(152f) Role of Model-Based Design of Experiments in Pharmacokinetic/Pharmacodynamic Modeling and the Development of Personalized Medicines – Case of COVID-19

The advent of personalized medicine has opened promising avenues for the prevention and treatment of diseases by tailoring drug properties and efficacy to the individual needs of patients. This study explores ways to enable the production of personalized solid dosage forms and optimal doses by leveraging personalised pharmacokinetic/pharmacodynamic (PK/PD) models from clinical data. Currently, PK/PD models, which represent the whole population of patients, are built using clinical data obtained from blood analysis and preset sampling procedures e.g. regular sampling intervals. To build accurate personalized PK/PD models, optimal sampling strategies are critical to maximize the data information content and minimize model prediction uncertainties.

A Model-Based Design of Experiments (MBDoE) is proposed to build predictive and reliable personalized PK/PD based on flexible administration and blood sampling strategies and variable drug doses. The MBDoE can be implemented iteratively to reduce model uncertainty below a certain threshold while maximizing drug efficacy and reducing side effects. Consequently, the proposed method can deliver the optimal doses, administration frequency, and sampling strategy for each individual/patient. Consequently, the method can help reduce the overall costs and time of the clinical trials and improve their planning and reliability. Besides, more personalised, and patient-centric medicines can be identified and produced on demand. The initial model PK/PD parameters can be obtained from early clinical trials which can be refined and improved through MBDoE during later clinical trials (Phase 2 and Phase 3).

To demonstrate and validate the capabilities of the proposed method in the context of personalized dosage formulations, a case study focussed on the treatment of COVID-19 using Dexamethasone (Dex) is considered. Through the implementation of systematic MBDoE methodology which optimizes the dosage formulations and timing for blood sample collection, our study aims to reduce inherent risks by reducing uncertainties and facilitating the development of tailored PK/PD models. These reliable personalized models then inform the optimization of the doses and dosage formulations to meet the individual needs, promising a significant strides towards mitigating side effects while preserving and maximizing therapeutic efficacy.

Our findings suggest that personalized clinical trials, underpinned by rigorous PK/PD modelling and optimization strategies, hold considerable promise for enhancing drug safety and efficacy in the treatment of COVID-19 and potentially other diseases. This approach not only contributes to the evolving landscape of personalized medicine but also holds a strategic endeavour towards more patient-centric healthcare solutions.

Acknowledgements

The authors acknowledge funding from the UK Engineering and Physical Sciences Research Council (EPSRC), for Made Smarter Innovation – Digital Medicines Manufacturing Research Centre (DM2), EP/V062077/1.