Background: Biologics make up 30% of all drugs approved by the FDA, many of which display target-mediated drug disposition (TMDD). Optimal design methodology for nonlinear mixed effects models was developed to maximize the efficiency of design variables for accurate parameter estimation in drug development. Utilization of optimal design approaches can be advocated ethically and economically. We evaluated the impact of reducing and optimizing the First-in-Human (FIH) study design on precision of TMDD parameter estimation. Methods: A two-compartment TMDD model for an experimental monoclonal antibody with Michaelis-Menten approximation and linear elimination was built in mrgSolve R package. The R package PopED was used to perform evaluation and optimization of all designs. A full FIH study design with 4 doses, 6 subjects/dose and 12 PK sampling times over 84 days was compared to reduced and optimized sampling durations of 2, 21, and 35 days and only two dose groups. A Line Search optimization algorithm was used, clustering of sample times was allowed, and constraints for sampling times were identified based on the design being optimized. D-efficiency was estimated and the impact on the precision of the linear and non-linear parameter estimates was evaluated and confirmed via simulations. Results: Sixteen sampling designs in total were evaluated. Leveraging optimal sampling approaches was found to be needed to obtain precise model parameters for shorter duration designs while its value for longer designs was limited. Optimizing the study design to include only the two lower doses and an optimized PK sampling scheme of just 5 sampling times over 35 days (60% shorter study, Figure 1) did not compromise the precision of the parameter estimates. Higher doses can have a sparse PK sampling scheme if needed for safety purposes. Shortening the study further to 21 days (~ 30 % shorter than linear t1/2), was not adequate to estimate Km precisely. Conclusion: Optimal design approach is a powerful tool to expedite the FIH study of biologics, maximize the certainty of the TMDD parameter estimates, minimize subject inconvenience and facilitate study conduct.
