Extracorporeal membrane oxygenators with light-diffusing fibers for treatment of carbon monoxide poisoning: Experiments, mathematical modeling, and performance assessment with unit cells.

UNLABELLED: BACKGROUND AND OBJECTIVES: Approximately 50,000 emergency department visits per year due to carbon monoxide (CO) poisoning occur in the United States alone. Tissue hypoxia can occur at very low CO concentration exposures because CO binds with a 250-fold higher affinity than oxygen to hemoglobin. The most effective therapy is 100% hyperbaric oxygen (HBO) respiration. However, there are only a limited number of cases with ready accessibility to the specialized HBO chambers. In previous studies, we developed an extracorporeal veno-venous membrane oxygenator that facilitates exposure of blood to an external visible light source to photo-dissociate carboxyhemoglobin (COHb) and significantly increase CO removal from CO-poisoned blood (photo-extracorporeal veno-venous membrane oxygenator [p-ECMO]). One objective of this study was to describe in vitro experiments with different laser wavelength sources to compare CO elimination rates in a small unit-cell ECMO device integrated with a light-diffusing optical fiber. A second objective was to develop a mathematical model that predicts CO elimination rates in the unit-cell p-ECMO  device design upon which larger devices can be based.

STUDY DESIGN/MATERIAL AND METHODS: Two small unit-cell p-ECMO devices consisted of a plastic capillary with a length and inside diameter of 10 cm and 1.15 mm, respectively. Either five (4-1 device) or seven (6-1 device) gas exchange tubes were placed in the plastic capillary and a light-diffusing fiber was inserted into one of the gas exchange tubes. Light from lasers emitting either 635 nm or 465 nm wavelengths was coupled into the light-diffusing fiber as oxygen flowed through the gas exchange membranes. To assess the ability of the device to remove CO from blood in vitro, the percent COHb reduction in a single pass through the device was assessed with and without light. The Navier Stokes equations, Carreau-Yesuda model, Boltzman equation for light distribution, and hemoglobin kinetic rate equations, including photo-dissociation, were combined in a mathematical model to predict COHb elimination in the experiments.

RESULTS: For the unit-cell devices, the COHb removal rate increases with increased 635 nm laser power, increased blood time in the device, and greater gas exchange membrane surface-to-blood volume ratio. The 6-1 device COHb half-life versus that of the 4-1 device with 4 W at 635 nm light was 1.5 min versus 4.25 min, respectively. At 1 W laser power, 635 nm and 465 nm exhibited similar CO removal rates. The COHb half-life times of the 6-1 device were 1.25, 2.67, and 8.5 min at 635 nm (4 W), 465 nm (1 W), and 100% oxygen only, respectively. The mathematical model predicted the experimental results. An analysis of the in vivo COHb half-life of oxygen respiration therapy versus an adjunct therapy with a p-ECMO device and oxygen respiration shows a reduction from 90 min to as low as 10 min, depending on the device design.

CONCLUSION: In this study, we experimentally studied and developed a mathematical model of a small unit-cell ECMO device integrated with a light-diffusing fiber illuminated with laser light. The unit-cell device forms the basis for a larger device and, in an adjunct therapy with oxygen respiration, has the potential to remove COHb at much higher rates than oxygen therapy alone. The mathematical model can be used to optimize the design in practical implementations to quickly and efficiently remove CO from CO-poisoned blood.