Population Pharmacokinetics of Midazolam and Metabolites during Venoarterial Extracorporeal Membrane Oxygenation in Neonates
Maurice J. Ahsman (1), Manon Hanekamp (2), Enno D. Wildschut (2), Dick Tibboel (2), Ron A.A. Mathot (1)
(1) Department of Hospital Pharmacy (Clinical Pharmacology Unit), Erasmus University Medical Center, ‘s-Gravendijkwal 230, P.O. Box 2040, 3000 CA Rotterdam, The Netherlands;
Objectives: Midazolam is used to sedate neonates during extracorporeal membrane oxygenation (ECMO). During the procedure, midazolam pharmacokinetics are expected to change due to extracorporeal circulation and maturation. We present a population pharmacokinetic model for midazolam and major metabolites in neonates during venoarterial ECMO.
Methods: We included 20 patients on venoarterial ECMO, with a median postnatal age (range) of 0.79 (0.17-5.8) days and a body weight of 3.0 (2.7-3.9) kg at onset of ECMO. Median ECMO duration was 124 (70-275) h. Plasma concentrations (293 in total) were measured at introduction and discontinuation of midazolam infusion (100 - 300 mg/kg/h). Nonlinear mixed-effects modeling (NONMEM) was used to describe the pharmacokinetics of midazolam (MDZ), 1-hydroxymidazolam (OHM) and its glucuronide (HMG). A 2-compartment model for MDZ and 1-compartment for OHM and HMG were used to describe the data. PK parameters were allometrically scaled.
Results: Median MDZ volume of distribution in a median patient increased from 1.4 to 4.9 L/kg (98-343 L/70kg) within the first 8h. Median MDZ and OHM clearance increased 3-fold in the first 5 days, from 2.6 to 7.6 mL/kg/min and 9.3 to 29 mL/kg/min (4.9-15 and 18-56 L/h/70kg, respectively), whereas HMG clearance remained constant (1.0 mL/kg/min or 1.9 L/h/70kg). Postnatal age was positively correlated with clearance of OHM, but not MDZ or HMG. Interpatient variability estimates on clearances and volumes of distribution ranged from 87% to 129%. Concomitant catecholamines increased HMG clearance by 23%. HMG accumulated during ECMO whereas MDZ and OHM were effectively cleared; median HMG concentration after 7 days was 4500 ng/mL. Simulations in 1000 individuals show that a dose regimen of 0.50 mg/kg of midazolam followed by continuous infusion of 0.30 mg/kg/h for 6 h, and 0.15 mg/kg/h thereafter, provides adequate plasma concentrations (400 ng/mL MDZ) for sedation of ECMO-patients. The dose will have to be increased after approximately 5 days to compensate for increased clearance.
Conclusions: During ECMO, MDZ and OHM clearance increase. Collinearity complicates differentiation between changes caused by maturation, the extracorporeal circuit, or disease progression. A dose of 0.50 mg/kg followed by 0.30 mg/kg/h for 6 h, and 0.15 mg/kg/h thereafter, will lead to adequate sedation of ECMO-patients. Large unexplained interpatient variability warrants careful titration on sedation and side effects.
