Comparative clinical trial of the variability factors of the exposure indices used for the drug monitoring of two tacrolimus formulations in kidney transplant recipients

Abstract Background Several studies found differences in tacrolimus whole blood trough levels (C0) or area-under-the curve (AUC) between the twice-daily (Tac-BID) and once-daily (Tac-OD) formulations given to kidney transplant recipients at equal doses. As C0 is widely used as a surrogate of the AUC for individual dose adjustment, this study investigated the correlation and proportionality between C0 and the 24h-AUC, depending on the formulation, time post-transplantation, pharmacogenetics traits and other individual characteristics. Methods 45 adult kidney transplant recipients were randomized to receive either Tac OD or Tac BID. On days 8 ± 1 (D8) and 90 ± 3 (month 3, M3), blood samples were collected over 24 h in both groups. Tacrolimus concentrations were determined using HPLC–MS/MS and common CYP3A5, CYP3A4 and ABCB1 genotypes characterized using allelic discrimination assays. Tacrolimus population pharmacokinetics was studied in the two patient groups using the Iterative Two Stage (ITS) technique, considering a one-compartment model with two gamma laws to describe the absorption phase. Bayesian estimation based on the C0, C1 h and C3 h concentrations was employed to estimate individual Tac AUC 0-12h and AUC 12-24h (for Tac BID), or AUC 0–24h (for Tac OD). Multiple linear regression was used to evaluate the influence of Tac formulation, post-transplantation period, recipient gender, existing glucose metabolism disorders, and CYP3A5, CYP3A4 and ABCB1 genotypes on C0, AUC 0–24h and the AUC-to-trough concentration ratios. Results The Full Analysis Set comprised 22 patients on Tac OD and 20 on Tac BID. Tac exposure indices as well as their time evolution were similar in the two groups. Multi-linear modeling analysis showed that the Tac dose was higher with Tac-OD than Tac-BID, on D8 than at M3 and in CYP3A5 expressors (p  0–24h was significantly higher on D8 than at M3. The AUC 0–24h /C0 ratio was not affected by the drug formulation and the polymorphisms studied, but it was significantly lower on D8 than at M3 (p = 7.8 × 10 −5 ). In contrast, both the post-transplantation period (p = 1.53 × 10 −4 ), and CYP3A5 expression (p = 0.003) had a significant influence on the AUC 0–24h /C24 h ratio, explaining 19% and 12% of its variability, respectively. Consistently, for both Tac formulations, the AUC 0–24h was better correlated with C24 h than C0, and for Tac-BID the AUC 0-12h was better correlated with C12 h than C0. Conclusions This study confirms that the precisely timed 12h- or 24h-post-dose blood concentration (as opposed to the vaguely defined ‘trough level’) is a convenient surrogate of the 24h-AUC of tacrolimus for the two TAC formulations over the first 3 months post-transplantation. Still, for a given C24 h value, AUC 0–24h was higher on D8 and in CYP3A5 expressors. Bayesian estimation of AUC 0-12h for TAC BID and AUC 0–24h for TAC OD is feasible using only 3 time points within the first 3 h, thus giving access to the actual overall exposure.