Mycophenolate: Better Value through Monitoring?

This is the authors' abstract. We don't add key points for this paper.

In the 10 years since the introduction of mycophenolate mofetil (MMF), the emphasis of antirejection therapy has shifted from absolute potency to targeted efficacy—benefit with reduced toxicity. Concerns about side effects have similarly refocused from acute problems such as infection to chronic issues such as malignancy. Drug level monitoring is universally adopted to achieve these objectives for both ciclosporin and tacrolimus, but monitoring of mycophenolic acid (MPA) is still far from widespread. In this brief analysis, we review recent developments that strengthen the case for MPA monitoring and examine whether it can be a cost-effective process. A 10-fold variation in MPA pharmacokinetics has been widely reported in transplant populations (1–3). Now, Mercé and colleagues in this issue of Transplantation (1), describe a similar variability in the pharmacodynamic activity of low predose MPA drug levels. Coupled with the narrow therapeutic index and lack of an established real-time indicator of MPA efficacy, these alone are justifiable criteria for therapeutic drug monitoring. However, two additional sources of variability exist: clinical and pharmacological modifiers of drug handling (2) and genetic variants that influence MPA metabolism (4). To appreciate the impact of both, key facts about the metabolism and excretion of MPA need emphasis. As reviewed in detail by van Gelder and Shaw (2), MPA is extensively and rapidly absorbed from both MMF (CellCept) and the enteric-coated formulation of mycophenolate sodium (Myfortic). It is metabolized both in the intestine and liver by at least three UDP-glucuronosyltransferase enzymes, UGT1A8, UGT1A9, UGT1A10, and the primary phenolic glucuronide metabolite (MPAG) is excreted in both urine and bile. Biliary MPAG is hydrolysed by bacterial glucuronidases and extensively reabsorbed as MPA, contributing an estimated 10−>60% of systemic MPA exposure. Numerous clinical and pharmacological variables and UGT1A9 genetic variants influence these processes (Table 1) and apply to solid organ transplantation, bone marrow grafting, oncology, and autoimmunity alike.TABLE 1: Impact of clinical, pharmacological and genetic variables on mycophenolate exposureThe impact of these variables is considerable: MMF doses during ciclosporin co-medication are around twice those needed with tacrolimus; MPA exposure is 30–50% lower with long-term glucocorticoids than without; MPA levels during antibiotic therapy and hypoalbinaemia may fall by 50%, while those in severe renal dysfunction may vary 2-fold; finally, MPA levels in individuals with promoter variants of UGT1A9 fall by 50% (4). Given the established relationship of low MPA exposure to rejection episodes and of high levels to adverse events such as leucopenia, gastrointestinal (GI) disturbances (1–3) and possibly malignancy, we believe that the realistic issue is not whether but how to monitor MPA drug exposure. Which monitoring approach is best for MPA? The two emerging preferences are levels in predose samples and measurements of AUC—the area under the curve of drug concentration versus time. These are monitored relative to consensus targets of 1–3.5 mg/L and 30–60 mg.h/L, respectively. Single samples are most convenient and economic but can be misleading if delayed or variable absorption of drug occurs. This applies to enterohepatically recirculated MPA, where differing comedication profoundly affects the magnitude of the recirculated MPA peaking around 8 hours postdose, and maybe to the enteric formulation, where reports exist of multiple and late absorption peaks. Practically, AUC measurements are expensive and onerous for the patient, clinical staff and the laboratory, but may be useful in managing difficult cases. Limited sampling strategies for estimating AUC are a less demanding compromise but are not widely validated, such as with different co-medication and in different patient populations. The choice of analytical technique will be governed by local expertise, equipment and budget. The gold standard method is liquid chromatography (LC) coupled preferably to mass spectrometry (MS) for detection. This is a highly sensitive method that also allows measurement of free drug as an additional index of efficacy/toxicity (5). Measurements of MPAG and the acyl glucuronide of MPA, implicated in GI toxicity (2, 5), may also assist interpretation. Immunoassay provides invaluable guidance and considerable benefits over categorical dosing, but may overestimate levels because of cross-reactivity with MPA acyl glucuronide. The chosen analytical method will clearly determine the cost:benefit relationship, as will testing frequency. The latter is best related to need rather than dogma—and the need arises when dose, clinical condition, interactive co-medication, graft function, renal function or time posttransplant change significantly (Table 1). Where LC or LC-MS technology already exists, introduction of MPA monitoring will mainly increase demands for operator time. Otherwise, immunoassay of trough levels is likely to be preferred but has significant additional consumable costs. In the UK, each analysis by immunoassay will cost the equivalent of two days’ medication with the recommended initial 1.5 g MMF dose given bd. This is based on undiscounted prices of CellCept and the Dade-Behring EMIT 2000 Mycophenolic acid immunoassay kit used with 20% wastage (accommodating calibration and controls). Although higher doses of MMF may be needed early after transplantation (1), a median daily dose of only 500 mg bd (range 0.25 to 4 g) was administered to the 249 stable adult liver recipients tested at our centre on 1125 occasions during in- and out-patient visits over the last 12 months. Only 0.8% required more than the 1.5 g bd recommended currently in liver transplantation and 74% received less than 1g MMF bd—exemplifying the dose reductions that can subsidize the monitoring programme as suggested earlier (3). Further savings should arise from the reduced treatment of, and hospitalization from, fewer episodes of acute rejection and adverse events. Although corroboration of benefit awaits the results of ongoing concentration controlled trials (2), we conclude that MPA monitoring can provide cost-effective improvements in management of patients prescribed CellCept and Myfortic. Even with drug level measurements, the true status of immunosuppression in certain patients will remain interpretatively challenging. Particularly problematic are those in whom low serum albumin levels, high MMF doses, but low total MPA exposure exist, and that increasing population where lower doses (and low levels) of multiple immunosuppressive agents with distinct mechanisms of action are combined. These cases emphasize the need for innovative immunosuppressive monitoring methods typified by the complementary pharmacodynamic approaches assessed by Mercé and colleagues (1).