Response to Hemostatic Complications During Neonatal Extracorporeal Membrane Oxygenation: Roller Pump and Centrifugal Pump Driven Circuits.

Abstract

To the Editor: Vermeer et al.1 recently published a case series and in vitro analysis of acute oxygenator failure manifested as a rise in preoxygenator pressure without evidence of gas transfer failure. As practitioners in the field who have fielded similar questions from colleagues describing similar issues, we applaud them for describing this problem that has been informally discussed in meetings and discussion boards for several years. The debate in neonatal extracorporeal life support (ECLS) between roller and centrifugal pumps has a long history.2 While the focus has primarily been on sources of increased hemolysis, the acute failure of the ECLS circuit by oxygenator clogging has received less attention. Vermeer et al.1 began using the RotaFlow in neonatal ECLS because of supply shortages and the availability and experience with adult ECLS. After a disturbing trend of complications, they embarked on a formal assessment of three systems (roller pump, RotaFlow, and PedIVAS) and associated failure rates in the neonatal population. The authors described anticoagulation, heat generation, and overall blood velocity as plausible mechanisms generating the cellular debris that led to the clogging phenomenon. In discussing the velocity, they came very close to what we would consider to be the fundamental problem inherent in the design of any centrifugal pump, which is their best efficiency point (BEP). The BEP is the point at which energy imparted to spinning the impeller generates the most forward bulk flow from the pump. The RotaFlow and PedivAS were designed to operate at 5 and 1.5 L/min, respectively. Schöps et al.3 used computational fluid dynamics to describe zones of recirculation, which do not contribute to the bulk flow of the blood, within adult devices operated at 1 L/min or less. Consequently, blood remains in the pump shear field for an extended time, significantly increasing the risk for hemolysis, which was demonstrated in vitro. Vermeer et al.1 have now provided in vivo evidence for the in silico and in vitro data provided by Schöps et al.3 However, we believe there is still more to the story. Several other factors are likely at play in neonatal ECLS, including cellular fragility4 and hemoglobin greater than 12–13 g/dl, which has been shown to be an independent factor leading to blood damage for neonatal ECLS regardless of pump type.5 A shrinking pool of devices and trained healthcare providers has led to the utilization of devices for the greatest number of patients (adults) in the most vulnerable population (neonates), and these are conditions for which these devices were never designed. Even the PediVAS, which is a pediatric device, is designed to operate at flow nearly three times that required for neonatal ECLS. We believe the study described by Vermeer et al.,1 Schöps et al.,3 Jenks et al.,5 and others in this space are important starting points for the ultimate discussion that a true neonatal ECLS centrifugal blood pump needs to be developed. We hope that such study continues and drives manufacturers to consider the needs of this population, ultimately allowing clinicians to more safely provide the life-saving support that this population needs.