|Home | About | Journals | Submit | Contact Us | Français|
Mycophenolic acid (MPA)is the active component of mycophenolate mofetil (MMF). Low MPA exposure is associated with a higher incidence of acute GVHD and possibly worse engraftment. Therapeutic plasma targets have been proposed in hematopoietic cell transplantation (HCT), however, are difficult to achieve in adult patients with MMF doses of 2 g/day. Mycophenolate pharmaco-kinetics was prospectively studied in adults undergoing nonmyeloablative HCT who received MMF 3 g/day with CYA. The first 15 individuals received 1.5 g every 12 h and the second 15 received 1 g every 8 h. Sampling was performed in each patient with i.v. and oral administration. There were no differences in total or unbound MPA 24-h cumulative area under the curves (AUCs), concentrations at steady state (Css)or troughs between the two dosing regimens (all P>0.01). The previously proposed total MPA Css target of 3 μg/ml and trough ≥1 μ/ml were achieved in only 13–27% and 20–53% of patients, respectively, on 3 g/day. However, the 3 g/day regimens readily achieved satisfactory unbound 24-h cumulative AUC targets of 0.600 μg*h/ml in 87–100% of subjects. There appears to be no significant difference in daily MPA exposure when MMF of 3 g/day is divided into two or three equal doses.
Mycophenolic acid (MPA) is the active form of mycophenolate mofetil (MMF) which is a common component of the immunosuppressive regimens after nonmyeloablative allogeneic hematopoietic cell transplantation (HCT). MMF is most commonly combined with a calcineurin inhibitor for promotion of engraftment and prevention of GVHD after nonmyeloablative HCT.1–3 MMF is rapidly hydrolyzed by esterase in the blood, gut wall, liver and tissues to MPA.4,5 MPA is then glucuronidated hepatically and extrahepatically by UDP-glucuronosyl transferase (UGT) enzymes to the primary inactive metabolite, MPA glucuronide (MPAG).4,6 MPAG is excreted into the urine and gut through the bile. Available MPAG in the gut is de-glucuronidated back into MPA and reabsorbed through enterohepatic recycling. MPA is 97% bound to plasma albumin and possibly other proteins.7 Only unbound MPA is pharmacologically active.
Mycophenolate pharmacokinetics are altered in HCT recipients, resulting in lower MPA exposures relative to kidney transplant recipients receiving the same dose.4,8–13 As a result, doses commonly used in kidney recipients may not be optimal in HCT recipients. MPA exposure–response relationships have been extensively studied in organ transplantation and have been extensively reviewed else-where. 14–17 Although conflicting reports have appeared, a recent randomized trial comparing fixed dose MMF (1 g twice daily) to MPA concentration controlled shows that acute rejection and biopsy-proven acute rejection rates are lower in kidney recipients who are concentration-controlled. 18 In addition, the recent randomized Opticept trial in kidney transplantation (n=720) showed that MMF dosed by concentration control with low doses of calcineurin inhibitor was not inferior to fixed dose MMF (1 g b.i.d.) and standard doses of calcineurin inhibitors.19 These data for the first time showed that minimization of calcineurin inhibitors and steroids could be accomplished when the mycophenolate was optimized through concentration-controlled dosing. In both studies, concentration-controlled patients received higher MMF doses than the fixed dose arms. We previously showed that lower unbound MPA concentrations were associated with higher rates of acute GVHD, and low total MPA trough concentrations were associated with poorer engraftment after nonmyeloablative HCT.13 Giaccone et al.12 showed that low total MPA exposure was associated with lower donor T-cell chimerism. On the basis of these data, therapeutic plasma targets have been proposed.
In our previous study, approximately 50% of the patients fell below the proposed unbound MPA target with the standard MMF dose of 2 g/day.13 Our standard dose of MMF was increased to 3 g/day; however, the dosing regimen (1 g every 8 h or 1.5 g every 12 h) that would best achieve the targets was not known. Considering logistical issues and convenience, the use of every 12-h dosing was desirable, but given the rapid half-life of MPA it was not known if this regimen would achieve sufficient exposures. The objective of this study was to compare pharmacokinetic measures in patients receiving MMF 1 g every 8 h and 1.5 g every 12 h and to determine whether the higher doses would achieve the proposed therapeutic targets.
A total of 30 adult patients with advanced or high-risk hematologic malignancies who underwent HCT were studied. Patients were eligible if they were receiving a nonmyeloablative conditioning regimen, ≥18 years of age and had a pretransplant serum creatinine ≤2.0mg per 100 ml. No patients had received MMF within 2 weeks prior to transplantation. Women of childbearing potential had a negative pregnancy test prior to HCT. The protocol was approved by the University of Minnesota Cancer Protocol Review Committee and Institutional Review Board. All patients gave written informed consent.
Patients received CY 50 mg/kg on day –6 i.v. plus fludarabine 40 mg/m2/day beginning on day –6 for 5 days (total dose, 200 mg/m2), and TBI 200 cGy as a single fraction on day –1. Patients who had not received chemotherapy in the prior 6 months were also given equine antithymocyte globulin (15 mg/kg i.v. every 12 h, on days –6, –5 and –4). Filgrastim of 5 μg/kg was started when ANC was <1000 cells/μl and continued until ANC was more than 2500 cells/μl for 2 consecutive days.
All patients received prophylaxis with MMF and CYA. MMF (i.v., Cellcept, 3 g/day) was started on day –3 and converted to oral therapy (12×250 mg capsules per day) in week 1 as tolerated using a 1:1 i.v. to oral dose conversion. MMF (i.v.) was infused over 2 h. CYA was also initiated on day –3 at a dose of 2.5 mg/kg every 12 h i.v. and adjusted to maintain whole-blood HPLC trough concentrations of 200–400 ng/ml. All patients received antibiotic prophylaxis and blood product support as described previously.20
Two consecutive dosing cohorts were studied while inpatient for HCT. The first 15 subjects were administered MMF at 1.5 g every 12 h i.v. and orally, and the second 15 subjects were given 1 g every 8 h i.v. and orally. Pharmacokinetic sampling was performed on i.v. and oral therapy in each subject. Patients not able to convert to oral therapy in week 1 were subsequently dropped from the study and replaced. Doses were not modified based on pharmacokinetic results. Serum creatinine (SCr) and weight were obtained on the day of pharmacokinetics. Total bilirubin, alanine transaminase (ALT), albumin and CYA trough concentrations were obtained within 48 h of each pharmacokinetic sampling period. Creatinine clearance (CrCl) was estimated by the Cockcroft and Gault equation using ideal body weight.21 In patients whose actual body weight was less than the ideal weight, the actual body weight was used. Transplant-related outcome data was retrieved from the Blood and Marrow Transplant Program Database, which prospectively collects data for all patients receiving HCT at the University of Minnesota.
Pharmacokinetics was obtained over one steady-state dosing interval between days –1 and +5 post transplant while on i.v. therapy. The patient was converted to oral therapy and sampling was repeated over one steady-state dosing interval between days 5 and 14 post transplant. Steady state was defined as a minimum of three identical MMF doses given by the same route of administration. For i.v. therapy, blood samples were collected at 0 (immediately prior) and 2, 4, 6 and 8 h after the start of infusion. Subjects receiving every 12-h dosing were additionally sampled at 12 h. For oral therapy, blood samples were collected at 0 (immediately prior) and 1, 2, 4, 6 and 8 h after administration. Subjects receiving every 12-h dosing were also sampled at 12 h.
Blood samples of 5ml were collected from a central venous catheter into 10 ml purple top tubes containing EDTA as the anticoagulant and placed on wet ice. Samples were centrifuged at 2000 r.p.m. for 15 min within 1 h of collection. Each spun plasma sample was aliquoted into two 1.5 ml Eppendorf screw cap vials.
Plasma concentrations of total and unbound MPA and total MPAG were measured on each sample on an Agilent 1100 series (Agilent Technologies, Wilmington, DE, USA) HPLC system with UV detection as described previously.22 The validated assays were linear in the range of 0.025–10 μg/ml for total MPA, 1–100 μg/ml for total MPAG and 0.001–0.5 μg/ml for unbound MPA. Any sample above the linear range was diluted and re-analyzed. Assay accuracy and intra- and interday variability as percentage of relative standard deviation were 97–111%, 0.6–4.4% and 1.0–6.0% for total MPA; 97–101%, 0.9–1.4% and 0.7–4.8% for MPAG; and 96–118%, 0.7–5.4% and 0.5–7.8% for unbound MPA, respectively.
For each patient, steady-state concentration time data was analyzed with noncompartmental analyses using WinNonLin (version 5.2, Pharsight Corporation) software. AUC0–tau to the time of last measured concentration (8 or 12 h) was estimated using the log-linear trapezoidal method. Concentration at steady state (Css) was AUC0–tau divided by the dosing interval. A cumulative 24-h AUC exposure was calculated by multiplying the measured AUC0–12 by 2 (for the every 12-h regimen) or AUC0–8 by 3 (for the every 8-h regimen). Cmax was taken as the largest observed plasma concentration. Steady-state Cmin (trough) was the last sample drawn (8 or 12 h) post dose. The free fraction of MPA was determined by calculating the ratio of unbound MPA AUC to total MPA AUC.
The distribution of pharmacokinetic measures was described using medians and ranges. The Wilcoxon rank sum test was used to provide comparisons of independent pharmacokinetics measures between the dosing cohorts. The sign test was used to compare paired differences of the pharmacokinetic measures within the dosing cohorts. To compare the frequencies of patients achieving the therapeutic targets, χ2 or Fisher’s exact tests were used to compare proportions between the dosing cohorts and the sign test was used to compare proportions within dosing cohorts. Only P-values ≤0.01 were considered statistically significant due to the large number of comparisons. To assess the association between pharmacokinetic exposure measures and biochemical markers, univariate linear regression analysis was performed for total and unbound MPA. AUCs and troughs against CrCl, SCr, total bilirubin, ALT, albumin, actual body weight, height and body surface were area calculated using actual body weight.
Patient characteristics by MMF dosing cohort are summarized in Table 1. There were no differences in gender, age, weight, body surface area, conditioning regimen, stem cell source or donor type between the dosing cohorts (P≥0.11) (Table 1). There were no differences in CrCl, total bilirubin or CYA trough concentrations between the two dosing cohorts (all P≥0.02) (Table 2).
Total MPA pharmacokinetic comparisons are shown in Table 2 and concentration time profiles are given in Figure 1. Trough concentrations were slightly lower but not significant (P≥0.04) after i.v. dosing relative to oral dosing. The total MPA cumulative 24-h AUC for the 1 g every 8 h and 1.5 g every 12 h cohorts was not different. For i.v. dosing, the cumulative 24-h AUC after 1 g every 8 h was 55.59 (22.68–101.99), and after 1.5 g every 12 h it was 60.90 (35.89–127.24) μg*h/ml (P=0.34). For oral dosing, the cumulative 24 h AUC with 1 g every 8 h was 49.49 (31.71–77.71) μg*h/ml and with 1.5 g every 12 h it was 56.39 (35.40–97.28) μg*h/ml (P=0.43). Trough concentration targets ≥1 μg/ml were achieved in 20–53% of subjects and the proportion achieving that target was not different between the regimens (P≥0.06) (Table 3). The Css target of ≥3 μg/ml was uncommonly reached (13–27% of subjects) on either dosing regimens and was not different between the dosing regimens (P≥0.5) (Table 3).
Small secondary peaks between 6 and 12-h post-dose after oral therapy were observed after every 12-h and every 8 h dosing in only six and three subjects, respectively. After i.v. therapy, secondary peaks after every 12-h and every 8-h dosing were observed in only one subject in each group. The lack of secondary peaks is consistent with limited enterohepatic recycling (Figure 1). The intrapatient correlation between MPA AUC0–6 and AUC0–8 for the every 8-h cohort was high (r2≥0.90). The intrapatient correlation between AUC0–6 and AUC0–12 for the every 12-h cohort was similar (r2≥0.86, P≥0.1). These data also suggest lack of substantial recycling.
To determine whether any patient characteristics were associated with total exposures, linear regression analyses between the total MPA pharmacokinetic measures and SCr, CrCl, ALT, total bilirubin, height, weight, age and body surface area were conducted. All correlations were weak with an r2<0.5 (P≥0.25).
Unbound MPA pharmacokinetic results are shown in Table 2 and concentration time profiles given in Figure 1. Trough concentrations were slightly lower but not significant (P≥0.02) after i.v. dosing relative to oral dosing. The cumulative 24-h AUC for the 1 g every 8 h vs 1.5 g every 12-h cohort was not different (P≥0.22). For i.v. dosing, the cumulative unbound MPA 24-h AUC after 1 g every 8 h was 0.942 (0.389–1.722) μg*h/ml and with 1.5 g every 12 h it was 1.080 (0.610–2.194) μg*h/ml (P=0.25). For oral dosing, the unbound cumulative 24-h AUC on the 1 g every 8 h was 0.987 (0.542–1.505) μg*h/ml and with 1.5 g every 12 h was 0.964 (0.638–1.540) μg*h/ml (P=0.68). All patients achieved the target cumulative 24-h unbound AUC of 0.600 μg*h/ml with MMF 1.5 g every 12 h, whereas 13 of 15 (87%) of patients achieved the target on the 1 g every 8-h dosing regimen (P≥0.14).
MPA unbound fraction was not different between the dosing cohorts (P≥0.84) (Table 2). The median (range) unbound MPA fraction for the 30 pharmacokinetic sets was 1.80% (1.18–4.30).
To determine whether any of the common clinical parameters were associated with unbound exposures, linear regression analyses between the unbound MPA pharmacokinetic measures and CrCl, SCr, albumin, ALT, total bilirubin, height, weight, age and body surface area were conducted. All correlations were weak with an r2<0.5 (P≥0.26), suggesting that no single patient factor would be useful in estimating exposure. However, no biochemistry measures were substantially abnormal; therefore, these data cannot be extrapolated to subjects with abnormal kidney or hepatic function.
Total MPAG pharmacokinetic results are shown in Table 2 and concentration time profiles are given in Figure 1. MPAG trough concentrations in both dosing cohorts were significantly higher after oral compared with i.v. administration (P≤0.01). The cumulative 24-h MPAG AUC after 1 g every 8 h after i.v. and oral administration were 1396.26 (660.76–3473.03) and 1637.86 (1275.70–4213.55) μg*h/ml, respectively (P=0.04). The cumulative 24-h AUC after 1.5 g every 12 h after i.v. and oral administration was 1447.26 (995.19–2296.50) and 1679.86 (958.87–2842.61) μg*h/ml, respectively (P=0.02). Overall, the ratio of MPAG AUC to total MPA AUC was 30.85 (10.53–132.90) μg*h/ml. MPAG concentrations were higher in patients with lower CrCl concentrations. There was a modest correlation between MPAG AUCs, trough concentrations, and SCr and CrCl (r2=0.51–0.62).
All patients engrafted (ANC>500 cells/μl × 3 consecutive days) in the every 8 h (median (range) day of engraftment, day 10 (6–38)) and every 12 h (day 12 (0–31)) cohorts. Acute GVHD grades II–IV developed in three and six patients in the MMF every 8-h and every 12-h dosing cohorts, respectively. The relationships between MPA exposure, and engraftment and GVHD were not analyzed given the small sample size and heterogeneity of donor source.
We previously found that most patients administered MMF 2 g/day did not achieve the proposed therapeutic ranges.13 Our clinical protocols were modified to administer MMF 3 g/day, although it was not known whether this would achieve the desired targets and which dosing interval (every 12 or every 8 h) best achieved the targets. We found no significant differences in the pertinent MPA pharmacokinetic measures (total trough, total Css or 24-h cumulative unbound AUC) between 1 g every 8-h and 1.5 g every 12-h regimens.
Dosing regimens for mycophenolate in HCT have been primarily derived from data in kidney transplantation (reviewed elsewhere).4,8,14,17 With adoption of the standard starting mycophenolate doses used in kidney transplant, HCT recipients receive lower MPA exposure.9–13,23–27 We speculate that these differences are due to physiologic differences peri-transplant between the kidney and HCT recipients that alter drug disposition. These differences include renal function, chemotherapy effects (for example, mucositis, hepatic injury, poor calorie intake, cellular death and regrowth of normal tissue), prophylactic antibiotic use and higher severity of illness, which may alter bioavailability, metabolism, excretion and/or enterohepatic recirculation. Data support the notion of chemotherapy-induced changes in MPA disposition. Fifty-five HCT recipients were given MMF pretransplant and again in week 1 post transplant.13 Total and unbound MPA AUCs were higher in pretransplant than in week 1 post transplant, and pre- and post-MPA AUCs were weakly correlated (r2≤0.19). We and others have studied the pharmacokinetics of mycophenolate in HCT recipients.9–13,23–27 Most data show that MPA exposures are 30–50% lower in HCT recipients compared to organ transplant recipients receiving the same dose when combined with CYA. MPA oral bioavailability is also lower in HCT recipients relative to other populations.5,25,28 We recently reported that children undergoing myeloablative HCT require higher MMF doses relative to pediatric kidney recipients to achieve comparable unbound MPA exposures.29
MPA exposure–response relationships in combination with CYA after nonmyeloablative HCT recipients have been evaluated in two studies.12,13 The data support a minimal required exposure for optimal outcomes; however, differ on the most pertinent pharmacokinetic measures (trough, AUC, Css; total or unbound). Eighty-seven adult patients undergoing HCT were treated with a preparative regimen of fludarabine (200 mg/m2, n=80) or BU (8 mg/kg, n=7), plus CY (50 mg/kg) and TBI (200 cGy).13 MMF (1 g every 12 h) and CYA were given beginning on day –3. A total MPA trough concentration <1 μg/ml in week 1 post transplant was associated with a lower rate of neutrophil engraftment. A low unbound MPA exposure (AUC0–12 <0.300 μg*h/ml or cumulative 24-h AUC <0.600 μg*h/ml) in week 1 post transplant was associated with more frequent acute grades II–IV GVHD(58% vs 35%, P=0.05). In another study, 85 adult HCT recipients of unrelated BM (n=6) or PBSC (n=79) received fludarabine (90 mg/m2) and TBI (2 Gy) in combination with MMF and CYA.12 A total MPA Css <3 μg/ml was associated with donor T-cell chimerism <50% (P=0.03). MPA exposure measures were not associated with acute GVHD. The differences observed in these two studies may be due to dissimilarity in stem cell source and/or the preparative regimen. Although both regimens are considered nonmyeloablative, the first regimen contains modest dose of CY and higher doses of fludarabine, which may be more immunosuppressive and/or convey more toxicity. Preparative regimen toxicity and graft source may influence the necessary intensity of post transplant immunosuppression required for good clinical outcome. MPA therapeutic ranges and duration of therapy may therefore differ depending on these factors. Mycophenolate exposure– response relationships have not been well described when MMF is combined with tacrolimus, although the combination has been safely administered in HCT.30,31
Proposed MPA target concentrations (total trough ≥1 μg/ml, total Css ≥3 μg/ml and cumulative 24-h unbound AUC>0.600 μg*h/ml) are achieved in 50% or less of nonmyeloablative HCT recipients receiving MMF 2 g/day.12,13 Owing to the concerns of low exposure with 2 g/day, MMF 3 g/day (1 g three times daily orally) was administered and pharmacokinetics were studied in 41 recipients of nonmyeloablative HCT.12 All exposure measures were higher than previously observed at 2 g/day. On day 7 post transplant, the total MPA Css and trough concentration were a mean (range) of 3.1 (1.1–6.0) and 2.5 (0.4–19.3) μg/ml, respectively. The cumulative 24-h unbound AUC was 0.855 (0.246–2.5) (n=32) μg*h/ml. In a nonrandomized analysis, 71 patients receiving an unrelated peripheral blood mononuclear graft were treated with a nonmyeloablative preparative regimen containing fludarabine (90 mg/m2) and TBI (2 Gy) followed by MMF 15 mg/kg (~1 g) twice daily and CYA.32 Graft rejection and acute GVHD grades II–IV rates were 15 and 52%, respectively. Given the high rate of rejection and GVHD, postgrafting immune suppression was enhanced by increasing MMF dosing to three times daily (3 g/day). In the second cohort of patients (n=103) treated with the higher MMF dose, graft rejection was reduced to 5% (P=0.004), although the acute GVHD rate (53%) was unchanged (P=0.37). Documented viral and fungal infections were slightly higher in the MMF three times daily group (P≤0.04). Overall, the data suggested that MMF 3 g/day provides better MPA exposure than 2 g/day and is associated with better clinical outcomes after nonmyeloablative HCT. After myeloablative HCT, MMF doses>3 g/day did not show a clinical benefit.25 In a phase 1 trial (n=30), the three cohorts received increasingly higher doses of MMF (15mg/kg every 12, 8 and 6 h) in combination with CYA. Increasing doses did not appear to decrease the risk of acute GVHD. At the 15 mg/kg every 8-h dose level (~3 g/day), MPA exposure was similar to that achieved in organ transplant with an acceptable rate of GVHD. Outcomes were not better at the highest dose level and there was a suggestion of more toxicity.
In organ transplantation, MPA exposures are higher when MMF is combined with tacrolimus compared with CYA.33–37 It is not clear if this also occurs in HCT subjects. Recently Haentzschel et al.31 reported a pilot study (n=29) targeting total MPA AUC0–12 (goal 35–60 μg*h/ml) in combination with tacrolimus. Patients were initially given MMF 1.5 g every 12 h i.v., AUCs were measured on days 3, 7 and 11, and doses adjusted to achieve the target, if necessary. On day 3, the median observed MPA AUC0–12 was 35.1 μg*h/ml, which is similar to what we observed (median 30.45 μg*h/ml) in our study on the same dose. MPA trough concentrations were measured in 14 children receiving MMF and tacrolimus.30 Total MPA trough concentrations were low (≤0.6 μg/ml) when receiving MMF 15 mg/kg every 12 h. Doses were subsequently increased to MMF 600–1200 mg/m2 every 6 h in some children to achieve a trough concentration >1 μg/ml. Hence, it remains unclear if tacrolimus enhances MPA exposure in HCT.
As total MPA targets were poorly achieved in this study, it is tempting to consider higher MMF doses. However, the unbound MPA cumulative 24-h AUC of 0.600 μg*h/ml was readily achieved on the 3 g/day dosing regimen. The upper limit of MPA toxicity has not been defined in HCT. In organ transplantation, high MPA exposures have been associated with greater rates of hematotoxicity and infection. 38,39 MMF toxicity has also been reported in patients with normal total MPA concentrations but dramatically elevated unbound concentrations.40,41 Given the potential inaccuracies and difficulties in interpreting total concentrations in the setting of altered protein binding, we believe that unbound concentrations are more likely a better reflection of overall immunosuppressive activity and that doses of 3 g/day are sufficient in most patients.
The applicability of our pharmacokinetic data to other HCT settings is not known. MPA pharmacodynamic studies have not been conducted with myeloablative-, non-fludarabine-based preparative regimens or in combination with tacrolimus. Given the greater toxicity of the myeloablative regimens, MPA pharmacokinetic disposition may be different, particularly following oral administration.
This work was supported by grants from the National Institutes of Health (NIH), National Cancer Institute 5K23CA096622 (PJ) and a seed grant from the University of Minnesota Academic Health Center (PJ). The expert technical assistance of Jason Dagit and Jim Fisher is gratefully acknowledged.
Drs Jacobson and Weisdorf have in the past received financial support from Roche Pharmaceuticals.