Challenges in the Pharmacokinetics of Therapeutic Proteins

Due to the complex structure and complicated disposition pattern of therapeutic macromolecules, their pharmacokinetic interpretation has many challenges. Two of these challenges were investigated in this dissertation: 1) the error of classical bioavailability assessment observed during subcutaneous (SC) administration of therapeutic macromolecules that undergo target-mediated drug disposition (TMDD) and 2) the ontogeny of the neonatal Fc receptor (FcRn) expression along with its effect on the pharmacokinetics of monoclonal antibodies (mAbs) during development. TMDD often well describes the pharmacokinetics of therapeutic proteins that have high specificity and affinity of binding to their target receptors. The target receptors can be saturated by therapeutic proteins under therapeutic concentration due to their limited expression and availability in the body. Consequently, clearance through this pathway will reach its maximum and nonlinear pharmacokinetics will be observed upon further increasing dose if TMDD is a major elimination process. This, in turn, will impact the bioavailability estimation. Bioavailability estimations based on the classic AUC approach can be erroneous in this situation, mainly due to the incorrect assumption of dose-independent constant clearance that cannot be applied to therapeutic proteins that undergo TMDD. To shed light on this issue, a simulation study was performed with two model drugs: filgrastim and denosumab. Their published structural pharmacokinetic models and model parameters were employed in the simulations of plasma concentration-time profiles at different IV and SC doses. The bioavailability was calculated as the ratio of dose-normalized AUC after SC administration to that after IV administration. The overestimation was extreme when high SC and low IV doses of both protein drugs were used for the estimations, whereas excessive underestimation was observed with the combination of low SC and high IV doses. These biases in the bioavailability estimation resulted from the transition from low plasma concentration (at low doses) to high plasma concentration (at high doses), which shifted the major elimination pathway from TMDD to the unspecific linear clearance pathway. The changes in clearance resulted in parallel changes in dosenormalized AUCs and were very dynamic in the dose range of 0.1 – 5 µg/kg for filgrastim and below 60 mg for denosumab; thus caution is necessary when bioavailability of these two therapeutic proteins is estimated in these dose ranges using conventional method. To minimize the error of conventional bioavailability estimation of protein drugs that undergo TMDD, the bioavailability should be estimated at similar IV and SC doses or the assessment should be performed in dose ranges that yield constant dose-normalized AUCs (0.01 – 0.1 µg/kg or 5 – 10000 µg/kg for filgrastim, and 60 – 210 mg for denosumab). Moreover, an alternative estimation method could be applied, which determines the ratio of IV and SC doses that generate equally shaped concentration-time profile by applying a variable rate IV infusion, thereby resulting in equal AUCs as suggested by others. FcRn has been evidenced as a salvage pathway from lysosomal clearance for mAbs and Fc conjugated proteins; thus it can prolong the existence of these protein drugs in systemic circulation. The ontogeny of FcRn expression and its effect on the pharmacokinetics of mAbs should be of special concern if therapeutic mAbs are used in both pediatrics and adults. The down-regulation of FcRn during the development may shortened the half-life of therapeutic mAbs observed in adults. To address this problem, FcRn expression was quantified in various organs of C57BL/6J mice from postnatal days 2 through 70, the pharmacokinetics of AMG589 were studied in different age groups of C57BL/6J mice, and the correlation between the FcRn expression levels and the pharmacokinetics of AMG589 at various developmental stages of mice were explored using a nonlinear-mixed effects modeling-based population pharmacokinetic approach. FcRn showed ontogenetic changes in liver, lungs, and kidneys. Two-fold increases in FcRn expression were observed in liver and lungs of 10-day-old mice, whereas FcRn expression in the kidneys was doubled in 10- and 42-day-old mice. However, the ontogeny of FcRn expression could not be correlated to the prolonged persistence of AMG589 observed in 42 day old mice. A population pharmacokinetic approach revealed that after accounting for the effect of body weight by allometric scaling, age and FcRn expression in skin influenced the pharmacokinetics of AMG589 in different age groups of mice. Decreasing volume of distribution of AMG589 was observed during development. Interestingly, clearance of AMG589 was negatively correlated with the expression of FcRn in the skin, even though FcRn expression in skin did not show any ontogeny. These results suggest that body weight, age, and FcRn expression in skin could affect the pharmacokinetics of fully-human mAbs. However, regardless of the species difference in physiology, body weight should be considered during dosage regimen design, especially for pediatric patients who show a highly dynamic change in body size at early age. In summary, the findings in this dissertation have pointed out the weakness of the classical bioavailability estimation for protein drugs that undergo TMDD and have determined the factors that should be considered for dose adjustments of therapeutic mAbs in different-aged populations