Effects of neonatal extracorporeal membrane oxygenation circuits on drug disposition

Abstract

Background: It has been observed that neonatal patients receiving extracorporeal membrane oxygenation (ECMO) require larger doses of sedatives than do non-ECMO patients to achieve similar levels of sedation. Objective: The purpose of this study was to determine whether drug disposition of midazolam and morphine, 2 commonly used sedatives, are affected by the ECMO circuit. Methods: This was a 3-part study. First, a retrospective analysis of prescription charts for 10 ECMO neonates and 10 neonates who had undergone cardiac surgery was conducted to confirm that the former group required larger doses of sedatives. Second, uptake of midazolam and morphine by the polyvinylchloride (PVC) and silicone components of the neonatal ECMO circuits was evaluated in vitro. Third, known concentrations of midazolam were injected into a complete neonatal ECMO circuit and analyzed after the oxygenator to determine the effects of the circuit components on the bioavailability of midazolam. Results: Mean ± SD total dose of midazolam for the ECMO neonates was 21,283 ± 10,446 μg versus 9599 ± 5504 μg for the cardiac surgery group ( P = 0.008). Mean ± SD total dose of morphine for the ECMO neonates was 2364 ± 1280 μg versus 1158 ± 457 μg for the cardiac surgery group ( P = 0.017). There was significant uptake (68%) of midazolam by the PVC and silicone components of neonatal ECMO circuits, whereas morphine uptake of 16% was observed only during contact with the PVC component. After 40 minutes of running a new ECMO circuit, the concentration of midazolam leaving the circuit was <50% of that injected. Conclusions: There is significant interaction between the PVC and silicone components of an ECMO circuit and the sedative drugs that pass through them. These interactions may affect the bioavailability of these drugs and hence the doses required to induce a therapeutic effect.