Characterization of Protein Binding on an Extracorporeal Membrane Oxygenation (ECMO) Circuit Following the Priming Procedure

Purpose of the Study Extracorporeal membrane oxygenation (ECMO) circuit priming procedures are crucial for removing air bubbles and the albumin in priming solution is thought to create a protective layer against blood exposure and protein binding.1 However, the binding profile to the ECMO circuit after priming remains unexplored. This study aims to characterize proteins bound to the ECMO circuit after priming, and the corresponding priming solutions using data-dependent acquisition mass spectrometry (DDA-MS). Main Results Information on the ECMO circuit information, priming procedures, circuit collection, sample preparations, and experimental protocols for DDA-MS are detailed in Appendix 1, Supplemental Digital Content, https://links.lww.com/ASAIO/B305. An unused ECMO circuit was collected after 4 hour priming procedures, and circuit samples from five different sites (Figure 1) were collected for this study.Figure 1.: The basic diagram of ECMO circuit. Circuit samples were collected from site 1: postvenous cannula, site 2: prepump head, site 3: post-pump head to oxygenator, site 4: post-oxygenator, site 5: pre arterial cannula. ECMO, extracorporeal membrane oxygenation.The median total protein concentration of the five ECMO circuit samples was 3.24 μg/ml. Specifically, site 1 showed the highest total protein concentration of 51.10 μg/ml, followed by site 2 and site 5 of 12.72 μg/ml and 3.24 μg/ml, respectively, whereas no protein was detected from site 3 nor site 4. The total protein concentration of 20% albumin solution, Plasma-Lyte 148 solution, and the priming reagent mix were 204.00, 15.71, and 25.63 μg/ml, respectively. Due to insufficient protein concentration across other sites, only the site 1 ECMO circuit sample, 20% albumin solution, Plasma-Lyte 148 solution, and the priming reagent were analyzed using DDA-MS. In total, 129 different proteins were identified among all samples, with 121 proteins specifically identified bound to the ECMO circuit sample. Although 27, 33, and 32 proteins were identified among 20% albumin solution, Plasma-Lyte 148 solution, and the priming reagent mix, respectively. The Venn diagram of the identified proteins demonstrated the overlap between samples (eFigure 1, Supplemental Digital Content, https://links.lww.com/ASAIO/B305). Principal Component Analysis (PCA) analysis showed minimal separation between the 20% albumin solution and the priming reagent, whereas significant separation between Plasma-Lyte 148 solution and the circuit samples (eFigure 2, Supplemental Digital Content, https://links.lww.com/ASAIO/B305). The detailed list of identified proteins for each sample is presented in eTable 1, Supplemental Digital Content, https://links.lww.com/ASAIO/B305. The pathway analysis of the 121 proteins bound to the ECMO circuit sample showed involvement across 67 pathways divided into 15 categories. Specifically, immune system, hemostasis, and signal transduction represented the top three pathway categories with the greatest number of pathways involved (Figure 2). The detailed pathway analysis results are included in eTable 2, Supplemental Digital Content, https://links.lww.com/ASAIO/B305.Figure 2.: Pathway analysis of the 121 proteins bound to the ECMO circuit. ECMO, extracorporeal membrane oxygenation.Discussion In clinical practice, priming of the ECMO circuit before use in patient care serves as a preventive measure, which mitigates the blood exposure to the ECMO circuit through precoating with albumin. In this study, we demonstrated that besides albumin, a significant number of other proteins were identified bound to the ECMO circuit sample. In addition, complex protein compositions were identified in 20% albumin solution and Plasma-Lyte 148 solution. According to the product information,2 the 20% albumin solution is manufactured from human plasma using predominantly chromatographic techniques. Although it is prepared by undergoing rigorous purification processes, trace amounts of other residual proteins may remain in the final product due to the limitation of the purification methodology. Despite the fact that the Plasma-Lyte solution, being declared as a sterile, nonpyrogenic isotonic solution,3 various proteins may still be introduced during the manufacturing process. However, given the low protein concentration detected in the original Plasma-Lyte solution and the subsequent dilution with the 20% albumin solution before priming, the total quantity of identified proteins is very limited. The complex bound proteome of the ECMO circuit sample could be attributed by several factors and indicates potential impacts to patients during the following ECMO support. In this study, the total protein amount bound to different sites showed significant intra-circuit variation, with site 1 having the highest protein concentration. This could be due to the flow pressure variation between different sites of the ECMO circuit during priming process, which leads to differences in protein adsorption between sites.4 Our study showed a significantly greater number of proteins bound to site 1 ECMO circuit sample when compared with the 20% albumin solution and Plasma-Lyte 148 solution. This difference may be attributed to preexisting proteins on the ECMO circuit, which potentially be introduced during the manufacture or circuit setup before priming, or to limitation of protein identification in the 20% albumin solution. The overwhelming presence of albumin in the 20% albumin solution could potentially cause an ion suppression effect of DDA-MS, which obscures the detection of other less abundant proteins and hinders the comprehensive protein identification.5 Although for the ECMO circuit sample, the albumin was diluted by mixing with Plasma-Lyte 148 solution, but also the binding protein composition depends on the binding affinity and evolves with the priming time. Specifically, the less abundant proteins with higher binding affinity in the priming solutions were selectively accumulated to the ECMO circuit during the priming procedures. It is hypothesized that upon connection to the patient, these bound proteins would continue to interact with other blood components.6 The pathway analysis based on proteins bound to the ECMO circuit showed involvement of multiple pathways, which may have potential impacts to patients during ECMO support. However, this result needs to be interpreted with caution, due to the inability to confirm whether bound proteins have potentiating or inhibiting effects. Future studies will need to further investigate the impacts of these bound proteins on the patient outcomes during ongoing support. Given the variability in ECMO circuit priming solutions and procedures across different institutions,7 the method described in this paper offers a potential framework for future research aimed at characterizing the proteins bound to ECMO circuits following different priming procedures. This will enable a deeper understanding of how these clinical practice variations may uniquely affect patient outcomes during subsequent ECMO support. In conclusion, this study characterized proteins bound to the ECMO circuit after priming procedures and detailed protein profiles in the corresponding priming solutions. The total protein concentration bound to the primed ECMO circuit showed significant intra-circuit variation. Complex protein profiles were identified among the primed ECMO circuit and the corresponding priming solutions. Acknowledgment The authors like to thank the participants and their families for participating in this study. The authors like to acknowledge funding from the Australian National Health and Medical Research Council (NHMRC) grant APP1129317. Additionally, the authors also like to acknowledge the staff from the RCH PICU and perfusion teams at the RCH for obtaining the ECMO circuit.