Search

When using the extracorporeal capillary membrane oxygenator (sample A) for ECMO treatments of COVID-19 severely ill patients, which is dominantly used in Japan and worldwide, there is a concern about the risk of SARS-CoV-2 scattering from the gas outlet port of the membrane oxygenator. Terumo has launched two types of membranes (sample A and sample B), both of which are produced by the microphase separation processes using polymethylpentene (PMP) and polypropylene (PP), respectively. However, the pore structures of these membranes and the SARS-CoV-2 permeability through the membrane wall have not been clarified. In this study, we analyzed the pore structures of these gas exchange membranes using our previous approach and verified the SARS-CoV-2 permeation through the membrane wall. Both have the unique gradient and anisotropic pore structure which gradually become denser from the inside to the outside of the membrane wall, and the inner and outer surfaces of the membrane have completely different pore structures. The pore structure of sample A is also completely different from the other membrane made by the melt-extruded stretch process. From this, the pore structure of the ECMO membrane is controlled by designing various membrane-forming processes using the appropriate materials. In sample A, water vapor permeates through the coating layer on the outer surface, but no pores that allow SARS-CoV-2 to penetrate are observed. Therefore, it is unlikely that SARS-CoV-2 permeates through the membrane wall and scatter from sample A, raising the possibility of secondary ECMO infection. These results provide new insights into the evolution of a next-generation ECMO membrane.

The objective of this study is to clarify the pore structure of ECMO membranes by using our approach and theoretically validate the risk of SARS-CoV-2 permeation. There has not been any direct evidence for SARS-CoV-2 leakage through the membrane in ECMO support for critically ill COVID-19 patients. The precise pore structure of recent membranes was elucidated by direct microscopic observation for the first time. The three types of membranes, polypropylene, polypropylene coated with thin silicone layer, and polymethylpentene (PMP), have unique pore structures, and the pore structures on the inner and outer surfaces of the membranes are completely different anisotropic structures. From these data, the partition coefficients and intramembrane diffusion coefficients of SARS-CoV-2 were quantified using the membrane transport model. Therefore, SARS-CoV-2 may permeate the membrane wall with the plasma filtration flow or wet lung. The risk of SARS-CoV-2 permeation is completely different due to each anisotropic pore structure. We theoretically demonstrate that SARS-CoV-2 is highly likely to permeate the membrane transporting from the patient’s blood to the gas side, and may diffuse from the gas side outlet port of ECMO leading to the extra-circulatory spread of the SARS-CoV-2 (ECMO infection). Development of a new generation of nanoscale membrane confirmation is proposed for next-generation extracorporeal membrane oxygenator and system with long-term durability is envisaged.

Hemoconcentration membranes used in cardiopulmonary bypass require a pore structure design with high pure water permeability, which does not allow excessive protein adsorption and useful protein loss. However, studies on hemoconcentration membranes have not been conducted yet. The purpose of this study was to analyze three-dimensional pore structures and protein fouling before and after blood contact with capillary membranes using the tortuous pore diffusion model and a scanning probe microscope system. We examined two commercially available capillary membranes of similar polymer composition that are successfully used in hemoconcentration clinically. Assuming the conditions of actual use in cardiopulmonary bypass, bovine blood was perfused inside the lumens of these membranes. Pure water permeability before and after bovine blood perfusion was measured using dead-end filtration. The scanning probe microscopy system was used for analysis. High-resolution three-dimensional pore structures on the inner surface of the membranes were observed before blood contact. On the other hand, many pore structures after blood contact could not be observed due to protein fouling. The pore diameters calculated by the tortuous pore diffusion model and scanning probe microscopy were mostly similar and could be validated reciprocally. Achievable pure water permeabilities showed no difference, despite protein fouling on the pore inlets (membrane surface). In addition, low values of albumin sieving coefficient are attributable to protein fouling that occurs on the membrane surface. Therefore, it is essential to design the membrane structure that provides the appropriate control of fouling. The characteristics of the hemoconcentration membranes examined in this study are suitable for clinical use.