Pharmacometrics to facilitate decision-making in pain research and perioperative management.

Pharmacometrics is a discipline tightly associated with drug development. Nevertheless, the use of pharmacometrics is not limited to the process of developing a drug per se and could also be applied to support other aspects such as the characterization of biomarkers in areas with low success rates. In this thesis, we aimed to mitigate the lack of quantitative tools for the clinical management of pain and perioperative care, two areas that could benefit from the application of model-informed precision dosing (MIPD). Particularly, we focused on developing population pharmacokinetics/pharmacodynamics (popPK/PD) methods to improve the intraperioperative administration of opioids and neuromuscular blockade agents and assist in the selection of patients requiring red blood cells transfusions, which lack an optimal solution. Throughout the different sections, several examples of the application of MIPD in these clinical settings are presented. The thesis is organized as follows: The Introduction provides an overview of the evolution of pharmacometrics, its main components and the current state of the field to contextualise the methodologies that will be developed in the following sections. The Methods detail the clinical studies, measured variables and the different pharmacometric techniques that were used in this thesis. The Results provide a comprehensive description of the evaluation, validation and clinical applicability of the developed popPK/PD models, which constitute the main findings of this work. First, we present two pharmacodynamic models to relate painful stimulation with pupillary and movement reflexes in the presence of opioids. These models constitute a semi-mechanistic framework to evaluate the use of pupillary reflex as a biomarker for pain during surgical procedures, providing anaesthetists with a quantitative tool to individualise drug administration. Then, we propose a popPK/PD model to characterise the effects of rocuronium in post-tetanic counts during profound neuromuscular blockade. This work compares the performance of two published PK models and serves as a basis for optimising rocuronium administration during surgery. Later, a quantitative model to describe haemoglobin concentrations in the early perioperative period using variables collected in everyday surgical procedures is presented. This model was subsequently used to evaluate the risk of requiring a blood transfusion in different clinical situations and provide recommendations on the administration of blood and fluid therapy. The Discussion provides an overview of the highlights, the applicability, as well as the limitations of this work taking into consideration the present and future of the field. Finally, the Conclusions summarise the most relevant findings of this thesis in both English and Spanish. The Appendix contains two pieces of work not directly linked with the primary objective of this thesis. In the first one, an overview of PAINCARE, a public-private collaboration aiming to improve the clinical success rate of novel analgesics, is provided. The second one proposes a population PK model for piperacillin in critically ill children with or without continuous kidney replacement therapy, an area with high medical need due to the vulnerability of this population