03 (GraphPad Software, Inc., La Jolla, CA, USA).ResultsThe ex vivo circuits were maintained under physiological conditions for 24 hours with no complications during the run. Four hundred samples sellckchem (80 per drug) were analyzed. The changes in drug concentrations in the ex vivo model are summarized in Figures Figures11 and and2,2, and the actual drug concentrations determined in plasma samples from the control jars and the ECMO circuits are presented in Table Table1.1. In this paper, only the data for fentanyl, morphine, midazolam, meropenem, and vancomycin are presented. Validated assays are being developed for the remaining study drugs, and the results will be made available in due course.Figure 1Percentage of drug remaining in extracorporeal membrane oxygenation circuits and the controls plotted against time.

Lipohilic drugs such as fentanyl and midazolam were significantly sequestered in the circuit despite being stable in the controls. Morphine …Figure 2Percentage of meropenem and vancomycin remaining in extracorporeal membrane oxygenation circuits and the controls plotted against time. Meropenem was unstable in the controls; however, the extracorporeal membrane oxygenation circuit losses significantly …Table 1Measured study drug concentrations in the control jars and the extracorporeal membrane oxygenation circuits at 5 minutes and at 1, 6, 12, and 24 hoursTesting confirmed that all baseline plasma samples were free of study drugs. There were no statistically significant differences in drug recoveries between the four circuits.

The mean drug recoveries from the circuits and the control jars at 24 hours relative to baseline were, respectively, fentanyl 3% and 82%, morphine 103% and 97%, midazolam 13% and 100%, meropenem 20% and 42%, and vancomycin 90% and 99%. Up to 70% of fentanyl and 50% of midazolam were lost in the circuit within the first hour of ECMO run. Fentanyl levels were undetectable in the circuit by 24 hours. This may be related to the lipophilicty of these drugs. Morphine, which is less lipophilic than fentanyl, was stable in both the circuit and the controls. Antibiotics were less significantly affected. The hydrophilic and minimally protein-bound drug meropenem was stable in the circuit and the controls in the first 120 minutes, and 62% of the drug was recovered from the circuit at 6 hours. This was statistically significant (P = 0.

01) even after accounting for spontaneous degradation (21%). There was no significant loss of the moderately protein-bound AV-951 hydrophilic drug vancomycin in the circuit at 12 hours (P = 0.41) or 24 hours (P = 0.26).DiscussionThis is the first systematic investigation of drug disposition in the adult ECMO circuitry. The findings highlight the role of the circuit in altering the PK of sedative, analgesic, and antibiotic drugs during ECMO and clearly show that there is considerable between-drug variability in the degree of drug sequestration.