Challenges in collecting pharmacokinetic and pharmacodynamic information in an intensive care setting: PK/PD modelling of clazosentan in patients with aneurysmal subarachnoid haemorrhage

This paper describes the pharmacokinetic/pharmacodynamic modelling of clazosentan in patients with aneurysmal subarachnoid haemorrhage (aSAH), and the impact of collecting data in an intensive care unit (ICU) setting. Factors influencing data quality, analysis, and interpretation are provided with recommendations for future clinical studies in ICU settings. CONSCIOUS-2 was a phase III study involving 1,157 patients with aSAH. Secured by surgical clipping, patients were infused with clazosentan or placebo for up to 14 days post-aSAH. Clazosentan exposure relationships with vital signs, QT intervals, and AST/ALT values as well as efficacy and safety endpoints were characterised using population PK/PD and logistic regression models. Clazosentan clearance was influenced by age, sex, Asian origin, and disease status at baseline, and increased with time. Volume of distribution showed a sex difference. Exposure had no relationship with any efficacy endpoint or ALT/AST values, but was related to the increasing probability of lung complications. Blood pressure decreased proportionally to clazosentan concentrations, and the presence of clazosentan was associated with QT interval increases. Implausible values in the concentration data reflect the specific ICU challenges, possibly arising from PK sampling from the infusion arm or haemodilution. Population PK/PD modelling of CONCIOUS-2 data provided clinically relevant knowledge about various effects of clazosentan in the aSAH patient population in a real clinical setting. The quality of data and analyses could be improved by the collection of additional data and stricter training of study personnel. Differences in clinical practice between sites and geographical regions are more challenging to overcome.