Dose optimization is generally based on probability of target attainment (PTA), ensuring the selected dose can maintain the drug exposure above the pre-defined target threshold. However, the target threshold is not always available for some treatments, especially in pediatric dosing. This work introduces a statistical framework for dose optimization when the pharmacokinetic (PK) or pharmacokinetic-pharmacodynamic (PKPD) target is unknown. The framework involves simulating PK or PKPD outcomes in typical patients with standard dose regimen to establish a reference profile, then simulating outcomes at various dose scenarios (i.e., all available tablet strengths) in the population of interest. Statistical distances between the empirical cumulative distribution functions (ECDFs) of the PK or PKPD outcomes from all possible dosing regimens are calculated and compared to the reference profile. The optimal dose assignment is commonly based on the most equivalent PK or PKPD outcomes, determined by the dose with the minimum statistical distance to the reference profile. This approach can be applied to both models with categorical and continuous covariates. For known PK and PKPD target outcomes, the optimal dose is selected to maintain the outcome above the assigned target, while in populations with unknown targets, optimal dosing is chosen to generate equivalent PK and PKPD outcomes as the typical population. The developed pharmacometric method provides an unbiased, robust, and reproducible framework for dose optimization in populations with unknown target levels. It offers a valuable approach when traditional PTA-based methods are not applicable due to the absence of established target thresholds, maximizing overall treatment outcomes by balancing the risk of sub-therapeutic exposure and toxicity of the available dose strengths. The developed methods was also implementing in a freely accessible Shiny web-application to maximize its usefullness.
Reference:
- Chotsiri, P., Yodsawat, P., Hoglund, R. M., Simpson, J. A., & Tarning, J. (2025). Pharmacometric and statistical considerations for dose optimization. CPT: pharmacometrics & systems pharmacology, 14(2), 279–291. https://doi.org/10.1002/psp4.13271
- https://pharmacology.shinyapps.io/dose_optimization/