PEDS4: A Stacked Silicon Membrane Artificial Placenta Oxygenator to Treat Extreme Prematurity

Background: Each year in the United States, approximately 10,000 infants are spontaneously delivered between 20 and 23 weeks of gestation but cannot survive due to their underdeveloped lungs. An artificial placenta, which performs extracorporeal gas exchange, may offer a potential treatment for these patients by enabling survival and lung development outside the uterus. We previously demonstrated the feasibility of silicon membranes (SM) as gas exchange membranes. These rigid flat-plate membranes enable precision design of the blood flow path to optimize shear forces, laminar blood flow, pressure drop, priming volume, and gas-exchange. In this study, we build on our previous proof-of-concept studies and construct a clinical-scale microfluidic artificial placenta oxygenator using a stacked array of SM. Methods: Multilayered devices containing a total of 30 microfluidic channels and a membrane surface area of 144 cm2 were constructed. Each channel consisted of 2 SM, stacked using 500 μm-thick silicone gaskets spacers. They were aligned within a 3D-printed housing and sealed with silicone adhesive and epoxy. Completed devices were tested at a flow rate of 60 mL/min on the benchtop and in an extracorporeal pig model. Results: In the benchtop setup, devices demonstrated an oxygen flux (mL/min) of 0.431±0.006 (mean±standard error) and 0.493±0.005 at 1 and 2.5 PSI of sweep gas pressure, respectively. In the extracorporeal model, oxygen flux was 0.293±0.040, 0.481±0.058, 0.677±0.066, and 1.298±0.177 at 2.5, 5, 7.5, and 10 PSI, respectively of sweep gas pressure. Pressure drop across the device was 37 mmHg. Total priming volume was 9.2 mL. No gross thrombosis was noted during the study. Conclusion: This work demonstrates the feasibility of an SM-based oxygenator to provide clinical-scale extracorporeal gas exchange. The total oxygen flux measured by this device is approximately equal to the oxygen demand of a 20week fetus. Furthermore, the low pressure drop and small priming volume will enable pumpless operation. Future studies will focus on optimizing the blood flow path to maximize hemocompatibility.