A recent in vitro study [13] reported meropenem and vancomycin recovery of 89% and 67% at 180 minutes in neonatal circuits that used a centrifugal pump and polypropylene hollow-fiber membrane oxygenators. Whereas the meropenem recovery at 180 minutes is comparable to our results, vancomycin recovery was much lower in the neonatal circuits. In contrast, the fentanyl and midazolam circuit losses seen moreover in this study are consistent with the results of the neonatal circuit studies [13,26-28]. Morphine appears to be relatively stable in both neonatal and adult ECMO circuits and may be the preferred analgesic during ECMO. Future clinical studies should compare the efficacy of different classes of drugs to rationalize sedation and analgesia during ECMO.

Studies that compare drug losses in clinically used versus new neonatal circuits have demonstrated significant variability in drug sequestration between the used and new circuits [10,13,28]. Consequently, it is still unclear whether there is saturation of the drug-binding sites in the ECMO circuit over time. Similarly, the effect of priming with various solutions on drug sequestration is not well characterized in the available literature. Drug sequestration in blood-primed circuits has been shown to be much lower than that in crystalloid-primed circuits [27]. It is possible that some of the blood components may compete with drugs for circuit-binding sites. Even though ECMO circuits are not routinely primed with blood prior to their use in adult patients, the circuits get primed with the patient’s own blood once ECMO is commenced.

In our study, we tried to replicate the clinical situation ex vivo which allowed us to study the interactions between the drug and the device in the absence of disease-related factors which independently can significantly alter PK [29,30]. Repeat dose experiments are required in long-term model systems to estimate the degree of circuit saturation with time. The concurrent presence of several other physically compatible study drugs in the circuit and control jars mimicked the clinical scenario in which patients receive these drugs concurrently, but the drugs may have had an impact on competitive binding to blood proteins or the circuit components. The presence of a reservoir bladder may have influenced the circuit drug losses.

Similarly, quantification of drug lost in control jars because of binding of drugs to the polypropylene container was not feasible. However, the results confirm the findings of neonatal ECMO circuit studies.ConclusionsThis ex vivo study highlights the role of the ECMO circuit in altering PK during Anacetrapib ECMO. These alterations are more pronounced for lipophilic drugs and may result in therapeutic failure. Morphine may be a useful alternative to fentanyl in a patient with escalating sedative and analgesic drug requirements.