Characterization of Trauma-Induced Coagulopathy Functional Profile in a Microfluidic Device

Introduction:: Trauma is the leading cause of death in young individuals. An estimated 25% of trauma patients present with trauma-induced-coagulopathy (TIC), which increases mortality 400%. The underlying mechanisms of TIC are not fully understood. Moreover, the ability to reliably assess coagulopathy in blood from trauma patients is limited as clinical assays have many limitations: they lack conditions of flow and biological surfaces, are typically over- or under-potentiated by agonists, and can have long turnaround times. However, many mechanistic components of hemostatic function are flow-dependent, including key aspects of platelet function. We have developed a microfluidic model that recapitulates injury to human vasculature under both venous and arterial flow settings (Fig. 1A). We used this model to characterize flow-dependent hemostatic function of trauma patient blood samples to better understand platelet dysfunction in the setting of TIC.


Materials and Methods:: Samples were collected from Level I trauma patients (N=19) within 60 minutes of arrival and perfused at venous (N=19; shear 150/s) and arterial (N=16; 3500/s) flow through a microfluidic device which comprised an injury site (IS) coated with collagen and tissue factor to simulate the subendothelial matrix exposed during vessel injury. Time to seal the IS was recorded (bleeding time, BT). Healthy volunteers were used as controls. Brightfield microscopy images of the IS were collected to analyze the resulting clot morphology by determining end mean intensity, in addition to measuring change in total clot surface area (SA) at the injury site over 1200 seconds (s). A subset of samples (trauma N=6, healthy N = 4) were incubated with fluorescent antibodies specific for CD41 prior to flow to visualize and quantify platelet contributions to clot formation via mean fluorescence intensity (MFI) data. Clinical data including complete blood counts, blood gas parameters, and injury information were collected. Prothrombin time (PT), activated partial thromboplastin time (aPTT), fibrinogen concentration (FGN), von Willebrand Factor (VWF) antigen levels (VWF:Ag), and D-dimer were also measured in platelet poor plasma derived from patient samples. Results are reported as mean ± standard error of the mean (SEM) and compared by t-tests. Correlation analyses (Pearson’s R) were performed on patient data and functional microfluidic data to identify patterns to generate hypotheses which were further explored with linear regression.


Results, Conclusions, and Discussions:: Trauma patients were 78% male, 61% blunt injury, age 42±2y, and had an Injury Severity Score (ISS) of 18±4. Trauma patients had longer BTs than controls for both venous (776±93s vs 317±48s; p=0.010) and arterial (890±92s vs 327±47s; p< 0.001) flow regimens. Trauma patients had slower deposition rates (Fig. 1B) but similar end clot SA for venous (2,197±131µm² vs 2,495±2µm²; not significant (ns)) and arterial (1,821±249µm² vs 2,346±97.77 µm²; ns) flow. Max fold change (FC) of trauma sample platelet MFI was 1.608±0.643 vs healthy (3.848±0.760; p=0.056) in venous flow conditions, and 1.095±0.218 (trauma) vs 4.013±0.846 (healthy, p< 0.01) in arterial shear (Fig. 1C). Despite having similar PT values (13.6±0.3s vs 13.3±0.2s), trauma samples had lower FGN (290.2±27.5mg/dl vs 404.7±21.5mg/dl; p=0.012) and aPTT (27.9±0.8s vs 37.6±1.8s; p< 0.001) than healthy samples; with higher VWF:Ag (317.3± 28.2% vs 111.7± 17.6%; p< 0.001) and D-dimer (0.258±0.015µg/mL vs. 6.566±1.761 µg/mL; p=0.029) when compared to healthy controls. Correlation analysis revealed trends of reduced protein content associated with longer venous BT (Fig. 1D-G). Interestingly, reduced red blood cell count correlated with higher arterial clot growth rate and reduced blood oxygen saturation similarly correlated with a faster arterial BT.


We observed increased bleeding times in samples collected from trauma patients at both venous and arterial flow settings, clearly demonstrating the impaired hemostatic function in trauma samples. SA analysis of the clot at the IS showed a slower initial rate of formation of clot at the IS yet resulted in similar end SA. However, these clots had drastically reduced platelet content, identifying impaired flow-based platelet function induced by trauma. Correlation analysis further confirmed that reduced VWF levels, important for platelet interaction with injury, are associated with slower venous BT. Furthermore, lower D-Dimer in trauma samples was also associated with increased BT, suggesting perhaps that patients with a high degree of in vivo hemostatic activity are more likely to have improved ex vivo clotting.


While still depositing substantial amounts of clot under flow, ex vivo microfluidic analysis from trauma samples demonstrated reduced hemostatic capacity, underpinned by aberrant platelet function, plasma protein levels, and oxygenation-related parameters.




Acknowledgements (Optional): :

References (Optional): :