Pharmacokinetics of mycophenolic acid and determination of area under the curve by abbreviated sampling strategy in Chinese liver transplant recipients

Objectives: This study aimed to: (i) define the clinical pharmacokinetics of mycophenolic acid (MPA) in Chinese liver transplant recipients; and (ii) develop a regression model best fitted for the prediction of MPA area under the plasma concentration-time curve from 0 to 12 hours (AUC(12)) by abbreviated sampling strategy. Methods: Forty liver transplant patients received mycophenolate mofetil 1g as a single dose twice daily in combination with tacrolimus. MPA concentrations were determined by high-performance liquid chromatography before dose (C-0) and at 0.5 (C-0.5), 1 (C-1), 1.5 (C-1.5), 2 (C-2), 4 (C-4), 6 (C-6), 8 (C-8), 10 (C-10) and 12 (C-12) hours after administration on days 7 and 14. A total of 72 pharmacokinetic profiles were obtained. MPA AUC12 was calculated with 3P97 software. The trough concentrations (C-0) of tacrolimus and hepatic function were also measured simultaneously. Multiple linear regression analysis was used to establish the models for estimated MPA AUC12. The agreement between predicted MPA AUC12 and observed MPA AUC12 was investigated by Bland-Altman analysis. Results: The pattern of MPA concentrations during the 12-hour interval on day 7 was very similar to that on day 14. In the total of 72 profiles, the mean maximum plasma concentration (C-max) and time to reach Cmax (t(max)) were 9.79 +/- 5.26 mg/L and 1.43 +/- 0.78 hours, respectively. The mean MPA AUC(12) was 46.50 +/- 17.42 mg center dot h/L (range 17.99-98.73 mg center dot h/L). Correlation between MPA C-0 and MPA AUC12 was poor (r(2) = 0.300, p = 0.0001). The best model for prediction of MPA AUC12 was by using 1, 2,6 and 8 hour timepoint MPA concentrations (r(2) = 0.921, p = 0.0001). The regression equation for estimated MPA AUC(12) was 5.503 + 0.919 center dot C-1 + 1.8710 center dot C-2 + 3.176 center dot C-6 + 3.664 center dot C-8. This model had minimal mean prediction error (1.24 +/- 11.19%) and minimal mean absolute prediction error (8.24 +/- 7.61 %). Sixty-three of 72 (88%) estimated MPA AUC12 were within 15% of MPA AUC12. Bland-Altman analysis also revealed the best agreement of this model compared with the others and a mean error of +/- 9.89 mg center dot h/mL. Conclusion: This study showed the wide variability in MPA AUC(12) in Chinese liver transplant recipients. Single timepoint MPA concentration during the 12-hour dosing interval cannot reflect MPA AUC(12). MPA AUC(12) could be predicted accurately using 1, 2, 6 and 8 hour timepoint MPA concentrations by abbreviated sampling strategy.