Abstract
Objective:
We sought to identify potential drivers behind resuscitative endovascular balloon occlusion of the aorta (REBOA) induced reperfusion coagulopathy using novel proteomic methods.
Background:
Coagulopathy associated with REBOA is poorly defined. REBOA Zone 1 provokes hepatic and intestinal ischemia that may alter coagulation factor production and lead to molecular pathway alterations that compromises hemostasis. We hypothesized that REBOA Zone 1 would lead to reperfusion coagulopathy driven by mediators of fibrinolysis, loss of coagulation factors, and potential endothelial dysfunction.
Methods:
Yorkshire swine were subjected to a polytrauma injury (blast traumatic brain injury, tissue injury and hemorrhagic shock). Pigs were randomized to observation only (controls, n=6), or to 30 minutes of REBOA Zone 1 (n=6) or REBOA Zone 3 (n=4) as part of their resuscitation. Thromboelastography was used to detect coagulopathy. Enzyme-Linked Immunosorbent Assays (ELISA) assays and mass spectrometry proteomics were used to measure plasma protein levels related to coagulation and systemic inflammation.
Results:
After the polytrauma phase, balloon deflation of REBOA Zone 1 was associated with significant hyperfibrinolysis (TEG results: REBOA Zone 1 35.50% vs Control 9.5% vs Zone 3 2.4%, p<0.05). In the proteomics and ELISA results REBOA Zone 1 was associated with significant decreases in coagulation factor XI and coagulation factor II, and significant elevations of active tPA, PAP complexes, and syndecan-1 (p<0.05).
Conclusion:
REBOA Zone 1 alters circulating mediators of clot formation, clot lysis, and increases plasma levels of known markers of endotheliopathy, leading to an reperfusion induced coagulopathy compared to REBOA Zone 3 and no REBOA.
Keywords: Combat casualty care, blast traumatic brain injury, hemorrhagic shock, tissue injury, trauma resuscitation, resuscitative endovascular balloon occlusion of the aorta, trauma induced coagulopathy, hyperfibrinolysis
Introduction
Resuscitative endovascular balloon occlusion of the aorta (REBOA) is an adjunct to resuscitate critically injured patients with noncompressible truncal hemorrhage. The balloon can be inflated in Zone 1 which is within the thoracic aorta between the left subclavian artery and the celiac trunk, or in Zone 3, which is in the abdominal aorta distal to the renal arteries. Zone 1 inflation is suggested for the management of intraabdominal and retroperitoneal hemorrhage and Zone 3 inflation for the treatment of pelvic and lower extremity hemorrhage[1]. Recently the military has recommended Zone 1 REBOA for resuscitation of combat casualties (2,3)
While the visceral ischemic consequences of Zone 1 aortic occlusion are well recognized and have prompted the development of new partial REBOA devices, the potential impact on coagulation has been largely unexplored [1]. Trauma induced coagulopathy is driven dominantly by hypoperfusion which is further exaggerated by tissue injury and traumatic brain injury [4]. Thus, while Zone 1 REBOA can reduce subdiaphragmatic hemorrhage it could paradoxically result in greater blood loss through reperfusion induced coagulopathy, which we reported recently [5]. Morrison et al observed coagulopathy in a swine model of severe non-compressible hemorrhage resuscitated with Zone 1 REBOA [6].
Using our swine polytrauma model we sought to evaluate REBOA’s potential to promote coagulopathy due to visceral ischemia in a combat scenario. We hypothesized that Zone 1 REBOA would be associated with more severe coagulopathy compared to inflation in Zone 3. We then sought to identify the potential mediators driving REBOA-induced coagulation using mass spectrometry proteomics and targeted ELISAs.
Swine Model
This experimental animal model was approved by the Institutional Animal Care and Use Committee under protocol #1050. All animal care occurred in a fully accredited Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC) facility and all research procedures were conducted in compliance with the Animal Welfare Act and the principles of Guide for the Care and Use of Laboratory Animals, National Research Council. Male Yorkshire swine weighing between 45–58 kg were used for all experiments. Animal preparation was performed according to previously published protocols [3]. Briefly, on the day of the experiments, animals were anesthetized using a combination of propofol and fentanyl. The right axillary artery as well as the right and left femoral arteries and veins were cannulated for invasive blood pressure monitoring, blood removal, and ER-REBOA catheter (Prytime Medical, Boerne, TX) placement. The injury series was comprised of blast traumatic brain injury (bTBI), bilateral open femur fractures, and hemorrhagic shock to a targeted base excess of −10 mmol/L.
The bTBI was performed inside a Mobile Shock Tube laboratory (Applied Research Associates, Littleton, Co) where a 55psi shock wave was directed at the swine’s head while wearing a Kevlar vest to help protect their chest and abdomen. Bilateral femur fractures were created through muscle cutdowns and a captive bolt stunner (Blitz-Kerner, Turbocut JOBB GmbH, Germany) was fired directly on the exposed femur. The fracture was confirmed by digital and visual inspection and packed with gauze to minimize bleeding. The HS phase was produced through rapid controlled bleeding from each femoral artery until the swine reached a mean arterial pressure (MAP) of 15 mmHg and an EtCO2 reading of 20 mmHg. During the blood removal period, swine were randomized to either a REBOA intervention group (Zone 1 or Zone 3) or to a non-REBOA control group. Swine randomized to the REBOA groups underwent complete balloon occlusion in either Zone 1 or Zone 3, or no-occlusion, for the last 30 minutes during the HS phase, and resuscitated to a MAP of 35 mm HG with 5% human albumin (1000mL, Grifols Biologicals Inc., Los Angeles, CA). After the 30 minutes of balloon occlusion (or no-occlusion for the control group), animals were resuscitated with a combination of albumin and shed blood.
The animals were observed for 240 minutes following. Vital signs were monitored continuously and serial blood draws were performed (baseline, end of shock, 30-, 60-, 120-, 180-, and 240-minutes after the end of shock). Whole blood was used for thrombelastography (TEG). Citrate native (CN) and tissue plasminogen activator (tPA) challenge TEGs were performed at each timepoint and the R-time, maximum amplitude (MA), angle, and percent lysis at 30 minutes (LY30) were recorded. The tPA TEG was used because of the reported relative resistance of swine to fibrinolysis (7). Platelet free plasma was centrifuged from citrated whole blood samples and flash frozen in liquid nitrogen for ELISA assays and mass spectrometry (MS)-based proteomics analysis.
MS-Based Proteomics
The proteomics workflow was conducted according to our previously described protocol[8]. Briefly, the stored plasma samples were digested with trypsin after reduction and alkylation. Resulting peptide samples underwent liquid chromatography separation coupled with tandem MS. Raw DIA files were processed using Spectronaut software (Biognosys, Schlieren, Switzerland) and the Uniprot database was queried for sus scrofa specific peptide identification. Normalization and global analyses of proteomics data was performed using MetaboAnalyst software. Further, targeted searches were performed to assess significant differential abundances of coagulation factors and anti-coagulant inhibitors, complement proteins, known regulators of fibrinolysis, and known markers of endothelial activation. Identified proteins were analyzed using percent change from baseline. Proteins below the MS detection limit at baseline were categorized using 3 cutoffs based on previously published fold changes [8]: below detection, <20-fold above detection level, and >20-fold above detection level. Proteins of interest and analysis methods can be viewed in Supplemental Digital Content (SDC) 1.
ELISA assays
Due to the known limitations in identification of certain proteins by MS, the following proteins were analyzed using Enzyme Linked Immunosorbent Assays (ELISA) at select timepoints: tissue plasminogen activator (tPA), plasminogen activator inhibitor (PAI-1), and tPA-PAI-1 complex (Innovative Research, Inc, Novi, MI); plasminogen, plasmin-antiplasmin complex (PAP), antithrombin (ATIII), thrombin, and thrombin-antithrombin complex (TAT) (MyBiosource, San Diego, CA, ng/mL). Proteins detected by ELISA were analyzed using absolute concentration values.
Statistical Analysis
Data are presented as means and standard deviations (SD) for normally distributed variables and as medians and interquartile ranges if skewed. Baseline changes and between groups comparisons for TEG parameters, ELISAs, and MS-identified proteins were performed at each timepoint using linear mixed models for repeated measures, except for those MS-identified proteins that had relative intensities below detection levels at baseline. Changes in these proteins were analyzed using the designated fold-change cutoffs with χ2 test or Fisher’s Exact Test. Overall significance was declared at p<0.05 with multiple comparisons adjusted by false discovery rate. For the tPA-challenge TEG LY30 results, a Box-Cox power transformation was performed prior to the linear mixed model analysis. Group sample sizes were previously determined and reported based on expected coagulopathy variables as previously described [3].
Results
Six swine were randomized to the No-REBOA group, six to Zone 1 REBOA, and five to Zone 3 REBOA. One swine in the Zone 3 group expired during instrumentation due to cardiac dysfunction and was excluded from analysis. The percentage of total blood volume removed (49 ± 9.4%) to achieve the target MAPs and EtCO2 values was similar between groups (p>0.5), and all animals spent a similar amount of time in HS (50min [IQR: 45, 60; p>0.5]). Base excess at the start of resuscitation after REBOA was significantly lower in the Zone 1 REBOA group compared to the No-REBOA and the Zone 3 groups, respectively, at 60 minutes (−13.3±4.4 vs −5.3±5.7 vs −5.0±1.4 mEq/L, p<0.01), 120 minutes (−9.0±6.0 vs 0.5±5.8 vs 1.0±2.9 mEq/L, p<0.005), and 180 minutes (−4.6±5.4 vs 5.2±4.4 vs 6.0±3.7 mEq/L, p<0.005) into resuscitation.
Changes in TEG outcomes (R-time, MA, angle, LY30) are shown in Table 1. Coagulation abnormalities were prominent in the Zone 1 REBOA group and were most pronounced at 60 minutes after the end of HS. The coagulopathy identified in the Zone 1 REBOA group was characterized by prolonged R-time, depressed angle and MA, and an increased LY30 compared to baseline and the other groups. Mass spectrometric analysis identified a total of 854 plasma proteins. Global analysis of the proteomics data using partial least squares-discriminant analysis (PLS-DA) identified distinct separation as a function of swine group. The Zone 1 REBOA group drove the highest separation as measured by component 2 of the PLS-DA model which explained 15.4% of variance within the data (Figure 1A). Further elaboration of the proteomics data was performed using ANOVA to determine significant differential abundances between swine groups. The top 50 significantly different proteins between swine group as measured by ANOVA were visualized using a heat map (Figure 1B). The most notable differential abundances were identified in the Zone 1 REBOA group. On a proteome wide level, the most significant changes in plasma abundances occurred in the Zone I REBOA group. To determine the biological role of the significantly differential proteins, analyses were performed on select proteins as described below.
Table 1.
Coagulation Results from Thromboelastography (TEG). Values shown are mean (SD) other median [IQR]. Citrate Native TEG is used to quantify all values except for LY30, which is measured using the tPA-challenge TEG.
| Value | Time | No-AO | Zone 1 REBOA | Zone 3 REBOA |
|---|---|---|---|---|
| R-Time (sec) | Baseline | 7.30 (0.86) | 6.23 (2.22) | 6.90 (5.80) |
| End Of Shock (0min) | 4.70 (2.37) | 5.51 (1.27) | 5.55 (1.95 | |
| 30 min | 4.47 (2.06) | 9.70 (6.92) x | 3.05 (0.98) | |
| 60 min | 5.88 (1.95) | 11.77 (4.69) * x | 4.03 (1.21) | |
| 240min | 5.18 (2.08) | 7.18 (2.10) | 6.40 (0.56) | |
| Angle (degrees) | Baseline | 68.68 (2.85) | 72.72 (6.40) | 63.33 (16.89) |
| End Of Shock (0min) | 71.32 (8.46) | 72.43 (3.25) | 70.18 (3.97) | |
| 30 min | 71.26 (4.90) | 59.26 (19.80) * x | 73.92 (4.47) | |
| 60 min | 70.15 (4.38) | 55.15 (11.20) * x | 74.90 (3.28) | |
| 240min | 71.93 (4.82) | 62.7 (6.71) | 67.97 (1.70) | |
| Maximum Amplitude (mm) | Baseline | 76.58 (1.74) | 78.00 (4.79) | 75.13 (5.25) |
| End Of Shock (0min) | 71.80 (4.10) | 74.00 (5.34) | 71.00 (3.00) | |
| 30 min | 68.83 (5.30) * | 62.90 (9.51) * | 68.00 (7.35) * | |
| 60 min | 66.41 (4.38) * | 59.23 (9.64) * x | 70.10 (3.72) | |
| 240min | 67.25 (4.10) * | 62.80 (4.75) * | 64.13 (11.85) * | |
| LY30 (%) | Baseline | 0.95 [0.6, 2.1] | 2.45 [1.8, 3.5] | 5.15 [2.2, 9.0] |
| End Of Shock (0min) | 3.6 [1.3, 14.0] | 2.4 [1.4, 4.3] | 2.1 [1.5, 2.9] | |
| 30 min | 9.55 [0.5, 30.6] | 21.1 [2.7, 31.9] * | 2.7 [0.8, 4.2] | |
| 60 min | 9.56 [0.5, 30.6] | 35.50 [21.2,58.2] * x | 2.50 [0.90,3.9] | |
| 240min | 1.70 [1.6, 5.9] | 1.30 [1.1, 1.5] | 1.70 [0.6, 2.5] |
p<0.05 compared to group’s baseline
p<0.05 compared to other 2 models
Figure 1. Global Proteomic Changes Depended Upon Intervention.
Normalization and statistical analyses were performed using MetaboAnalyst software.
A) Partial least squares – discriminant analysis (PLS-DA) of swine proteomes across the monitored time course showed distinct clustering dependent on the intervention. The highest separation was observed on component 2 (15.4%) which was driven by the Zone 1 REBOA intervention.
B) Heat map of the top 50 proteins that were significantly different between interventions as measured by ANOVA. Protein intensities were normalized and exhibited on a scale from −4 (blue) to 4 (red). Time points were ordered from left to right per swine for each intervention. Proteins were hierarchically clustered to show groups with similar intensities.
Coagulation and Complement Activation
Significant changes in coagulation factors II, V, XI, XIII chain B, fibrinogen chains, and complement C3 and complement factor H are shown in Table 2. There were no significant changes in the intensities of MS-identified coagulation factors VII, IX, X, XII, and XIIII chain A, or in complement proteins C5 and complement factor I after adjusting p-values using FDR. Trends in these coagulation factors are shown in SDC2. Plasma relative intensities of Factors II and XI declined significantly in the Zone 1 group only, while the relative intensity of factor V declined significantly in the No-REBOA group only. Proteomic levels of Factor II reflect both prothrombin and thrombin levels and are shown in Figure 2. Factor XIII chain B, and fibrinogen chains α, β, and γ relative intensities declined significantly in all groups throughout the experiment. The relative intensities of complement protein C3 and complement Factor H declined significantly in the Zone I REBOA group only. Thrombin plasma concentrations had a significant decline in the Zone 1 REBOA group only (thrombin EILSA concentrations are shown in Figure 2).
Table 2.
Markers of clot formation and complement activation from mass spectrometry results. All values are mean with SD.
| Protein (relative % change from baseline) | Time (from end of shock) | No-AO | Zone 1 REBOA | Zone 3 REBOA |
|---|---|---|---|---|
| FII | 30min | 2.18 (7.17) | −13.47 (6.20) * | −4.42 (10.06) |
| 60min | −3.63 (8.73) | −15.54 (7.64) * | −3.23 (6.48) | |
| 120min | −10.61 (5.81) | −15.09 (8.05) * x | 2.76 (9.25) | |
| 240min | −9.16 (7.17) | −17.94 (4.26) * | −2.50 (4.29) | |
|
| ||||
| FV | 30min | −2.64 (20.83) | −16.75 (13.85) | −11.45 (11.06) |
| 60min | −8.68 (15.08) | −18.19 (15.44) | −12.73 (8.86) | |
| 120min | −18.85 (1.44) | −5.38 (19.79) | −5.82 (14.80) | |
| 240min | −21.35 (10.00) * | −17.99 (6.03) | −15.43 (8.99) | |
|
| ||||
| FXI | 30min | 10.14 (22.00) | −16.91 (4.64) | −14.78 (8.72) |
| 60min | −1.32 (13.18) | −16.97 (9.05) * x | −6.82 (11.08) | |
| 120min | −8.69 (7.32) | −14.82 (10.34) * x | −8.37 (7.79) | |
| 240min | 1.34 (8.50) | −13.85 (9.86) | −12.46 (9.67) | |
|
| ||||
| FXIII B Chain | 30min | −24.30 (10.82) * | −35.66 (13.66) * | −23.58 (11.25) * |
| 60min | −28.66(11.42) * | −30.88 (11.57) * | −18.84 (4.15) * | |
| 120min | −30.30 (5.94) * | −23.92 (15.28) * | −20.07 (5.03) * | |
| 240min | −30.81(6.43) * | −28.07 (8.03) * | −23.03 (873) * | |
|
| ||||
| Fibrinogen α Chain | 30min | −26.57 (32.18) | −27.13 (25.12) | −8.64 (35.02) |
| 60min | −14.88 (31.91) | −30.72 (33.61) | −24.85 (37.98) | |
| 120min | −25.88 (33.88) | −28.33 (24.84) | −44.64 (26.11) * | |
| 240min | −16.78 (30.43) | −31.73 (35.28) * | −20.72 (53.85) | |
|
| ||||
| Fibrinogen β Chain | 30min | −25.12 (36.08) | −39.71 (27.23) * | −26.43 (9.28) |
| 60min | −24.16 (32.10) | −43.56 (13.61) * | −32.75 (21.35) | |
| 120min | −33.36 (24.65)* | −28.39 (37.15) * | −37.54 (21.35) * | |
| 240min | −23.66 (11.06) | −35.98 (36.52) * | −16.11 (26.52) | |
|
| ||||
| Fibrinogen γ Chain | 30min | −28.68 (24.10) * | −33.21 (20.80) * | −23.31 (13.28) |
| 60min | −23.62 (20.14) * | −29.87(23.89) * | −24.43 (30.07) | |
| 120min | −27.51 (28.23) * | −26.83 (24.46) * | −35.13 (26.49) * | |
| 240min | −26.31 (17.27) * | −32.50 (33.76) * | −22.52 (32.60) | |
|
| ||||
| Complement C3 | 30min | 2.10 (14.48) | −8.52 (5.42) | −0.45 (4.13) |
| 60min | 0.32 (12.09) | −13.96 (6.40) * | −4.57 (4.44) | |
| 120min | −7.58 (4.12) | −7.82 (20.59) | 2.36 (14.73) | |
| 240min | −11.89 (4.22) | −17.32 (6.07) * | −2.85 (9.33) | |
|
| ||||
| Complement Factor H | 30min | −2.06 (15.68) | −8.32 (9.36) | 1.83 (11.53) |
| 60min | −1.39 (14.79) | −11.03 (8.14) | 0.55 (1.42) | |
| 120min | −5.29 (8.48) | −14.73 (12.71) * | −5.05 (17.94) | |
| 240min | −6.96 (6.49) | −14.50 (13.00) * | −6.71 (19.46) | |
p<0.05 compared to group’s baseline
p<0.05 compared to other 2 models
Figure 2. Changes in Thrombin (ELISA) and FII (Mass Spectrometry).
Mean Thrombin concentrations determined by ELISA plotted with standard deviation. Proteomic levels of Factor II (reflecting both prothrombin and thrombin) are graphed as mean change from baseline plotted with standard deviation. Plotted p-values represent significant differences of Zone 1 REBOA group at that time compared to baseline.
Hypocoagulation
Plasma concentrations of ATIII were significantly higher in the Zone 1 REBOA group at 60min post-shock (22.22 ± 4.66, p=0.049) and 240min post-shock (21.16 ± 4.68, p=0.049) compared to baseline (16 ± 4.68 ng/mL). There were no changes in relative intensity levels of anticoagulant proteins protein C, protein S, and protein C endothelial receptor throughout the experiment in any group.
Hyperfibrinolysis
Percent changes of the relative intensity of MS-identified proteins involved in fibrinolysis are shown in Table 3. Plasminogen, TAFI and α−2-antiplasmin decreased significantly from baseline in the Zone 1 REBOA group only. Five pigs (83%) in the Zone 1 REBOA group had a >20-fold or more increase in the MS relative intensity of tPA, while only two pigs (33.3%) in the No-REBOA group and 2 pigs (50%) in the Zone 3 group had this increase (p=.047). Four pigs (66.7%) in the Zone 1 group experienced a >20-fold increase in PAI-1 (50%), while two pigs (40%) in the Zone 3 group and three pigs (50%) in the No-REBOA group had this increase (p=0.16).
Table 3.
Relative percent change in plasma levels of proteins regulating fibrinolysis from mass spectrometry results. tPA and PAI-1 values were compared among groups using cutoffs thresholds due to being below detection level at baseline and other timepoints (represented as 0.00). Values presented as mean (SD) except for tPA and PAI-1 which are shown as medians [IQR] when above detection level.
| Protein (relative % change from baseline) | Time (from end of shock) | No-AO | Zone 1 REBOA | Zone 3 REBOA |
|---|---|---|---|---|
| Plasminogen | 30min | −9.46 (17.05) | −14.90 (5.58) | −6.30 (12.38) |
| 60min | −10.10 (14.34) | −15.32 (7.41) | −3.03 (4.69) | |
| 120min | −10.41 (9.61) | −27.40 (19.34) * | −17.22 (32.08) | |
| 240min | −14.25 (9.65) | −24.24 (7.50) * | −20.49 (29.16) | |
|
| ||||
| α−2-antiplasmin | 30min | 5.17 (14.69) | −11.92 (7.20) | 0.60 (6.81) |
| 60min | 1.43 (15.33) | −16.33 (9.34) * | 1.67 (5.39) | |
| 120min | −4.00 (4.96) | −9.56 (17.42) | 11.67 (18.90) | |
| 240min | −3.76 (8.64) | −15.50 (6.84) | 5.96 (16.32) | |
|
| ||||
| TAFI | 30min | −2.61 (11.32) | −27.87 (8.85) * | −16.45 (9.84) |
| 60min | −4.35 (17.44) | −31.08 (6.69) * | −12.61 (11.59) | |
| 120min | −10.19 (9.35) | −20.09 (22.35) * | 1.47 (10.44) | |
| 240min | −4.64 (8.61) | −26.66 (15.90) * | −8.36 (1.62) | |
|
| ||||
| PAI-1 | 30min | 0.00 | 0.00 | 0.00 |
| 60min | 0.00 | 0.00 | 0.00 | |
| 120min | 0.00 | 0.00 | 0.00 [0–1868] | |
| 240min | 2224 [0–2583] | 1721 [1523– 2389] | 2388 [1855–2639] | |
|
| ||||
| tPA | 30min | 0.00 | 3640 [0– 6041] | 0.00 |
| 60min | 0.00 | 3318 [0– 5382] | 0.00 [0–2580] | |
| 120min | 0.00 | 0.00 | 0.00 | |
| 240min | 0.00 | 0.00 | 0.00 | |
p<0.05 compared to group’s baseline
p<0.05 compared to other 2 models; Plasminogen, α−2-antplasmin, and TAFI are shown as mean (SD); tPA and PAI-1 are shown as median and IQR
tPA= tissue plasminogen activator; PAI-1= plasminogen activator inhibitor-1; TAFI=thrombin activatable fibrinolysis inhibitor; SD= standard deviation; IQR= interquartile range
Changes in the concentrations of tPA, PAP, tPA-PAI-1 complex, and plasminogen measured by ELISA are shown in Figure 3. There were significant increases in tPA and PAP plasma concentrations in the Zone 1 REBOA group only. All groups had a significant increase in PAI-1 concentrations at 240min post-shock (Zone 1: 63.04 ± 36.60 vs 2.11 ± 0.70 ng/mL, p= 0.0003; Zone 3: 46.59 ± 42.82 vs 1.80 ± 0.76 ng/mL, p=0.019; No-REBOA: 86.99 ± 43.78 vs 3.88 ± 0.93 ng/mL, p=0.0003); Compared to plasma baseline concentrations, the tPA-PAI-1 complex increased significantly in Zone 1 at 240 minutes (136.38 ± 49.92 vs 10.16 ± 4.31 ng/mL. p=0.003) and in Zone 3 REBOA at 60 minutes post-shock (109.84 ± 43.03 vs 8.77 ± 43.98 ng/mL. p=0.041).
Figure 3. Changes in the concentrations of tPA, PAP, tPA-PAI complex, and plasminogen from ELISA results.
Mean ELISA concentrations are plotted with standard deviations.
Endotheliopathy
Significant changes in the relative intensities of markers of endotheliopathy are shown in Table 4. All groups had significant increases in the relative intensity of plasma vWF. The No-REBOA group and the Zone 1 REBOA group had significant increases in L-selectin compared to baseline and the Zone 3 group. The relative intensity of syndecan-1 increased >20-fold in 5 pigs (83%) in the Zone 1 REBOA group, while no pigs in the No-REBOA group or Zone 3 REBOA group manifested this increase (p=0.001).
Table 4.
Markers of endothelial activation and basement/interstitial membrane disruption. Values presented as mean (SD), except for SDC-1 which is represented as median [IQR]. SDC-1 values were compared among groups using cutoffs thresholds due to being below detection level at baseline and other timepoints (represented as 0.00).
| Protein (relative % change from baseline) | Time (from end of shock) | No-AO | Zone 1 REBOA | Zone 3 REBOA |
|---|---|---|---|---|
| SDC-1 | 30min | 0.00 | 3354 [2152– 4969] | 0.00 |
| 60min | 0.00 | 3655 [1381– 6143] | 0.00 [0– 697] | |
| 120min | 0.00 | 1605 [0– 2899] | 0.00 | |
| 240min | 0.00 | 1039 [0– 2366] | 0.00 [0– 1039] | |
|
| ||||
| vWF | 30min | 63.52 (25.13) * | 100.41 (55.68) * | 88.36 (48.41) * |
| 60min | 82.25 (35.75) * | 117.31 (42.05) * | 132.64 (52.15) * | |
| 120min | 101.18 (39.01) * | 151.99 (67.18) * | 90.09 (31.82) * | |
| 240min | 114.14 (40.31) * | 173.84 (100.48) * | 116.21 (24.93) * | |
|
| ||||
| L-Selectin | 30min | 44.76 (26.88) * | 53.90 (36.77) * | 28.60 (17.17) |
| 60min | 56.01 (32.98) * | 56.50 (32.63) * | 7.03 (14.67)X | |
| 120min | 46.98 (19.39) * | 41.39 (30.83) * | -15.78 (10.15)X | |
| 240min | 52.72 (31.18) * | 44.15 (33.81) * | 7.10 (22.44) | |
|
| ||||
| Collagen type I α−1 chain | 30min | 149.88 (105.24) * | 275.83 (101.95) * X | 98.52 (71.47) |
| 60min | 58.78 (76.91) | 200.81 (110.24) * X | 40.88 (14.39) | |
| 120min | 24.72 (48.41) | 97.81 (52.61) * | 34.70 (16.02) | |
| 240min | 16.98 (40.63) | 59.64 (43.04) | 12.20 (25.32) | |
|
| ||||
| Collagen type I α−2 chain | 30min | 165.16 (140.34) * | 313.28 (115.40) * X | 122.15 (55.24) * |
| 60min | 81.74 (101.00) * | 193.01 (92.79) * | 60.09 (54.57) | |
| 120min | 46.32 (54.30) | 85.19 (66.34) | 68.07 (36.09) | |
| 240min | 35.80 (52.23) | 89.84 (52.23) | 35.46 (38.88) | |
|
| ||||
| Collagen type XVIII α−1 chain | 30min | 115.83 (27.20) * | 166.75 (42.37) * | 88.96 (44.90) * |
| 60min | 98.26 (57.10) * | 144.64 (47.90) * | 87.06 (14.65) * | |
| 120min | 78.02 (57.13) * | 143.58 (81.89) * | 55.80 (34.58) | |
| 240min | 55.83 (42.19) * | 143.53 (76.76) * | 57.61 (48.73) | |
p<0.05 compared to group’s baseline
p<0.05 compared to other 2 models
All groups had significant increases in the relative intensities of collagen type XVIII α−1 chain from the basement membrane of blood vessels. The relative intensity of collagen type XV α−1 chain increased >20-fold in 100% of the Zone 1 group, in three (50%) of the No-REBOA pigs, and in three (60%) of the Zone 3 pigs (p=0.06). The relative intensity of collagen type IV α−2 chain increased >20-fold in 100% of the Zone 1 group, in four (67%) of the No-REBOA pigs, and in three (40%) of the Zone 3 pigs (75%) (p= 0.11).
The relative intensity of collagen type I α−2 chain from the interstitial matrix of the vasculature increased in all groups. The No-REBOA and Zone 1 REBOA groups had significant increases in collagen type I α−1 chain. There were no significant changes in the relative intensities of collagens type III α−1 chain and collagen type VI α- 1, 2, and 3 chains.
Discussion
This study investigated the potential mechanisms driving Zone 1 REBOA induced coagulopathy using comprehensive proteomic analyses combined with select ELISAs to identify the proteomic changes that may be contributing to reperfusion coagulopathy in a military relevant DCBI swine model resuscitated with REBOA. Our data indicate 30 minutes of Zone 1 REBOA is associated with an exaggerated coagulopathy characterized by poor clot initiation, weak clot strength, and an increased rate of fibrinolysis. This hypocoagulopathic profile appears to be driven by multiple events including coagulation factor depletion, enhanced fibrinolysis, and endothelial activation..
Clotting factor depletion following trauma occurs immediately after injury and is associated with worse outcomes[9–11]. In this swine study the Zone 1 REBOA group was associated with further decreases in coagulation factors FII and FXI. Clinical studies have associated prolonged PT with a primary deficiency in Factor II, and prolonged aPTT is primarily explained by deficiencies in Factor XI (intrinsic pathway)[12]. The prolonged R-time detected by CN TEG in the Zone 1 REBOA group is partially due to the depletion of these coagulation factors. The depressed CN TEG MA seen in all groups in this study may be accentuated by a reduction in factor XIII a, as all groups has significant decline in XIII b chain which is the part of the circulating Factor heterodimer of a and b that binds to fibrinogen. Factor XIII is responsible for crosslinking fibrin and stabilizing the clot[13], and prior clinical studies have shown that patients with deficiencies in factor XIII a also have depressed MA and angle[14], as well as poor clot stability with increased rates of clot lysis[15].
Additionally, all groups had significant decline in the plasma fibrinogen chains (α, β, and γ), with significant decreases in the Zone 1 persisting throughout the resuscitation period while Zone 3 had only temporary reduction. Fibrinogen is synthesized by the liver[15] and has been shown to be the first coagulation factor to decrease following hemorrhagic shock [16]. Low fibrinogen levels at admission are associated with increased rates of transfusion requirements and risk of mortality [17–19]. The transient decline in fibrinogen chains followed by return to near baseline levels in the Zone 3 REBOA group compared to the prolonged depletion levels at 240 minutes in the Zone 1 group may be due to the benefits of preserved liver perfusion with Zone 3 balloon inflation.
Perhaps the most striking findings in this study is the overwhelming increase in the rate of clot lysis with concurrent increases in profibrinolytic drivers and decreases in antifibrinolytic inhibitors in the Zone 1 REBOA group. Elevated fibrinolysis is found in approximately 15% of severely injured patients, and is associated with increased rates of massive transfusion and mortality [20, 21]. While the exact mechanism driving pathological hyperfibrinolysis is unclear, increased clot lysis has been attributed to increased tPA release overwhelming available PAI-1 levels, leading to uninhibited plasminogen activation [9]. The results from our ELISAs and mass spectrometry findings are consistent with this mechanism. In the Zone 1 REBOA group, increased tPA levels were detected only at the same timepoints in which elevated fibrinolysis was detected by our tPA challenge TEG. Additionally, PAP complex levels, a known marker of plasmin generation, were increased only in the Zone 1 group at 30- and 60- minutes after balloon deflation. Declines in α−2-antiplasm and TAFI in the Zone 1 group were also seen at these same timepoints, further contributing to the hyperfibrinolysis [23]. PAI-1 levels increased in all groups near the end of the experiment, and this delayed surge in PAI-1 is consistent with previous animal and clinical studies[9]. This later spike in PAI-1, along with the short half-life of tPA as it is cleared by the liver, and continued resuscitation efforts following balloon deflation likely drive recovery from the temporary hyperfibrinolytic state in the Zone 1 animals. Clinical studies have shown that overcorrection from a hyperfibrinolytic state to a prolonged depressed fibrinolytic state lasting greater than 24 hours ;ie, fibrinolysis shutdown, is associated increased mortality[24,25].
The significant increases in plasma levels of syndecan-1, as well as increases in vWF and L-selectin in the Zone 1 REBOA group suggests greater endothelial disruption due to the additional ischemic burden from thoracic aortic occlusion. The higher levels following Zone 1 occlusion are presumably due to the greater sensitivity of visceral endothelium to hypoperfusion. The protective glycocalyx is composed of proteoglycans and glycoproteins such as syndecan-1, hyaluronic acid, heparin sulfate and chondroitin sulfate[26]. Increased plasma circulating levels of these glycocalyx components indicate breakdown of this glycocalyx leading to increased vascular permeability and release of pro-inflammatory cytokines[27] Syndecan-1 is one of the most well-known biomarkers for quantifying endothelial dysfunction, and plasma levels of syndecan-1 are higher in patients with more severe shock and lower base excess upon admission[27,28]. In addition to being associated with worse outcomes, patients with elevated syndecan-1 levels have increased rates of hypocoagulation and hyperfibrinolysis[29–31].
These novel findings that Zone 1 REBOA is associated with worsening coagulopathy and endotheliopathy should be further investigated with a focus on elucidating the underlying mechanisms. Whether this worsening coagulopathy profile is due to hypoperfusion of the liver, the intestine, or due to more global tissue bed ischemia remains unknown. Swine models of liver ischemia have shown that 60 minutes of ischemia is associated with a decline in clotting factor activity [32]. But this time limit may be shortened by associated shock and tissue injury. Patients who undergoing liver resection and transplantation have significant declines in coagulation factors I, II, V, VII, X, IX, XI, and XII, as well as decreases in circulating levels of protein C and ATIII[33]. However, fibrinogen levels have also been shown to decrease in animal models with isolated intestinal ischemia without liver injury[34]. Thus, the significant decline in coagulation factors II and XI and the associated increase in R-Time may be predominantly driven by liver hypoperfusion further worsened by the intestinal ischemia from Zone 1 REBOA.
The hyperfibrinolysis associated with Zone 1 may be aggravated by endothelial activation following ischemia leading to tPA release, and further accentuated by decreased tPA clearance from the liver. Polytrauma in rodent models, show that tPA release occurs immediately following insult, and PAI-1 release occurs several hours later (35,36). Our tPA and PAI-1 findings are consistent with these observations. Additionally, Zheng et al compared hepatocyte and endothelial derived tPA release in response to local carotid thrombosis and found tPA release was due to endothelial stores, while hepatocyte derived tPA appeared to be responsible for basal tPA levels [37]. Thus, if the hyperfibrinolysis seen following Zone 1 REBOA is driven by overwhelming tPA release, the source is likely diverse endothelial sources.
This study has several limitations. First, while mass spectrometry techniques are an increasingly important component of translation investigation in our trauma patients, this method does not measure protease activity. These peptide levels reflect total protein availability and do not distinguish between free, activated, or complexed protein levels in plasma. It should also be acknowledged that these comprehensive mass spectrometry techniques provide a cost-efficient method to investigation many proteins; however, the inclusion of a large cohort of proteins for comparison increases the risk of a type 1 error, which we attempt to limit using FDR. The investigation using ELISA detected coagulation related proteins is also limited by a lack of known clinically relevant levels in swine. Compared to humans, swine are innately hypercoagulable, and swine plasminogen has been shown to be relatively resistant to tPA induced lysis. Finally, with 240 minutes of observation the downstream consequences of these findings beyond this 240 min period are unknown and ideally should be investigated in experiments with greater observation times.
In summary, the findings from this swine study highlight the potential consequences of Zone 1 REBOA to exacerbate hypocoagulation and hyperfibrinolysis in severely injured trauma patients. The dysregulation in several interacting inflammatory systems including the coagulation pathway, fibrinolysis system, and endothelial activation may limit the resuscitative benefits of complete Zone 1 occlusion. Whether partial REBOA catheters can attenuate reperfusion coagulopathy remains to be established.
Supplementary Material
ACKNOWLEDGMENTS
Salary support and research support for animal and laboratory costs were funded by the Department of Defense contract number W81XWH2010205. Research support was also provided by the National Institute of General Medical Sciences of the National Institutes of Health (T32 GM008315; 1RM1GM131968-01). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the Department of Defense.
Support:
This research is funded by the Department of Defense contract number W81XWH2010205. This contract provides salary support and research support for animal and laboratory costs. Research support is also provided by the National Institute of General Medical Sciences of the National Institutes of Health (T32 GM008315). The current major funding source is an RM-1 grant (1RM1GM131968-01).
Footnotes
DISCLOSURES
E.E.M. has patents pending related to coagulation and fibrinolysis diagnostics and therapeutic fibrinolytics and was a cofounder of ThromboTherepeutics. E.E.M. has received grant support from Haemonetics; Hemosonics; Werfen; Stago; and Prytime outside the submitted work. C.J.F is a clinical consultant for Prytime Medical. The remaining authors declare no conflicts of interest.
References
- 1.Brenner M, Bulger EM, Perina DG, et al. Joint statement from the American College of Surgeons Committee on Trauma (ACS COT) and the American College of Emergency Physicians (ACEP) regarding the clinical use of Resuscitative Endovascular Balloon Occlusion of the Aorta (REBOA). Trauma Surg Acute Care Open. 2018;3(1):e000154. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Butler FK Jr, Holcomb JB, Shackelford SA, et al. Advanced Resuscitative Care in Tactical Combat Casualty Care: TCCC Guidelines Change 18–01:14 October 2018. J Spec Oper Med. 2018;18:37–55 [DOI] [PubMed] [Google Scholar]
- 3.Morrison JJ, Ross JD, Rasmussen TE, Midwinter MJ, Jansen JO. Resuscitative endovascular balloon occlusion of the aorta: a gap analysis of severely injured UK combat casualties. Shock. 2014; 41:388–93. [DOI] [PubMed] [Google Scholar]
- 4.Moore EE, Moore HB, Kornblith LZ, et al. Trauma-induced coagulopathy. Nat Rev Dis Primers. 2021;7(1):30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Cralley AL, Moore EE, Kissau D, et al. Combat Casualty Relevant Dismounted Complex Blast Injury Model In Swine. J Trauma Acute Care Surg. 93(2S Suppl 1):S110–S118,2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Morrison JJ, Ross JD, Houston Rt, Watson JD, Sokol KK, Rasmussen TE. Use of resuscitative endovascular balloon occlusion of the aorta in a highly lethal model of noncompressible torso hemorrhage. Shock. 2014;41(2):130–7. [DOI] [PubMed] [Google Scholar]
- 7.Blecher TE, Gunstone MJ. Fibrinolysis, coagulation and haematological findings in normal large white-Wessex cross pigs. Br Vet J. 1969. Feb;125(2):74–81. [DOI] [PubMed] [Google Scholar]
- 8.Schaid TR Jr., Cohen MJ, D’Alessandro A, et al. Trauma Induces Intravascular Hemolysis, Exacerbated by Red Blood Cell Transfusion and Associated with Disrupted Arginine-Nitric Oxide Metabolism. Shock. 2023;59(1):12–9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Chapman MP, Moore EE, Moore HB, et al. Overwhelming tPA release, not PAI-1 degradation, is responsible for hyperfibrinolysis in severely injured trauma patients. J Trauma Acute Care Surg. 2016;80(1):16–23 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Woolley T, Gwyther R, Parmar K, et al. A prospective observational study of acute traumatic coagulopathy in traumatic bleeding from the battlefield. Transfusion. 2020;60 Suppl 3:S52–S61. [DOI] [PubMed] [Google Scholar]
- 11.Rizoli SB, Scarpelini S, Callum J, et al. Clotting factor deficiency in early trauma-associated coagulopathy. J Trauma. 2011;71(5 Suppl 1):S427–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Stettler GR, Moore EE, Moore HB, et al. Variability in international normalized ratio and activated partial thromboplastin time after injury are not explained by coagulation factor deficits. J Trauma Acute Care Surg. 2019;87(3):582–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Karimi M, Bereczky Z, Cohan N, Muszbek L. Factor XIII Deficiency. Semin Thromb Hemost. 2009;35(4):426–38. [DOI] [PubMed] [Google Scholar]
- 14.Martinuzzo M, Barrera L, Altuna D, et al. Effects of Factor XIII Deficiency on Thromboelastography. Thromboelastography with Calcium and Streptokinase Addition is more Sensitive than Solubility Tests. Mediterr J Hematol Infect Dis. 2016;8(1):e2016037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Mutch NJ, Koikkalainen JS, Fraser SR, et al. Model thrombi formed under flow reveal the role of factor XIII-mediated cross-linking in resistance to fibrinolysis. J Thromb Haemost. 2010;8(9):2017–24. [DOI] [PMC free article] [PubMed] [Google Scholar]; 15. Tennent GA, Brennan SO, Stangou AJ, O’Grady J, Hawkins PN, Pepys MB. Human plasma fibrinogen is synthesized in the liver. Blood. 2007;109(5):1971–4. [DOI] [PubMed] [Google Scholar]
- 16.Schlimp CJ, Schochl H. The role of fibrinogen in trauma-induced coagulopathy. Hamostaseologie. 2014;34(1):29–39. [DOI] [PubMed] [Google Scholar]
- 17.Rourke C, Curry N, Khan S, et al. Fibrinogen levels during trauma hemorrhage, response to replacement therapy, and association with patient outcomes. J Thromb Haemost.2012;10(7):1342–51. [DOI] [PubMed] [Google Scholar]
- 18.McQuilten ZK, Wood EM, Bailey M, Cameron PA, Cooper DJ. Fibrinogen is an independent predictor of mortality in major trauma patients: A five-year statewide cohort study. Injury. 2017;48(5):1074–81. [DOI] [PubMed] [Google Scholar]
- 19.Meizoso JP, Moore EE, Pieracci FM, et al. Role of Fibrinogen in Trauma-Induced Coagulopathy. J Am Coll Surg. 2022;234(4):465–73. [DOI] [PubMed] [Google Scholar]
- 20.Cotton BA, Harvin JA, Kostousouv V, et al. Hyperfibrinolysis at admission is an uncommon but highly lethal event associated with shock and prehospital fluid administration. J Trauma Acute Care Surg. 2012;73(2):365–70; discussion 70. [DOI] [PubMed] [Google Scholar]
- 21.Chapman MP, Moore EE, Ramos CR, Ghasabyan A, Harr JN, Chin TL, Stringham JR, Sauaia A, Silliman CC, Banerjee A. Fibrinolysis greater than 3% is the critical value for initiation of antifibrinolytic therapy. J Trauma Acute Care Surg. 2013. Dec;75(6):961–7; discussion 967.. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.reference
- 23.Bouma BN, Mosnier LO. Thrombin activatable fibrinolysis inhibitor (TAFI) at the interface between coagulation and fibrinolysis. Pathophysiol Haemost Thromb. 2003;33(5–6):375–81. [DOI] [PubMed] [Google Scholar]
- 24.Meizoso JP, Karcutskie CA, Ray JJ, Namias N, Schulman CI, Proctor KG. Persistent Fibrinolysis Shutdown Is Associated with Increased Mortality in Severely Injured Trauma Patients. J Am Coll Surg. 2017;224(4):575–82. [DOI] [PubMed] [Google Scholar]
- 25.Moore HB, Moore EE, Liras IN, et al. Acute Fibrinolysis Shutdown after Injury Occurs Frequently and Increases Mortality: A Multicenter Evaluation of 2,540 Severely Injured Patients. J Am Coll Surg. 2016; 222(4):347–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Weinbaum S, Tarbell JM, Damiano ER. The structure and function of the endothelial glycocalyx layer. Annu Rev Biomed Eng. 2007;9:121–67. [DOI] [PubMed] [Google Scholar]
- 27.Bunch CM, Chang E, Moore EE, Moore HB, et al. SHock-INduced Endotheliopathy (SHINE): A mechanistic justification for viscoelastography-guided resuscitation of traumatic and non-traumatic shock. Front Physiol. 2023. 27;14:1094845. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Rahbar E, Cardenas JC, Baimukanova G, et al. Endothelial glycocalyx shedding and vascular permeability in severely injured trauma patients. J Transl Med. 2015;13:117. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Gonzalez Rodriguez E, Ostrowski SR, Cardenas JC, et al. Syndecan-1: A Quantitative Marker for the Endotheliopathy of Trauma. J Am Coll Surg. 2017;225(3):419–27. [DOI] [PubMed] [Google Scholar]
- 30.Johansson PI, Stensballe J, Rasmussen LS, Ostrowski SR. A high admission syndecan-1 level, a marker of endothelial glycocalyx degradation, is associated with inflammation, protein C depletion, fibrinolysis, and increased mortality in trauma patients. Ann Surg. 2011;254(2):194–200 [DOI] [PubMed] [Google Scholar]
- 31.Richter RP, Joiner DM, Griffin RL, et al. Endotheliopathy is Associated with a 24-hour Fibrinolysis Phenotype Described by Low TEG Lysis and High D-Dimer after Trauma: a Secondary Analysis of the PROPPR Study. Ann Surg Open. 2022;3(1). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Hafstrom L, Korsan-Bengtsen K, Bengmark S. Changes in blood clotting and fibrinolysis after liver ischemia in pigs. Am J Surg. 1974;127(3):300–5. [DOI] [PubMed] [Google Scholar]
- 33.Mallett SV, Sugavanam A, Krzanicki DA, et al. Alterations in coagulation following major liver resection. Anaesthesia. 2016;71(6):657–68. [DOI] [PubMed] [Google Scholar]
- 34.Acosta S, Nilsson TK, Bergqvist D, Bjorck M. Activation of fibrinolysis and coagulation in non-occlusive intestinal ischaemia in a pig model. Blood Coagul Fibrinolysis. 2004;15(1):69–76. [DOI] [PubMed] [Google Scholar]
- 35.Hayakawa M, Tsuchida T, Honma Y, et al. Fibrinolytic system activation immediately following trauma was quickly and intensely suppressed in a rat model of severe blunt trauma. Sci Rep. 2021;11(1):20283. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Wu X, Darlington DN, Cap AP. Procoagulant and fibrinolytic activity after polytrauma in rat. Am J Physiol Regul Integr Comp Physiol. 2016;310(4):R323–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Zheng Z, Nayak L, Wang W, et al. An ATF6-tPA pathway in hepatocytes contributes to systemic fibrinolysis and is repressed by DACH1. Blood. 2019;133(7):743–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.