Supplementary file 1_Pharmacokinetic and pharmacodynamic analyses of nafamostat in ECMO patients: comparing central vein and ECMO machine samples.pdf

ObjectivesTo better understand nafamostat mesylate (NM) dose requirements during extracorporeal membrane oxygenation (ECMO), this study investigated its pharmacokinetic/pharmacodynamic (PK/PD) properties by comparing samples from the systemic circulation of patients and from the ECMO circuit. It specifically examined the relationship between NM concentration and activated partial thromboplastin time (aPTT) changes, aiming to provide a foundation for future dosing optimization.
MethodsIn this prospective study, 24 ECMO patients received a continuous infusion of NM through a dedicated stopcock located before the ECMO pump. This placement targets the anticoagulant effects of NM specifically to the ECMO circuit without substantially affecting the patient’s overall coagulation status. The starting dose was 15 mg/h, adjusted to keep the aPTT within a target range of 40–80 s. Blood samples were collected from both the patient’s central venous catheter and the ECMO circuit for PK/PD analysis using a nonlinear mixed effects model.
ResultsThe PK profiles of NM, derived from samples taken from both the patient’s catheter and the ECMO circuit, were best described by a two-compartment model. In the PK/PD models, the effect of NM on prolonging aPTT was described using a turnover model. NM was shown to inhibit the decrease in aPTT in the turnover model. In the patient model, the maximum inhibitory effect (Imax) of NM on the reduction of aPTT was 35.5%, and the concentration of NM required to achieve half of this maximum effect (IC50) was 350 μg/L. On the other hand, in the ECMO model, the Imax for aPTT reduction was 43.6%, with an IC50 of 581 μg/L.
ConclusionThe PK/PD models developed from samples collected from both the patient and the ECMO circuit indicate significant differences in PD. Given the observed variability and the high risk of bleeding in ECMO patients, a predictive model incorporating these differences and patient-specific variables could significantly improve anticoagulation management.