UCSF

In cases of severe lung or heart failure, extracorporeal membrane oxygenation (ECMO) is a life-saving therapy in which a patient’s blood is passed into a circuit outside of their body to provide respiratory support. The circuit’s main component is the membrane oxygenator that drives oxygen into the blood from a sweep gas source and removes excess carbon dioxide from the blood. At present, clinical use of ECMO is limited by its high risk profile, owing to two intertwined risks: thrombosis from the large circuit, and bleeding from the anticoagulation needed to prevent thrombosis. Improvements to the gas exchange efficiency and hemocompatibility of the oxygenator could enable the development of a longer-term supportive ECMO therapy, intended as a bridge-to-transplant or destination therapy for chronic lung failure. Here we describe a novel blood oxygenator concept based on parallel plate silicon membranes developed for high precision geometry, mechanical rigidity, and high efficiency membrane transport. Using these membranes, we create blood oxygenator prototypes consisting of arrays of silicon membranes, and endeavor to improve the efficiency and hemocompatibility of this concept.

First, multiple types of silicon membranes were evaluated systematically for mechanical rigidity and oxygen exchange efficiency, indicators of suitability for a future oxygenator. The combination of a silicon micropore membrane (SµM) and a 5 µm-thick polydimethylsiloxane (PDMS) layer maximized both qualities, withstanding over 260 cmHg of applied pressure and producing 0.03 mL/min of O2 flux. These membranes were then assembled into prototype flow cells, and tested for in vitro and in vivo oxygenation, successfully yielding an oxygen permeability of 1.92 ± 1.04 ml O2 STP/min/m2/cmHg. From this benchmark, we then attempted to optimize the surface hemocompatibility of the Si-PDMS composite through application of multiple polyethylene glycol (PEG)-based coatings. Although successful application of PEG to the surfaces was demonstrated, none of the coatings appeared to reduce protein adhesion to the SµM -PDMS membranes. Finally, we inserted turbulence-inducing spacer meshes into the channels of the SµM-PDMS prototypes to disrupt the transport boundary layer adjacent to the membranes, with the goal of substantially improving oxygenation. Though a threefold increase in oxygen flux was observed in vitro with the spacer meshes, the disruptive turbulence resulted in thrombosis and channel occlusion within the channels despite heavy anticoagulation of the blood. In summary, the work in this dissertation demonstrates the successful construction and testing of SµM-PDMS oxygenator prototypes, laying the foundation for future work to optimize this concept and create a large-scale blood oxygenator that can expand the clinical use of this life-saving therapy.