Predicting Method for the Human Plasma Concentration-Time Profile of a Monoclonal Antibody from the Half-life of Non-human Primates

Development of a New Method to Evaluate the Biodistribution of Antibodies Using Non-Radioactive Metal Labeling and Inductively Coupled Plasma Mass Spectrometry

PURPOSE

The use of radiolabeled compounds is associated with a number of limitations. Therefore, a new method for the radioisotope-free evaluation of antibody distribution using metal labeling and inductively coupled plasma-mass spectrometry (ICP-MS) was developed herein.

METHODS

Indium-labeled monoclonal antibodies were administrated intravenously to tumor-bearing mice and cynomolgus monkeys, and antibody concentrations in plasma and tissues were measured by ICP-MS. The results were compared with those obtained using a ligand binding assay (LBA) and radioisotope-labeled antibody administration. Indium-, terbium-, holmium-, and yttrium-labeled cetuximab were co-administered to one C57BL/6 J mouse for simultaneous PK and tissue distribution evaluations.

RESULTS

The administration of a radioactive or non-radioactive indium-labeled anti-human interleukin-6 receptor (hIL-6R) antibody to tumor-bearing hIL-6R transgenic mice resulted in similar plasma antibody concentration-time profiles by ICP-MS, a ligand binding assay (LBA), and gamma-ray detector. Liver, kidney, brain, spleen, and tumor concentrations of antibodies measured by ICP-MS were similar to those after the administration of radiolabeled anti-hIL-6R antibodies. Following the administration of indium-labeled cetuximab to cynomolgus monkeys, plasma antibody concentrations measured by ICP-MS were similar to those measured by LBA, and antibody concentrations in organs were evaluable by ICP-MS. The PK of all metals were similar to antibody PK evaluated by LBA, and concentrations in each tissue were equivalent among metals.

CONCLUSIONS

The assessment of antibody distribution using ICP-MS is a novel alternative to the traditional radiolabeled approach. It facilitates the assessment of antibody distribution in the early stages of drug discovery and accelerates the assessment of target engagement.

Review of the Existing Translational Pharmacokinetics Modeling Approaches Specific to Monoclonal Antibodies (mAbs) to Support the First-In-Human (FIH) Dose Selection

The interest in therapeutic monoclonal antibodies (mAbs) has continuously growing in several diseases. However, their pharmacokinetics (PK) is complex due to their target-mediated drug disposition (TMDD) profiles which can induce a non-linear PK. This point is particularly challenging during the pre-clinical and translational development of a new mAb. This article reviews and describes the existing PK modeling approaches used to translate the mAbs PK from animal to human for intravenous (IV) and subcutaneous (SC) administration routes. Several approaches are presented, from the most empirical models to full physiologically based pharmacokinetic (PBPK) models, with a focus on the population PK methods (compartmental and minimal PBPK models). They include the translational approaches for the linear part of the PK and the TMDD mechanism of mAbs. The objective of this article is to provide an up-to-date overview and future perspectives of the translational PK approaches for mAbs during a model-informed drug development (MIDD), since the field of PK modeling has gained recently significant interest for guiding mAbs drug development.

Developing a computational pharmacokinetic model of systemic snakebite envenomation and antivenom treatment

Snakebite envenomation is responsible for over 100,000 deaths and 400,000 cases of disability annually, most of which are preventable through access to safe and effective antivenoms. Snake venom toxins span a wide molecular weight range, influencing their absorption, distribution, and elimination within the body. In recent years, a range of scaffolds have been applied to antivenom development. These scaffolds similarly span a wide molecular weight range and subsequently display diverse pharmacokinetic behaviours. Computational simulations represent a powerful tool to explore the interplay between these varied antivenom scaffolds and venoms, to assess whether a pharmacokinetically optimal antivenom exists. The purpose of this study was to establish a computational model of systemic snakebite envenomation and treatment, for the quantitative assessment and comparison of conventional and next-generation antivenoms. A two-compartment mathematical model of envenomation and treatment was defined and the system was parameterised using existing data from rabbits. Elimination and biodistribution parameters were regressed against molecular weight to predict the dynamics of IgG, F(ab')₂, Fab, scFv, and nanobody antivenoms, spanning a size range of 15-150 kDa. As a case study, intramuscular envenomation by Naja sumatrana (equatorial spitting cobra) and its treatment using Fab, F(ab')₂, and IgG antivenoms was simulated. Variable venom dose tests were applied to visualise effective antivenom dose levels. Comparisons to existing antivenoms and experimental rescue studies highlight the large dose reductions that could result from recombinant antivenom use. This study represents the first comparative in silico model of snakebite envenomation and treatment.

Combination of cassette-dosing and microsampling for reduced animal usage for antibody pharmacokinetics in cynomolgus monkeys, wild-type mice, and human FcRn transgenic mice

PURPOSE

The aim of this study was to develop a useful antibody PK evaluation tool using a combination of cassette-dosing and microsampling in mice and monkeys in order to reduce the number of animals used.

METHODS

Cetuximab, denosumab, infliximab, and a mixture of the three antibodies, i.e., cassette-dosing, were administered intravenously to cynomolgus monkeys, C57BL/6J mice, and homozygous human neonatal Fc-receptor transgenic (Tg32) mice. Mouse blood was collected from one animal continuously via the jugular vein at nine points.

RESULTS

In cynomolgus monkeys, infliximab showed faster elimination in the cassette-dosing group than in the single-dose group. Anti-drug antibody production was observed, but the PK parameters of the clearance and distribution volume were similar in both groups. In C57BL/6J and Tg32 mice, each of the plasma concentrations-time profiles after cassette-dosing were similar to those after single dosing. PK evaluation using a combination of cassette-dosing and microsampling in mice may reduce the number of mice used by approximately 90% compared with the conventional method.

CONCLUSIONS

The combination of antibody cassette-dosing and microsampling is a promising PK evaluation method as a high-throughput and reliable with reduced numbers of mice and cynomolgus monkeys.

Prediction of Human Pharmacokinetics Profile of Monoclonal Antibody Using hFcRn Transgenic Mouse Model

Human pharmacokinetics (PK) profiles of monoclonal antibodies (mAbs) are usually predicted using non-human primates (NHP), but this comes with drawbacks in terms of cost and throughput. Therefore, we established a human PK profile prediction method using human neonatal Fc receptor (hFcRn) transgenic mice (TgM). We administered launched 13 mAbs to hFcRn TgM and measured the concentration in plasma using electro-chemiluminescence immunoassay. This was then used to calculate PK parameters and predict human PK profiles. The mAbs showed a bi-phased elimination pattern, and clearance (CL) (mL/d/kg) and distribution volume at steady state (V_dss) (mL/kg) ranges were 11.0 to 131 and 110 to 285, respectively. There was a correlation in half-life at elimination phase (t_1/2β) between hFcRn TgM and humans for 10 mAbs showing CL of more than 80% in the elimination phase (R² = 0.714). Human t_1/2β was predicted using hFcRn TgM t_1/2β; 9 out of 10 mAbs were within 2-fold the actual values, and all mAbs were within 3-fold. Regarding the predicted CL values, 7 out of 10 mAbs were within 2-fold the human values and all mAbs were within 3-fold. Furthermore, even on day 7 the predicted CL values of 8 out of 10 mAbs were within 2-fold the observed value, with all mAbs within 3-fold. These results suggest human PK profiles can be predicted using hFcRn TgM data. These methods can accelerate the development of antibody drugs while also reducing cost and improving throughput.

Improving priors for human monoclonal antibody linear pharmacokinetic parameters by using half-lives from non-human primates

Obtaining a good prior for the linear pharmacokinetics of new monoclonal antibodies (mAbs) would be an advantage not only for designing first-in-human (FIH) studies but also for stabilizing fitting of data with non-linear target-mediated disposition models. We estimated the pharmacokinetics from FIH studies for five mAbs using a two-compartment model, both separately and together, using a simple pool, a third hierarchical level of random effects for between mAb differences and non-human-primate half-lives as a predictor covariate for said differences. There was good agreement between compounds for the rapidly accessible central volume of 2.9 L (70 kg human), but clearances and peripheral volumes differed with terminal half-lives ranging from 15 to 28 days. The simple pool of human studies gave inter-individual variability estimates of 32% coefficient of variation (CV) for clearance and 33% CV for peripheral volume, larger than for separate fits (13-26% CV and 15-35% CV for clearance and volume respectively). Using third level hierarchical random effects gave inter-individual variability estimates close to those of separate fits (24% and 16% CV respectively). The between-mAb differences became predictable if non-human primate body weight scaled terminal half-life estimates were included as covariates on clearance and peripheral volume. In conclusion, ignoring inter-mAb variation leads to inflated estimates of inter-individual variability and unrealistic simulations for FIH studies. However, by using 70 kg body weight scaled terminal half-life estimates from non-human primates one can account for between-mAb differences and provide non-inflated priors for the linear pharmacokinetic parameters of new mAbs.