Effect of Unfractionated Heparin Dose on Complement Activation and Selected Extracellular Vesicle Populations during Extracorporeal Membrane Oxygenation

Abstract

Extracorporeal membrane oxygenation (ECMO) provides critical support for patients with severe cardiopulmonary dysfunction. Unfractionated heparin (UFH) is used for anticoagulation to maintain circuit patency and avoid thrombotic complications, but it increases the risk of bleeding. Extracellular vesicles (EVs), nano-sized subcellular spheres with potential pro-coagulant properties, are released during cellular stress and may serve as potential targets for monitoring anticoagulation, particularly in thromboinflammation. We investigated the impact of UFH dose during ECMO therapy at the coagulation-inflammation interface level, focusing on complement activation and changes in circulating large EV (lEV) subsets. In a post hoc analysis of a multicenter randomized controlled trial comparing two anticoagulation management algorithms, we examined lEV levels and complement activation in 23 veno-venous-ECMO patients stratified by UFH dose. Blood samples were collected at different time points and grouped into three phases of ECMO therapy: initiation (day 1), mid (days 3-4), and late (days 6-7). Immunoassays detected complement activation, and flow cytometry analyzed lEV populations with an emphasis on mitochondria-carrying subsets. Patients receiving <15 IU/kg/h UFH exhibited higher levels of the complement activation product C5a and soluble terminal complement complex (sC5b-9). Lower UFH doses were linked to increased endothelial-derived lEVs, while higher doses were associated with elevated RBC-derived and mitochondria-positive lEVs. Our findings suggest the potential theranostic relevance of EV detection at the coagulation-inflammation interface. Further research is needed to standardize EV detection methods and validate these findings in larger ECMO patient cohorts.

Keywords: ECMO; anticoagulation; biomarker; bleeding; complement; critical illness; extracellular vesicles; heparin; mitochondria; thrombosis.

Figure 1

Unfractionated heparin (UFH) dose and sample size across groups at different timepoints of ECMO therapy. Sample size (N) at initiation (low: 8; high: 7), mid (low: 5; high: 4) and late (low: 10; high: 9). Data represent single measurements of each patient per day grouped by ECMO phase and UFH dose. One-way analyses of variance or Kruskal–Wallis tests were performed to compare doses across phases. Data are shown as boxes and whiskers representing the median and interquartile range. If not indicated otherwise by lines and asterisks, differences in groupwise comparisons are non-significant. * = p < 0.05; *** p < 0.001.

Figure 2

Complement activation throughout ECMO treatment and in comparison to UFH dose regimens. Activated complement pathway components C3, C5 and sC5b-9 at absolute concentrations in plasma of ECMO patients at different timepoints of ECMO therapy (A–C). One-way analyses of variance or Kruskal–Wallis tests were performed to compare phases and doses. Data are shown as columns and whiskers representing mean and SD. Comparisons of UFH doses in all patients and at all time points undergoing ECMO (D–F). Data are plotted as columns and whiskers representing mean and SD. A Student t-test or Mann–Whitney test was performed to compare groups. Sample size (N) in ECMO phases: initiation (low: 8; high: 7), mid (low: 5; high: 4), and late (low: 10; high: 9). Sample size (N) in summary measures: UFH dose low: 23; UFH dose high: 20. If not otherwise indicated by lines and asterisks, differences in groupwise comparisons were non-significant. * = p < 0.05; ** = p < 0.01.

Figure 3

Annexin V+ events throughout the ECMO period and in comparison to UFH dose regimens, as analyzed by flow cytometry. Absolute and relative levels of Annexin V+ events in plasma of ECMO patients at different stages of ECMO therapy (A,B). One-way analyses of variance or Kruskal–Wallis tests were performed to compare phases and doses. Data are shown as columns and whiskers representing mean and SD. Effect of UFH dose on absolute and relative counts of Annexin V+ events in all patients and at all time points undergoing ECMO (C,D). Data are plotted as columns and whiskers representing mean and SD. A Student t-test or Mann–Whitney test was performed to compare groups. Sample size (N) in ECMO phases: initiation (low: 8; high: 7), mid (low: 5; high: 4), and late (low: 10; high: 9). Sample size (N) in summary measures: UFH dose low: 23; UFH dose high: 20. If not indicated otherwise by lines and asterisks, differences in groupwise comparisons were non-significant.

Figure 4

Absolute (events/μL, upper row) and relative counts (% of Annexin V positive events, lower row) of lEV populations of different cellular origins throughout different phases of ECMO therapy (A–E). One-way analyses of variance or Kruskal–Wallis tests were performed to compare phases and doses. Data are shown as columns and whiskers representing mean and SD. Sample size (N) in ECMO phases: initiation (low: 8; high: 7), mid (low: 5; high: 4), and late (low: 10; high: 9). If not indicated otherwise by lines and asterisks, differences in groupwise comparisons were non-significant. RBCs (red blood cells); TF (tissue factor). ** = p < 0.01.

Figure 5

Absolute (events/μL, upper row) and relative counts (% of Annexin V positive events, lower row) of lEV populations of different cellular origins depending on anticoagulation regimens throughout ECMO therapy (A–E). Student t-tests or Mann–Whitney tests were performed to compare doses. Data are shown as columns and whiskers representing mean and SD. p-values below 0.1 are given to indicate trends. Sample size (N) in summary measures: UFH dose low: 23; UFH dose high: 20. If not indicated otherwise by lines and asterisks, differences in groupwise comparisons were non-significant.

Figure 6

Absolute (events/μL) and relative count (% of Annexin V-positive events) of Mitochondria+ lEV populations throughout ECMO phases and in comparison to UFH dose regimens. Annexin V+- and MitoTracker DeepRed double-positive events in plasma of ECMO patients at different stages of ECMO therapy (A,B) and grouped by UFH dose (E,F). Annexin V+-, MitoTracker- DeepRed- and platelet CD42b-positive events in plasma of ECMO patients at different stages of ECMO therapy (C,D) and grouped by UFH dose (G,H). One-way analyses of variance or Kruskal–Wallis tests were performed to compare selected phases and doses. Data are shown as columns and whiskers representing mean and SD. A Student t-test or Mann–Whitney test was performed to compare groups based on UFH dose. Sample size (N) in ECMO phases: initiation (low: 8; high: 7), mid (low: 5; high: 4), and late (low: 10; high: 9). Sample size (N) in summary measures: UFH dose low: 23; UFH dose high: 20. If not indicated otherwise by lines and asterisks, differences in groupwise comparisons were non-significant. * = p < 0.05.