Population pharmacokinetics of midazolam and its metabolites during venoarterial extracorporeal membrane oxygenation in neonates

Midazolam is used to sedate children during extracorporeal membrane oxygenation (ECMO). Pharmacokinetic changes are expected because of extracorporeal circulation and maturation. We present a population pharmacokinetic model for midazolam and its major metabolites in neonates during venoarterial ECMO. We studied 20 neonates on venoarterial ECMO, with a median postnatal age of 0.79 (range 0.17-5.8) days and a bodyweight of 3.0 (range 2.7-3.9) kg at the onset of ECMO. The median ECMO duration was 124 (range 70-275) hours. Serum concentrations were measured at the initiation and discontinuation of the midazolam infusion (100-300 microg/kg/h). Analysis of concentrations of midazolam, 1-hydroxymidazolam and its glucuronide were performed using nonlinear mixed-effects modelling. A two-compartment model for midazolam and a one-compartment model for the metabolites 1-hydroxymidazolam and hydroxymidazolam glucuronide adequately described the data, with allometric scaling of all parameters. Following the start of ECMO, the volume of distribution of midazolam increased from 4.29 to 14.6 L/3 kg, with an elimination half-life of 1.85 hours. The median midazolam and 1-hydroxymidazolam clearance values increased 3-fold within the first 5 days (up to 1.38 and 5.31 L/h/3 kg, respectively), whereas hydroxymidazolam glucuronide clearance remained constant at 0.18 L/h/3 kg. Interpatient variability estimates of midazolam, 1-hydroxymidazolam and hydroxymidazolam glucuronide clearance and midazolam and hydroxymidazolam glucuronide volumes of distribution varied between 87% and 129%. Concomitant inotropic infusion increased hydroxymidazolam glucuronide clearance by 23%. After allometric scaling, clearance of midazolam and 1-hydroxymidazolam increases as a result of maturation or recovery from critical illness. In ECMO patients weighing 2.7-3.9 kg, continuously infused midazolam doses of 300 microg/kg/h for 6 hours and 150 microg/kg/h thereafter provide adequate serum concentrations for sedation. The dose must be increased substantially after 5-7 days. Hydroxymidazolam glucuronide accumulates during ECMO, providing an increased proportion of the overall effect, up to 34% after 7 days. Large unexplained interpatient variability warrants careful titration of sedation and adverse effects