Medhat is a final-year PhD candidate at the Department of Clinical Pharmacology and Pharmacy at Amsterdam University Medical Center. He holds both a Bachelor’s and Master’s degree in Bio-Pharmaceutical Sciences from Leiden University. His research focuses on population pharmacokinetic and pharmacodynamic (PK/PD) modeling to optimize drug therapy in complex patient groups, with a particular focus on the repurposing of imatinib. Specifically, he investigates how to balance efficacy and toxicity in patients with pulmonary arterial hypertension (PAH), and evaluates the changes in the pharmacokinetic changes in critically ill COVID-19 ARDS patients. Through his work, he aims to support individualized dosing strategies and improve therapeutic outcomes by integrating clinical pharmacology with advanced modeling approaches.

Abstract: Imatinib, a tyrosine kinase inhibitor used in oncology for chronic myeloid leukaemia (CML) and gastrointestinal stromal tumours (GIST), is being investigated for repurposing in COVID-19-associated acute respiratory distress syndrome (ARDS). However, dosing based on oncology targets may be ineffective in inflammatory conditions due to altered pharmacokinetics. We analysed total and unbound imatinib concentrations in COVID-19 ARDS patients compared to CML and GIST patients. Increased alpha-1-acid glycoprotein (AAG) during inflammation explained higher and more variable total imatinib levels, while inflammation and IL-6 receptor inhibitors affected metabolism. Notably, unbound drug levels remained unchanged but standard dosing showed no clinical benefit in COVID-19 ARDS. This highlights the importance of prioritizing unbound over total drug concentrations for accurate dosing in inflammatory states, where protein binding variability complicates interpretation. We propose a decision flowchart to guide clinicians on monitoring strategies. Additionally, simulation studies comparing models jointly predicting total and unbound concentrations revealed consistent parameter estimates but differed in accounting for covariate effects on protein binding, emphasizing the need for appropriate modeling to avoid misinterpretation. In conclusion, successful drug repurposing depends on integrating pharmacometric insights to account for disease-specific physiological changes and support precise dose optimization.