Mycophenolic Acid Pharmacodynamics and Pharmacokinetics Provide a Basis for Rational Monitoring Strategies

Mycophenolic acid (MPA), the active immunosuppressant form of the pro-drug mycophenolate mofetil (MMF), is a widely used component of immunosuppressive regimens in organ transplantation. This immunosuppressant is commonly administered in combination with a calcineurin inhibitor (CIN) + corticosteroid with or without induction therapy in the form of a polyclonal anti-T lymphocyte preparation or one of the available humanized monoclonal interleukin 2 inhibitors. There is strong evidence that maintenance CIN-MMF-steroid-based triple therapy, initiated in the early post-transplant period significantly reduces the risk of acute rejection in the first post-transplant year, when compared to double therapy regimens comprising CIN and steroids alone. After the initial phase of graft stabilization, there are several therapeutic approaches currently employed in transplant practice, including: (1) maintenance of a three-drug regimen, at reduced doses compared to the early post-transplant period; (2) elimination of corticosteroid; (3) reduction or elimination of CIN; (4) replacement of CIN with sirolimus in patients who are especially sensitive to the nephrotoxic effects of these agents; or (5) reduction or discontinuation of MMF, while retaining full-dose CIN or sirolimus and maintenance corticosteroids are among the other therapeutic options under investigation. Dose individualization of CIN is facilitated by target drug concentration monitoring, whereas for MMF, empiric dosing is most commonly practiced. A significant predictive value for assessment of the risks for acute rejection and hematologic side-effects has been shown for the 12-h dose interval mycophenolic acid (MPA) area under the concentration–time curve (AUC) and, although the data are less precise and with lower statistical significance, for the predose trough MPA concentration. Investigations have shown that there is a natural and significant pharmacokinetic variability of MPA as well as a sizable difference in the steady-state concentrations of MPA achieved, adjusted for mycophenolate mofetil (MMF) dose, depending on what concomitant immunosuppressant is used. These observations have together led to a heightened interest in MPA monitoring. Here, we consider our expanding knowledge of: (a) the effects of MPA (pharmacodynamics, PD) on proliferating T-lymphocyte functions, and the measurement of inosine monophosphate dehydrogenase (IMPDH) catalytic activity as potential indices of MPA immunosuppressive effects; (b) sources of MPA PK variability, particularly the effects of concomitant immunosuppressants on the steady-state MPA AUC; (c) the relationships between MPA PK and PD; (d) the relationship between MPA PK and clinical outcome. Based on these observations, we conclude with a proposed approach for individualization of MMF dosing as a replacement of a ‘one size fits all’ dosing scheme. Investigations of the effects of MPA on lymphocytes have focused primarily on: (i) assessment of the effects of MPA on proliferating lymphocyte functions in cell culture and whole blood of control and transplanted rats, and to a limited extent in human subjects; (ii) inosine monophosphate dehydrogenase (IMPDH) catalytic activity measured in blood lymphocytes of healthy volunteers and transplant patients over the 12-h dose interval; and (iii) measurement, in vitro, of the inhibition of proliferating cells by an aliquot of plasma from a transplant patient receiving MMF as an index of inhibition of IMPDH catalytic activity. The immunosuppressive effects of MPA have been evaluated by the in vitro assessment of lymphocyte functions in samples of whole blood obtained from rat controls or recipients of a heart allograft. The qualitative characterization of the effects of a range of concentrations of MPA on stimulated lymphocytes in aliquots of whole blood has been reported 1-3). In these investigations not only was new DNA synthesis found to be suppressed by MPA, as previously described, but suppression of the expression of several cell surface T-cell activation markers, CD-25 (interleukin-2 receptor α-chain), CD-134 (Ox40 receptor, member of the nerve growth factor/tumor necrosis factor superfamily), CD-71 (transferrin receptor), CD-11a (leukocyte function antigen-1 receptor α-chain, LFA-1), and CD-54 (intercellular adhesion molecule-1, ICAM-1), was detected as well (3). Evidence has also been provided for the suppression of cytokine production by limiting the number of cytokine-producing cells (4) and for increasing, or at least not inhibiting, apoptosis in lymphocytic and monocytic cell lines (5) or activated lymphoblasts (6). Three key observations in these investigations are: the statistically significant relationships between MPA concentrations and (a) the suppression of new DNA synthesis, (b) the expression of T-cell activation markers, and (c) in a rat heart allograft model, the statistically significant relationship between MPA AUC and the whole blood PD assays and the graft rejection scores (2, 3). The correlation between the MPA AUC and the graft histology score was r2 = 0.83 and for the MPA concentration at 6 h it was r2 = 0.85, whereas for the MPA trough, r2 was 0.57 for this outcome (3). Measurement of the catalytic activity of the enzyme target of the drug, IMPDH, in whole-blood lymphocytes in transplant patients receiving MMF, is another approach that has been used for assessment of the effects of MPA. MPA immunosuppression results from the reversible uncompetitive inhibition of IMPDH, a key enzyme required for sustaining the guanine nucleotide pool and de novo DNA synthesis in proliferating lymphocytes (7, 8). The assay of IMPDH in different blood compartments has produced conflicting results, probably due to the difficulties associated with reliable and robust analysis of this enzyme in patient samples 9-11). The most promising approach for this assay is measurement of the rate of xanthine monophosphate production by isolated mononuclear cells under controlled in vitro conditions using validated HPLC methodology 12-14). The results of all of the studies agree that there is considerable intersubject variability in baseline IMPDH activity, but the time course of IMPDH inhibition within a dose interval parallels that of MPA plasma concentration 9-14). Maximal IMPDH inhibition coincides with MPA Cmax and exceeds 50% compared to baseline activity in most of the studies. In a study in six stable renal transplant patients using the more reliable HPLC methodology for assessment of IMPDH activity, maximal MPA concentration of 25.9 ± 6.5 mg/L (measured by the EMIT method) was achieved at 1 h after oral MMF, with the MPA levels returning to baseline values by 4 h (14). The baseline IMPDH activity of 9.4 ± 3.3 nmol/h/mg mononuclear cell protein decreased rapidly in all six patients following administration of a 1-g dose of MMF. Maximal inhibition was achieved at 1 h, parallel to the time to maximal MPA concentration, with a mean inhibition of 87 ± 8% (range: 63–100%). IMPDH inhibition persisted in four of the patients for 4 h despite low MPA concentration values and all returned to baseline IMPDH activities by 11 h (14). Further studies in larger numbers of different transplant patient populations will be required to determine whether or not there are interpatient differences in enzyme kinetic parameters such as the Km, Ki or Vmax that could give rise to differences in response to a given plasma concentration of MPA, whether or not there are clinically significant genetic differences in IMPDH, and to determine the predictive performance of IMPDH inhibition for risks of acute rejection and side-effects. An interesting approach to the determination of MPA immunosuppressive activity is the measurement of the degree of inhibition of CEM cell proliferation using aliquots of serum from MMF-treated patients (11). Proliferation of this cell line is resistant to the effects of the CIN and corticosteroid, but is inhibited in a concentration-dependent manner by MPA, thereby reflecting IMPDH inhibition. Further investigations will be required to test the practicality of this approach for the evaluation of the PD effect of MPA in transplant patients receiving various combinations of concomitant immunosuppressants. New insights have been gained regarding the effects of MPA on the expression of cell surface markers of proliferation in activated T cells. Nevertheless, the primary effect of MPA is the inhibition of new DNA synthesis and arrest of cell cycling at the G1/S interface in proliferating T lymphocytes (8). The onset of the inhibition of IMPDH in proliferating lymphocytes by increasing concentrations of MPA appears to be rapid and reversible. The time course of inhibition of the proliferation and suppression of cell surface activation antigens needs more detailed evaluation in order to assess the delay time between MPA concentration changes and subsequent effects. Based on the investigations of the relationship of MPA concentration to PD effects there is a correlation between MPA concentration and effects. These observations support the use of MPA plasma concentration, especially the MPA AUC, as a surrogate marker for MPA effects. Investigations of MPA pharmacokinetics have been conducted in adult and pediatric renal and heart and adult liver and bone marrow transplant recipients 15-25). All of these studies show wide interpatient variability in drug exposure measured by the 12-h dose-interval MPA AUC or trough levels. Three examples of the variability of MPA exposure in transplant patients given a fixed dose of MMF are displayed in Figure 1. It is noteworthy that in the renal transplant patients the distribution was assessed during the first 2 weeks following transplant surgery, whereas for the heart transplant patients the time after surgery ranged from 1 to 20 months (average 7.5 months, median 6 months). The interpatient PK variability in liver transplant patients over the first 6 post-transplant weeks is even greater than that in renal or heart transplant patients (Figure 1). Mycophenolic acid (MPA) AUC values for cohorts of renal, heart and liver transplant patients. These data are adapted from (15, 19, 20), respectively. All MPA AUC values have been dose-normalized to 1 g mycophenolate mofetil (MMF) bid. The reported studies of MPA PK in recipients of hematopoietic cell transplants all indicate that in the majority of cases MPA exposure is low in comparison to solid organ recipients receiving equivalent doses of MMF 23-25). In most of these patients MPA concentrations from 4 to 12 h are at or near the limit of quantification of the method for measurement of the drug. These data suggest that enterohepatic cycling (EHC) is markedly reduced or absent in these patients. Furthermore, these data suggest that more effective MPA exposure can be provided by more frequent dosing (e.g. 3 or 4 × daily) and higher doses. Further studies are needed to determine the pathophysiological mechanisms responsible for the altered MPA PK in these patients and the most effective dosing strategy. Characterization of MPA metabolism has shown that the phenolic acid glucuronide, MPAG, is the primary metabolite and that it is pharmacologically inactive (26). The rate-limiting step in the clearance of MPA is its conversion to MPAG via the catalytic action of one or more isoforms of the UGT1 gene family of UDP-glucuronosyltransferase in the gastrointestinal tract, liver and possibly other tissues such as kidney. EHC of MPA is considered to be a significant contributor to the dose interval kinetics of MPA especially the post-distribution phase of the concentration–time curve (Figure 2). The contribution of EHC to the MPA AUC has been estimated to be approximately 37%, ranging from 10 to 61%, based on the effect of concomitant administration of cholestyramine in healthy volunteers on the MPA concentration vs. time profile (15). The finding of no significant difference in MPA Cmax, with vs. without cholestyramine, but a marked and highly significant reduction in MPA concentrations from 6 h onwards with cholestyramine is the experimental basis for the quantitative estimation of the contribution of EHC to the dose interval MPA profile. The occurrence of a secondary MPA concentration peak anywhere from 4 to 12 h following the morning dose of MMF is thought to be the result of EHC. Metabolic pathway for mycophenolate mofetil (MMF). The major metabolic steps are represented by the bold arrows. MPAG, mycophenolic acid 7-O-phenolic glucuronide; MPA, mycophenolic acid. Evidence for the formation of two minor metabolites, the acyl glucuronide and the phenolic glucoside, has been provided (27). The acyl glucuronide, but not the phenolic glucoside, metabolite of MPA inhibits IMPDH in vitro. Further investigations will be necessary to evaluate its possible immunosuppressive or toxic activities (27) including the following: (a) the possibility that MPA production from the labile acyl glucuronide metabolite could account for the inhibition of IMPDH during in vitro analyses; (b) the possibility that the acyl glucuronide moiety would interfere with the hydrogen bonding between the terminal carboxylate group on MPA and the hydroxyl and amide nitrogen in the serine 276 moiety of IMPDH. Disruption of those hydrogen bonds in the ‘inhibitor pocket’ of IMPDH by other MPA analogs with a modified terminal carboxylate moiety can cause a substantial reduction of inhibition in comparison to unaltered MPA (28); (c) whether it is possible for glucuronide metabolites to enter lymphocytes intact; (d) the recent suggestion that the acyl glucuronide may be involved in the development of the gastrointestinal side-effects associated with MPA (29) and therefore warrants measurement of this metabolite in patient samples. The lack of availability of a source of the pure material for measurement of this metabolite by investigators has limited the further investigation of the acyl glucuronide pharmacology. The primary sites and effects of drug–drug interactions involving other medications and MPA are likely to be decreased absorption in the gastrointestinal tract, inhibition of enterohepatic cycling and inhibition of transport of the primary phenolic glucuronide metabolite. Meal consumption just prior to oral intake of MMF delays absorption, causing a reduction in Cmax by about 25% but no change in the dose interval MPA AUC (30). The oral administration of ferrous sulfate (2 × 525 mg tablets) at the same time as MMF (4 × 250 mg capsules) caused an average 89.7% decrease of the MPA AUC in eight healthy volunteers, presumably by forming a poorly absorbable complex of iron with MMF (31). If the latter is the actual mechanism involved, this interaction can be avoided by appropriate timing of administration of MMF and iron tablets, analogous to the situation with antacids that lower the MPA AUC, presumably by decreasing absorption, when taken concomitantly. There is no evidence to date for interactions at the CYP3A4 or UDP-glucuronoyltransferase sites. It seems likely that further studies will reveal interactions at the UDP-glucuronosyltransferase site(s), since this is the rate-limiting step in MPA clearance and since there are known inducers and inhibitors of this enzyme family (32). Furthermore, since the catalytic amount of conjugating enzymes such as UDP-glucuronosyltransferase increases significantly during the first 3 years of life, age may be a significant source of intrapatient variability of MPA clearance, and therefore steady-state plasma concentrations, in the pediatric population (33, 34). Several investigations have shown that CsA lowers MPA concentrations. A study of the effect of CsA on MPA concentration showed that MPA trough concentration values were significantly lower during concomitant treatment with CsA + MMF + prednisone compared to the values achieved during treatment with MMF + prednisone alone (35). The trough concentrations of MPA increased significantly following discontinuation of CsA 6 months after transplant surgery. MPA trough concentrations in patients no longer receiving CsA averaged twice the values of patients who were still on CsA therapy at 9 months following transplant surgery with no change in the dose of MMF (35). A similar result was obtained in a cross-sectional investigation involving 18 kidney transplant patients (36). In a cross-sectional study involving 42 pediatric renal transplant patients, the average dose-normalized MPA AUC value was 82% higher in patients who received concomitant tacrolimus than in patients who received concomitant CsA and 45% higher than in patients who received no CIN therapy (37). The authors of this study noted that the mean MPA concentrations in the interval between 4 and 12 h after the MMF dose were comparable in the tacrolimus and the no-CIN patients but higher than the values for the CsA patients. This result supports the idea that CsA inhibits the transport of MPAG into bile from hepatocytes (38). According to this hypothesis CsA decreases the enterohepatic cycling of MPA and thereby reduces MPA concentrations in the 4- to 12-h window of the AUC profile. A study of the effect of CsA on MPA PK compared to tacrolimus or placebo in Lewis rats provided strong support for this suggested mechanism (38). According to the design of this investigation, cohorts of Lewis rats were treated with fixed daily doses of MMF and CsA or tacrolimus or placebo. After 2 weeks of dosing, the average MPA and MPAG AUC0−24 h-values for the MMF + tacrolimus and the MMF + vehicle groups were comparable. On the other hand, the average MPA AUC0−24 h-value in the CsA group was 37% and 41% lower than the value for the tacrolimus and vehicle groups, respectively. The corresponding average results for MPAG AUC0−24 h were comparable in the tacrolimus and vehicle groups, but approximately 60% higher in the CsA group. Analysis of the MPA AUC profiles in this study showed that the average MPA AUC0−2 h-values were comparable among all 3 cohorts. However, the average MPA AUC2−24 h for the CsA group was about one-half the corresponding value found for the tacrolimus and vehicle control groups (38). According to this analysis of the data, therefore, the post-absorption, post-distribution phase of the MPA AUC profile during which enterohepatic cycling of MPA is most prominent was suppressed in the CsA-treated animals. These observations support the proposed mechanism for the significantly higher value for steady-state, dose-adjusted MPA AUC in patients receiving concomitant tacrolimus and MMF compared to CsA + MMF: interruption of the EHC secondary to inhibition of MPAG transport across biliary duct epithelium into bile by CsA. The observation that tacrolimus inhibited human UDP-glucuronosyltransferase, in vitro, with an inhibition constant of 27.3 μg/L as compared to CsA, which had less of an inhibitory effect with an inhibition constant of 2518 μg/L, has led to the conclusion that this might be another factor that contributes to the differences in dose-adjusted MPA AUC in transplant patients depending on the concomitant immunosuppression(39). Emerging study data suggest that MPA pharmacokinetic behavior in patients receiving concomitant sirolimus therapy is comparable to the situation observed for tacrolimus. A randomized study in kidney transplant patients comparing the respective efficacies of a sirolimus-MMF vs. a CsA-MMF regimen showed that significantly higher dose-adjusted values for MPA Cmin and AUC, but not Cmax, were obtained in the patients who received combined sirolimus and MMF. Patients in the sirolimus-MMF subgroup experienced more toxicity and were more likely to need their MMF dose reduced or eliminated (40). The recommendation from this investigation, analogous to the situation for tacrolimus, is that MMF doses should be 0.5–0.75 g twice daily in patients treated with sirolimus + MMF in order to achieve the same MPA exposure as patients receiving CsA + twice-daily doses of 1 g MMF (40). An important MPA PK property is its tight and extensive binding to serum albumin (41). Inhibition of IMPDH and suppression of proliferation in mitogen-stimulated lymphocytes, in vitro, is dependent on free MPA concentration (41). In stable transplant patients the MPA free fraction ranges from 1 to 3%. The characterization of albumin binding and patient factors that can significantly alter MPA free fraction and free MPA AUC are described in detail elsewhere and will not be repeated here 15-18, 41, 42). In our experience the situations in which measurement of free, in addition to total, MPA concentration is useful for interpretation of MPA exposure include: (a) renal transplant patients with early poor function (16, 17); (b) patients with chronic renal failure (43, 44); (c) liver transplant patients in the early post-transplant period (22); (d) any patient with low serum albumin concentration and/or hyperbilirubinemia. In an investigation of MPA pharmacokinetics in pediatric renal transplant recipients, the free MPA AUC was a significant risk factor for hematologic side-effects, with a decision level of 0.4 mg*h/L (45) during the early post-transplant period. Above this value there was an increased risk for or with a of and of The AUC of MPA is at least lower in the first weeks than in the period months for the same dose of MMF in renal transplant patients. In most of the studies this data from the patients with renal function were not from those with early in one study this was the 20 of renal transplant patients with early function 4 on showed change approximately in the average MPA AUC with time after transplantation. In for of patients with renal the mean MPA AUC at 4 was lower than that in the patients with renal but it increased to a mean value comparable to that in patients with function by The average free fraction in the renal function group was higher than that in the early function group over the first 2 weeks but this was to equivalent by this in MPA free fraction is responsible for a in the clearance of the drug in the renal function group. This observation is at least in the of acute free fraction with a clearance mechanism for MPA In the of chronic renal mechanisms are likely to since MPA AUC values are comparable to those in stable patients, but free MPA AUC values can rise as as the patient at increased risk for hematologic toxicity (43, least 11 investigations of the relationship between MPA PK and clinical outcome have been eight in kidney and in heart transplant patients of the and are available for of the studies and these data are in 1. In studies CsA was the concomitant primary in one it was CsA and and in it was tacrolimus. corticosteroid therapy was in all of these studies. these investigations it is to conclude (i) MPA AUC has predictive value for the risk for acute rejection and in studies the risk for hematologic (ii) trough values of MPA also have predictive value for these outcome but are more and poorly with MPA AUC (iii) the area under the post-absorption, post-distribution phase of the MPA AUC between approximately 2 and 12 h following an MMF dose, is higher in to the MPA h when tacrolimus is the concomitant immunosuppressant compared to the situation involving (4) trough values may to have predictive value that can be used in clinical It is that the dose-interval MPA AUC is the most reliable index of risk for acute rejection and that there is poor correlation between MPA trough values and the MPA However, it is to a of plasma samples over the 12-h dose several investigators have evaluated limited for the estimation of MPA were obtained between of 12-h MPA AUC values using plasma samples at 4 or time and MPA AUC values our we use a estimation trough, and for the combination of MMF and tacrolimus and a estimation for the combination of MMF and CsA that of 3 samples within a period is the most estimation for the MPA AUC available and is a significant step in MPA therapeutic drug on a use the target range of and for patient we the distribution value for the respective transplant patient population MPA AUC are also the clinical of MPA trough values in comparison to the estimated MPA AUC value using the target range of or mg/L tacrolimus is concomitant which to the MPA AUC target range of It be that a key to therapeutic drug of MPA an effective for individualization of MMF dosing in transplant patients is between the and the clinical for the of the and clinical together with the of is as important as an and of MPA This suggested for effective immunosuppression a of a knowledge and of immunosuppressive drug PK and A suggested for MPA AUC in renal and heart transplant patients is an initial evaluation by the of the first post-transplant a determination 1 and at 1 at the following prior to a major change in immunosuppression such as CIN elimination or conversion to another CIN or to at the time of a major clinical such as hematologic or side-effects. that in order to a more of for the most effective use of MPA therapeutic drug monitoring, further investigations of the relationship between MPA concentration and clinical are including: (i) the use of immunosuppressive drug combinations not as widely (e.g. combination with tacrolimus and the combination with as has been with (ii) in patient populations not widely such as liver transplants and bone marrow transplant (iii) evaluation of the of MPA trough concentration vs. MPA AUC, estimated using validated and assessment of the of this A key step that will be required to effective drug for MPA of when it is at a or after such studies have been is the of the using validated methodology and with a of who to this