Background: Although the pharmacokinetics of midazolam in critically ill children has been described, there are no such reports in extracorporeal membrane oxygenation.
Methods: The pharmacokinetics of midazolam and 1-hydroxy midazolam after continuous infusion (50-250 microg. kg(-1). h(-1)) were determined in 20 neonates undergoing extracorporeal membrane oxygenation. Patients were randomized into two groups: group 1 (n = 10) received midazolam extracorporeally (into the circuit), and group 2 received drug via central or peripheral access. Blood samples for determination of plasma concentrations were taken at baseline, 2, 4, 6, 12, 18, and 24 h, then every 12 h. Population pharmacokinetic analysis and model building was conducted using WinNonMix (Pharsight Corporation, Mountain View, CA). The 1-hydroxy midazolam/midazolam metabolic ratio was determined as a surrogate marker of cytochrome P450 3A activity.
Results: The parameter estimates (n = 19) were based on a one-compartment model with time-dependent change in volume of distribution. Volume (mean +/- standard error) expanded monoexponentially from the onset of extracorporeal membrane oxygenation to a maximum value, 0.8 l +/- 0.5 and 4.1 +/- 0.5 l/kg, respectively. Consequently, plasma half-life was substantially prolonged (median [range]) from onset to steady-state: 6.8 (2.2-39.8) and 33.3 (7.4-178) h, respectively. Total body clearance was determined as (mean +/- standard error) 1.4 +/- 0.15 ml. kg-1. min-1. The median metabolic ratio was 0.17 (0.03-0.9). No significant differences were observed between the two groups with respect to parameter estimates. Simulations of plasma concentration profiles revealed excess levels at conventional doses.
Conclusions: These results reveal significantly increased volume of distribution and plasma half-life in neonates receiving extracorporeal membrane oxygenation. Altered kinetics may reflect sequestration of midazolam by components of the extracorporeal membrane oxygenation circuit.