-MPS was 720 mg every 12 hours, and it was adjusted based on the doctors' experience and patients' clinical manifestations. The study design was approved by the ethics committee of Zhongshan Hospital, Fudan University (No. B2012-109). The baseline characteristics of patients are shown in Table 1.
MPA Detection and AUC Calculation
At 2 weeks after the transplantation, blood specimens (2 mL each time) were collected predose and at 0.5, 1, 1.5, 2, 2.5, 3, 4, 6, 8, and 12 hours after the morning dose of EC-MPS. All the specimens were stored at 4°C overnight and were analyzed with enzyme-multiplied immunoassay technique (EMIT) in the clinic laboratory of the Zhongshan Hospital. The measured AUC was calculated with these 11 points of MPA concentrations with the linear trapezoidal rule.
The patients were randomly divided into 2 groups: the model group and the validation group. The randomization was performed by a computer treatment assignment in which randomized numbers were generated with the equation RAND(). LSS was used to build equations with the data of the model group. The MPA concentration of each time point was analyzed with linear regression, and the correlation coefficient (R2), standard error (SE) of estimation, residual and its mean square (ME), and P value were calculated separately. Considering precision and the clinical practicability, multiple linear regression with the method “forward” was done using MPA concentrations of 3 and 4 time points. Besides R2, prediction error (PE)% or bias, and absolute PE% or precision were introduced to evaluate the equations.15,23 Among them, a precision that was less than 15% was considered clinically acceptable.24 Bias and precision were defined as follows:
To further assess the external consistency of the equations, they were validated in the validation group. Time points of the equations were used as independent variables for multiple regression with the data of the validation group, and correlation coefficients were calculated. The equations from model group were used to calculate the estimated AUC, and bias and precision were also applied as mentioned above. The Bland–Altman test was used to evaluate the agreement between the measured AUC and estimated AUC, and the fixed range was defined as mean ± 1.96 SD. Correlation between estimated and measured AUC was depicted with scatter diagram.
Continuous variables (expressed as mean ± SD) were compared using the t test, and categorical variables were compared using the χ2 test. A value of P < 0.05 was considered to be statistically significant. Statistical analysis was performed using SPSS 18.0 for Windows (SPSS, Inc, Chicago, IL).
The baseline characteristics of the patients are listed in Table 1. The age, sex, body weight, the function of liver, the tacrolimus doses, and tacrolimus trough concentration showed no significant difference between the 2 groups. The estimated glomerular filtration rates were 68.8 ± 35.2 for the model group and 57.0 ± 24.3 for the validation group, which also showed no significant difference (P = 0.404).
The average dose of EC-MPS for all patients was 600.0 ± 121.7 mg every 12 hours, and the measured MPA AUC0–12 h was 43.04 ± 16.56 mg·h·L−1. The mean MPA Cmax was 10.7 ± 7.8 mg/L, and the maximum MPA concentration occurred at 4.8 ± 2.8 hours after the morning dose of EC-MPS, which was in accordance with the delayed absorption of the drug. The mean MPA concentration of each group is depicted in Figure 1. The curves present the same trend, where they rise sharply between 2.5 and 3 hours and reach the top at 4 hours (about 6 mg/L) and then decrease gently after 6 hours. There is only 1 apex obviously observed in the curves, but within 3–5 hours, they all keep at the peak level.
Details of pharmacokinetic profiles are shown in Table 2 and Figure 2. Although the mean MPA concentration of each group presents the same trend, the individuals are in great variability.
The correlations between MPA concentration at each time point and the measured AUC are shown in Table 3. The concentration at 4 hours has the strongest correlation with the full AUC (R2 = 0.580; adjusted R2 = 0.553; SE = 12.767; residual = 2607.789; P < 0.001). Others have poor estimation for full AUC.
Using the stepwise multiple linear regression, equations consisting of 3 and 4 time points were developed with the highest correlation coefficient (Table 4). For 3 time points, the equation is 6.629 + 8.029 × C0 + 0.592 × C3 + 1.786 × C4 (R2 = 0.910; adjusted R2 = 0.891; SE = 6.317; residual = 558.589; P < 0.001); for 4 time points, it is 3.132 + 5.337 × C0 + 0.735 × C3 + 1.783 × C4 + 3.065 × C8 (R2 = 0.959; adjusted R2 = 0.946; SE = 4.444; residual = 256.748; P < 0.001).
As shown in Table 5, when evaluated in the validation group, the 4 time point model has a much better performance than the 3 time point model [the 4 time point model: R2 = 0.873, bias = 0.505 (95% CI, −10.159 to 11.170), precision = 13.370 (95% CI, 5.186–21.555), and 77.8% of estimated AUCs was within 85%–115% of the measured AUCs; for the 3 time point model: R2 = 0.573, bias = 6.196 (95% CI, −10.627 to 23.018), precision = 21.286 (95% CI, 8.079–34.492), and 50.0% of estimated AUCs was within 85%–115% of the measured AUCs. In the Bland–Altman test, only 1 plotted difference exceeds the fixed range of the mean ± 1.96 SD in each model, but model 2 has a much better internal consistency (Fig. 3). The correlation between estimated and measured AUC in Figure 4 and the mountain plot of models 1 and 2 in Figure 5 also suggest a better estimation of model 2.
In this study, the pharmacokinetic profiles of the Chinese renal transplant recipients receiving EC-MPS were analyzed with LSS using EMIT. Compared with high-performance liquid chromatography, which was considered as the “gold standard” for drug concentration measurement, EMIT was reported for higher results. However, EMIT had a good linear correlation with high-performance liquid chromatography, making it a practical method for drug monitoring.25–28 More importantly, due to the high efficiency, convenience, and automation of EMIT, it has been more widely used in clinical setting. In this study, all samples were tested with EMIT, which should be considered when adjusting drug doses.
In accordance with the previous studies, large interindividual variability was shown in MPA exposure.20,29 MPA is metabolized by the UDP-glucuronosyltransferase (UGT) to a major metabolite, 7-O-glucuronide, and 7-O-glucuronide undergoes biliary excretion into the intestine via multidrug resistance protein 2 and solute carrier organic anion transporter.30 A series of studies have proven that the genetic polymorphisms of UGT and SLCO influence the pharmacokinetics of MPA.31 Food intake time, types of food ingestion, and gastric emptying may also have influence on it.9 Although for EC-MPS drug concentration monitoring is more difficult, it still highlights the value and importance of TDM. Various studies have shown the relationship between the MPA AUC and the risk of rejection and side effects. With TDM, the patients could receive personalized prescription.
Despite the interindividual variability, it shared the same trend within each group. The concentration of MPA increased sharply between 2.5 and 3 hours postdose, and the maximum occurred about 4 hours after the oral administration (peak time point: 4 hours postdose and mean MPA-Tmax: 4.8 hours postdose), which was in accordance with the majority of the individual profiles of patients. It was delayed compared with a previous study, which has reported that the median time to maximum MPA concentration was 2.0 hours.32 This result might reflect the metabolic characteristics of EC-MPS in Chinese population. There was only 1 obvious peak in the curves of our study. This may be due to the limitations of LSS. Within the range of 3–6 hours postdose, the inadequate sample points may conceal the second peak, which was also seen in other research.15
The results of univariate correlation analysis suggested that the predose MPA concentration was poor at predicting the systemic exposure of MPA (R2 = 0.187). Only the concentration at 4 hours postdose, which was also the peak of the concentration, had a relatively higher correlation, but it was still not reliable enough (R2 = 0.553). As a result, 3 time point (C0, C3, and C4) and 4 time point (C0, C3, C4, and C8) equations were developed. Although the frequency of sampling was acceptable, the time span needed was still impractical for outpatients. In our study, the same multiple regression method was performed with the variables within 2 hours postdose (C0, C0.5, C1, C1.5, and C2) so that it might have better practicability, but all parameters were excluded except C1.5 because of high P value (P > 0.10, data not shown). Fleming et al15 did the same effort, but the R2 for the early 4 points within 2 hours was only 0.292. De Winter et al and Pawinski et al also wanted to develop an equation using the time points within 3 hours postdose, but they failed because of biased and imprecise results.16,33 When C8 was added into the 3 time point equation in our study, it had a much better performance in the validation group. Because the reported enterohepatic recirculation was about 3–12 hours postdose and caused the second peak of MPA, the C8 was related to this phenomenon so that it greatly increased the precision of the equation.2,32 It is the pharmacokinetic characteristics of EC-MPS that decreases the convenience for TDM, especially for the outpatients.
A series of studies with similar parameters to evaluate the equations are shown in Table 6.9,15,16,19,22,33 It is in accordance with our results that a relatively feasible and precise equation should contain time points at a later stage. Compared with the results in Table 6, the equations developed in our study have a relatively better predictive performance, especially for the 4 time point equation, whose bias and precision for the validation group were 0.505 and 13.370, respectively. Considering that the SE of estimation was only 4.444, and about 80% of the estimated AUC was within 85%–115% of the measured AUC, this equation could be the reference for the treatment of inpatients in clinical setting.
This study has a number of limitations that should be considered. All cases were from a Chinese population, and the genetic polymorphism for MPA metabolism was not analyzed. Blood samples were stored at 4°C overnight for the practicability. This could cause a measurement error. The number of sampling time points for the measured AUC is limited, especially between 2 and 4 hours postdose, in which the MPA concentration increased sharply. Renal function of kidney transplant recipient achieves a stable condition mostly 2 weeks after the transplantation, and the pharmacokinetics of EC-MPS is relatively stable at this time point as well. Additionally, inappropriate MPA exposure in early stage after the kidney transplantation correlated to allograft rejection, infection, and myelosuppression; thus, MPA pharmacokinetic monitoring at an early stage after the kidney transplantation is of great importance. Based on these considerations, we chose 2 weeks after transplantation for the study time point. However, beyond 2 weeks after the transplantation, there are still important changes in dose-corrected MPA exposure, which could potentially change the overall performance of the currently obtained 4 point LSS at other (later) time points after the transplantation. Further validation is needed before LSS calculated by this predictive equation can be applied to patients receiving long-term EC-MPS. Moreover, the relatively small number of patients involved in this study amplified the bias. The models should thus be further tested with larger patient groups in more centers.
In summary, we developed optimal equations for the estimation of full MPA AUC0–12 h in the Chinese renal transplant recipients receiving EC-MPS and tacrolimus. The best 3 time point equation was 6.629 + 8.029 × C0 + 0.592 × C3 + 1.786 × C4, whereas the best 4 time point equation was 3.132 + 5.337 × C0 + 0.735 × C3 + 1.783 × C4 + 3.065 × C8. The latter one exhibited much better estimation ability than the former one. The long time span needed restricted its application, especially for the outpatients, but it could be a useful tool to guide the personalized prescription for the inpatients.
1. Halloran PF. Immunosuppressive drugs for kidney transplantation. N Engl J Med. 2004;351:2715–2729.
2. Bullingham RE, Nicholls AJ, Kamm BR. Clinical pharmacokinetics of mycophenolate mofetil. Clin Pharmacokinet. 1998;34:429–455.
3. Bunnapradist S, Ambuhl PM. Impact of gastrointestinal-related side effects on mycophenolate mofetil dosing and potential therapeutic strategies. Clin Transpl. 2008;22:815–821.
4. Johnston A, He X, Holt DW. Bioequivalence of enteric-coated mycophenolate sodium and mycophenolate mofetil: a meta-analysis of three studies in stable renal transplant recipients. Transplantation. 2006;82:1413–1418.
5. Budde K, Bauer S, Hambach P, et al. Pharmacokinetic and pharmacodynamic comparison of enteric-coated mycophenolate sodium and mycophenolate mofetil in maintenance renal transplant patients. Am J Transpl. 2007;7:888–898.
6. Burg M, Saemann MD, Wieser C, et al. Enteric-coated mycophenolate sodium reduces gastrointestinal symptoms in renal transplant patients. Transpl Proc. 2009;41:4159–4164.
7. Ortega F, Sanchez-Fructuoso A, Cruzado JM, et al. Gastrointestinal quality of life improvement of renal transplant recipients converted from mycophenolate mofetil to enteric-coated mycophenolate sodium drugs or agents: mycophenolate mofetil and enteric-coated mycophenolate sodium. Transplantation. 2011;92:426–432.
8. Langone AJ, Chan L, Bolin P, et al. Enteric-coated mycophenolate sodium versus mycophenolate mofetil in renal transplant recipients experiencing gastrointestinal intolerance: a multicenter, double-blind, randomized study. Transplantation. 2011;91:470–478.
9. Yao X, Huang H, Wei C, et al. Limited sampling strategy for mycophenolic acid in Chinese kidney transplant recipients receiving enteric-coated mycophenolate sodium and tacrolimus during the early posttransplantation phase. Ther Drug Monit. 2015;37:516–523.
10. Shaw LM, Holt DW, Oellerich M, et al. Current issues in therapeutic drug monitoring of mycophenolic acid: report of a roundtable discussion. Ther Drug Monit. 2001;23:305–315.
11. Kuypers DR, Le Meur Y, Cantarovich M, et al. Consensus report on therapeutic drug monitoring of mycophenolic acid in solid organ transplantation. Clin J Am Soc Nephrol. 2010;5:341–358.
12. Kuypers DR, De Jonge H, Naesens M, et al. Current target ranges of mycophenolic acid exposure and drug-related adverse events: a 5-year, open-label, prospective, clinical follow-up study in renal allograft recipients. Clin Ther. 2008;30:673–683.
13. Van Gelder T, Le Meur Y, Shaw LM, et al. Therapeutic drug monitoring of mycophenolate mofetil in transplantation. Ther Drug Monit. 2006;28:145–154.
14. Sommerer C, Muller-Krebs S, Schaier M, et al. Pharmacokinetic and pharmacodynamic analysis of enteric-coated mycophenolate sodium: limited sampling strategies and clinical outcome in renal transplant patients. Br J Clin Pharmacol. 2010;69:346–357.
15. Fleming DH, Mathew BS, Prasanna S, et al. A possible simplification for the estimation of area under the curve (AUC(0)(-)(1)(2)) of enteric-coated mycophenolate sodium in renal transplant patients receiving tacrolimus. Ther Drug Monit. 2011;33:165–170.
16. Pawinski T, Luszczynska P, Durlik M, et al. Development and validation of limited sampling strategies for the estimation of mycophenolic acid area under the curve in adult kidney and liver transplant recipients receiving concomitant enteric-coated mycophenolate sodium and tacrolimus. Ther Drug Monit. 2013;35:760–769.
17. De Winter BC, Van Gelder T, Glander P, et al. Population pharmacokinetics of mycophenolic acid: a comparison between enteric-coated mycophenolate sodium and mycophenolate mofetil in renal transplant recipients. Clin Pharmacokinet. 2008;47:827–838.
18. Yang SL, Gao X, Wang QH, et al. Use of limited sampling strategy for estimating area under concentration-versus-time curve of mycophenolate sodium in renal allograft recipients [in Chinese]. Zhonghua Yi Xue Za Zhi. 2013;93:3841–3846.
19. Sanchez Fructuoso AI, Perez-Flores I, Calvo N, et al. Limited-sampling strategy for mycophenolic acid in renal transplant recipients reciving enteric-coated mycophenolate sodium and tacrolimus. Ther Drug Monit. 2012;34:298–305.
20. Qiu K, Tian H, Wang W, et al. Pharmacokinetics of enteric-coated mycophenolate sodium in Chinese renal transplantation
recipients. Chin Med J (Engl). 2012;125:4226–4232.
21. Shah T, Tellez-Corrales E, Yang JW, et al. The pharmacokinetics of enteric-coated mycophenolate sodium and its gastrointestinal side effects in de novo renal transplant recipients of Hispanic ethnicity. Ther Drug Monit. 2011;33:45–49.
22. Capone D, Tarantino G, Kadilli I, et al. Evalutation of mycophenolic acid systemic exposure by limited sampling strategy in kidney transplant recipients receiving enteric-coated mycophenolate sodium (EC-MPS) and cyclosporine. Nephrol Dial Transpl. 2011;26:3019–3025.
23. Sheiner LB, Beal SL. Some suggestions for measuring predictive performance. J Pharmacokinet Biopharm. 1981;9:503–512.
24. Wong KM, Shek CC, Chau KF, et al. Abbreviated tacrolimus area-under-the-curve monitoring for renal transplant recipients. Am J Kidney Dis. 2000;35:660–666.
25. Martiny D, Macours P, Cotton F, et al. Reliability of mycophenolic acid monitoring by an enzyme multiplied immunoassay technique. Clin Lab. 2010;56:345–353.
26. Chen B, Gu Z, Chen H, et al. Establishment of high-performance liquid chromatography and enzyme multiplied immunoassay technology methods for determination of free mycophenolic acid and its application in Chinese liver transplant recipients. Ther Drug Monit. 2010;32:653–660.
27. Irtan S, Azougagh S, Monchaud C, et al. Comparison of high-performance liquid chromatography and enzyme-multiplied immunoassay technique to monitor mycophenolic acid in paediatric renal recipients. Pediatr Nephrol. 2008;23:1859–1865.
28. Premaud A, Rousseau A, Le Meur Y, et al. Comparison of liquid chromatography-tandem mass spectrometry with a commercial enzyme-multiplied immunoassay for the determination of plasma MPA in renal transplant recipients and consequences for therapeutic drug monitoring. Ther Drug Monit. 2004;26:609–619.
29. Li J, Liu Y, Huang J, et al. Evaluation of mycophenolic acid exposure using a limited sampling strategy in renal transplant recipients. Am J Nephrol. 2013;37:534–540.
30. Bernard O, Tojcic J, Journault K, et al. Influence of nonsynonymous polymorphisms of UGT1A8 and UGT2B7 metabolizing enzymes on the formation of phenolic and acyl glucuronides of mycophenolic acid. Drug Metab Dispos. 2006;34:1539–1545.
31. Han N, Yun HY, Kim IW, et al. Population pharmacogenetic pharmacokinetic modeling for flip-flop phenomenon of enteric-coated mycophenolate sodium in kidney transplant recipients. Eur J Clin Pharmacol. 2014;70:1211–1219.
32. Budde K, Glander P, Diekmann F, et al. Review of the immunosuppressant enteric-coated mycophenolate sodium. Expert Opin Pharmacother. 2004;5:1333–1345.
33. De Winter BC, Van Gelder T, Mathot RA, et al. Limited sampling strategies drawn within 3 hours postdose poorly predict mycophenolic acid area-under-the-curve after enteric-coated mycophenolate sodium. Ther Drug Monit. 2009;31:585–591.