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

Abstract

Background: Mycophenolic acid (MPA) is widely used in solid organ transplantation. MPA absorption from enteric-coated mycophenolate sodium (EC-MPS) is delayed, which results in a delayed enterohepatic recirculation and subsequently higher and more variable MPA 12-hour trough concentration and tmax values. Therefore, MPA trough level monitoring cannot be used to monitor MPA exposure in patients who are given EC-MPS. The aim of the study was to develop and validate a limited sampling strategy (LSS) for accurate prediction of the 12-hour area under the concentration-time curve (AUC0-12h) for MPA in patients who receive concomitant EC-MPS and Tacrolimus (Prograf or Advagraf) within 196 months posttransplantation. According to our knowledge, the LSS for MPA AUC estimation using high-performance liquid chromatography to determine MPA concentrations in plasma samples of kidney and liver transplant patients receiving EC-MPS and Tacrolimus (Advagraf) has not been previously evaluated.

Methods: Seventy-four renal and liver transplant patients receiving EC-MPS and concomitant tacrolimus (either Prograf or Advagraf) provided a total of 74 pharmacokinetic profiles. MPA concentrations were measured using a validated high-performance liquid chromatography method for 9 plasma samples collected at predose and at 0.5, 1, 2, 3, 4, 6, 9, and 12 hours after the morning dose of EC-MPS after an overnight fast. LSS were developed and validated by stepwise multiple regression analysis with the use of a 2-group method (test, n = 37; and validation, n = 37).

Results: The 3 and 4 time point equations using C1h, C3h, C9h and C1h, C2h, C3h, C6h, respectively, were found to be superior to all other models tested. When these LSS models were tested in the validation group, the results were acceptable [for 3 time points equation: r = 0.824, percentage of prediction error: 6.32 ± 25.75, 95% confidence interval (CI): -40.71 to 79.76; percentage of absolute prediction error: 27.45 ± 29.89, 95% CI: 0.04-199.92, predictive performance, 71% of estimated AUCs comprised within 85%-115% of the measured full MPA AUC, natural logarithmic residuals (ln) mean ± SD: -0.03 ± 0.24; for 4 time points equation: r = 0.898, percentage of prediction error: 3.32 ± 18.26, 95% CI: -49.35 to 51.06; percentage of absolute prediction error: 14.05 ± 11.89, 95% CI 0.13-49.86, percentage of predictive performance, 83% of estimated AUCs comprised within 85%-115% of the measured full MPA AUC, natural logarithmic residuals (ln) mean ± SD: -0.01 ± 0.19].

Conclusions: LSS equations using concentrations at 1, 3, and 9 hours or 1, 2, 3, and 6 hours time points provided the most reliable and accurate estimations of the MPA AUC in stable renal and liver transplant recipients treated with EC-MPS and tacrolimus. Further studies on independent groups of patients are required to confirm clinical utility of the presented LSS models.