Deep hypothermia reduces the predictive accuracy of the Eleveld propofol model: Population pharmacokinetic modelling in cardiac surgery

OBJECTIVES: To evaluate the predictive performance of the Eleveld target-controlled infusion model for propofol during deep hypothermia, characterise temperature-dependent changes in propofol pharmacokinetics, and explore the relationship between depth of anaesthesia monitoring and the perioperative stress response during deep hypothermic circulatory arrest. DESIGN: Prospective, single-centre, low-intervention study. SETTING: Tertiary cardiac surgery centre. PATIENTS: Thirty adult patients undergoing elective pulmonary endarterectomy with cardiopulmonary bypass and deep hypothermic circulatory arrest. METHODS: Serial blood samples (T0-T13) were obtained to determine propofol serum concentrations and endogenous norepinephrine levels. Model performance of the Eleveld target-controlled infusion system was assessed using linear regression across temperature phases, and population pharmacokinetic analysis was performed using nonlinear mixed-effects modelling. RESULTS: Predictive performance deteriorated during cooling and improved during rewarming. A two-compartment model identified temperature as a significant covariate on propofol clearance, with marked reduction during hypothermia and relative increase during rewarming. Processed EEG monitoring demonstrated profound cortical suppression (mean suppression ratio 98.4%, PSI 0.4). Despite this, norepinephrine levels decreased during cooling and increased significantly during reperfusion, indicating activation of the stress response in the absence of cortical activity. No intraoperative awareness was detected. CONCLUSIONS: Deep hypothermia alters propofol pharmacokinetics, reducing clearance and impairing model performance, while rewarming increases clearance. Autonomic stress responses may occur despite profound cortical suppression, highlighting limitations of EEG-based depth of anaesthesia monitoring during deep hypothermic circulatory arrest. These findings support temperature-adaptive dosing and warrant confirmation in prospective randomised studies.