Aline Fuchs, Tariq Afroz, Romain Ollier, Florence Guilhot, Andrea Pfeifer, Marie Kosco-Vilbois, Kasia Piorkowska
AC Immune SA, Lausanne, Switzerland
Introduction: Monoclonal antibodies (mAbs) are an established class of therapeutics. A key determinant of a mAb’s clearance (CL) is the binding affinity for the human neonatal Fc receptor (FcRn). Although preclinical PK studies are often performed in non-human primates (NHP), recently transgenic mice expressing the human FcRn have gained interest. This case study demonstrates the use of the Tg32 mice to aid in evaluating mAbs for candidate selection as well as predict human PK for linear CL.
Objective: To assess the value of the human FcRn transgenic mice (Tg32 mice) during preclinical development for early candidate screening and to support early human PK prediction.
Methods: One chimeric and four humanized mAbs were evaluated. Serum mAb concentrations were determined by ELISA after a single intravenous (iv) bolus administration of 40 mg/kg to male Tg32 mice and NHP with sampling over 6 and 8 weeks, respectively. One humanized mAb was also given at 4 mg/kg iv single dose. The mAb, trastuzumab, a humanized IgG1, was used as a reference. Samples with suspected anti-drug antibodies (ADA) were excluded from analysis. PK parameters were estimated by non-linear mixed effect (NLME) modelling and a 2-compartmental model applied. Linear elimination from the central compartment was systematically assessed. Saturable elimination and empirical target mediated drug disposition (TMDD) model combining linear and saturable elimination from central compartment were assessed when more than one dose level was tested. Typical PK parameters describing the PK in Tg32 and NHP for each antibody was scaled by allometric scaling using fixed exponent from Tg32 mice and NHP [1] to predict 70 kg human PK parameters and simulate human PK profile. Data preparation, exploration and model pre- and post-processing was performed with Rstudio (3.6.3) and the R package IQRtools (1.8.0). NLME models were run using Monolix (2019R1).
Results: A 2-compartmental model with linear elimination was chosen as it served to best characterize the PK profile for most mAbs except for the chimera in NHP where saturable elimination was selected. Due to limited data points from the Tg32 and low subject numbers in NHP, interindividual variability was fixed. PK parameters were estimated with relative standard error (RSE) < 35 %. For the reference mAb, trastuzumab, the human prediction from the data obtained from Tg32 was in accordance with published clinical data while the prediction from NHP showed a lower exposure as compared to the reported human data. However, it is reported that trastuzumab undergoes TMDD in NHP [2]. For the test mAbs, the human predictions were in accordance between the two species for humanized mAbs: low CL estimates and low volume of distribution (Vd) predicted from both Tg32 mice (CL < 0.112 mL/hr/kg, Vd < 67 mL/kg) and NHP (CL < 0.107 mL/hr/kg, Vd < 51 mL/kg) for all except one of the humanized mAb. In both species, the same candidate had the highest human predicted CL of 0.195 mL/hr/kg and 0.256 mL/hr/kg with the highest human predicted Vd of 89.4 mL/kg and 69.4 mL/kg from Tg32 and from NHP respectively. Simulation of human predicted concentration-time profile showed the lowest and shortest exposure for this candidate and the chimera whereas a prolonged exposure for the three others humanized mAb candidates was observed.
Conclusions: The results between Tg32 mice and NHP were in accordance as well as the ranking of the various test mAbs when predicting the human PK profile. Taken together, this case study demonstrates the value to use Tg32 mice to allow early prediction of the PK in humans, especially for clearance evaluation, before NHP studies are initiated. In this manner, using Tg32 mice would allow to select a limited set of candidates for NHP PK studies, thereby reducing the use of NHP and enhance the early selection of potential clinical candidates.
References:
[1] Betts et al. MABS 2018, 10:751.
[2] European Medicines Agency. Herceptin: EPAR Scientific Discussion.
Reference: PAGE 30 (2022) Abstr 10079 [www.page-meeting.org/?abstract=10079]
Poster: Drug/Disease Modelling - Other Topics