The absence of ADAMTS13 improves early outcomes in an experimental model of trauma with uncontrolled hemorrhage

Abstract

BACKGROUND:

Bleeding after trauma is aggravated by trauma-induced coagulopathy (TIC). In trauma patients with shock, ADAMTS13 (a disintegrin and metalloprotease with a thrombospondin type 1 motif, member 13) antigen is decreased, but its activity can be increased, possibly due to specific cleavage by plasmin. Increased ADAMTS13 activity could aggravate TIC and bleeding. Therefore, this study aimed to determine whether knocking-out ADAMTS13 is protective after trauma with uncontrolled bleeding. Furthermore, we examined the effect of plasmin inhibition with tranexamic acid (TXA) on ADAMTS13 antigen and activity.

METHODS:

Wild-type and ADAMTS13 knockout (ADAMTS13^KO) mice were anesthetized, mechanically ventilated, and subjected to traumatic injury with uncontrolled hemorrhage. In a separate experiment, wild-type mice underwent the same traumatic injury, but with additional blood withdrawal to induce shock and treatment with a single dose of TXA or vehicle. Outcomes included mortality, ADAMTS13 activity, von Willebrand factor (VWF) multimers, and rotational thromboelastometry (ROTEM).

RESULTS:

ADAMTS13^KO mice showed significantly lower mortality rates after trauma compared with wild-type mice (13% vs. 47%, P=0.046), with significantly higher VWF multimers. ROTEM parameters did not differ significantly between ADAMTS13^KO and wild-type mice. In the wild-type mice subjected to trauma and shock, there was a significant increase in ADAMTS13 activity, which correlated with shock severity. Treatment with TXA significantly reduced mortality, but had no significant effect on ADAMTS13 antigen or activity.

CONCLUSION:

Knocking-out ADAMTS13 is associated with improved early survival following trauma, demonstrating a role for ADAMTS13 in contributing to early TIC and bleeding. While ADAMTS13 activity increases after trauma and shock, its levels appear unaffected by TXA.

Traumatic injury is a major contributor to global mortality, with most deaths occurring early.^1–3 Up to half of all early trauma deaths is due to uncontrolled hemorrhage, which is exacerbated by trauma-induced coagulopathy (TIC), developing within minutes after injury.^2,4 Following traumatic injury, von Willebrand factor (VWF) multimers are secreted, which promote platelet adhesion at the injury site.⁵ VWF multimer size is regulated by the protease ADAMTS13 (a disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13), which cleaves VWF multimers, thereby limiting VWF-mediated platelet adhesion. In normal physiology, ADAMTS13 circulates in a closed conformation, with its activity self-regulated by interactions between the terminal CUB (complement components C1r and C1s, sea urchin protein Uegf, and bone morphogenetic protein 1) domains and the central spacer domain. ADAMTS13 shifts to a more open and active conformation upon binding to VWF, enhancing the enzymatic activity of ADAMTS13.^6–8

Circulating levels of high molecular weight VWF multimers are significantly increased after trauma and remain elevated for several days, while ADAMTS13 activity generally decreases in the first hours to days after injury, correlating with an increased risk of microvascular thrombosis and organ dysfunction.^5,9–11 However, we previously showed in a subset of severely injured patients with shock, ADAMTS13 can circulate in a hyperactivated form (ie, without its terminal thrombospondin or CUB domains), which has been associated with hyperfibrinolysis and higher early mortality.¹²

The mechanisms underlying these changes in ADAMTS13 following trauma are incompletely understood. Previous studies have demonstrated that plasmin can degrade ADAMTS13 in vitro, reducing the full-length protein in plasma while potentially increasing its activity via selective cleavage of the thrombospondin or CUB domains, which could be prevented by the addition of tranexamic acid (TXA).^12,13 TXA is associated with improved survival rates in bleeding trauma patients, but its effects on ADAMTS13 after trauma are unknown.¹⁴

This study builds upon previous findings and aims to determine whether knocking-out ADAMTS13 improves early survival in a murine model of traumatic injury with uncontrolled hemorrhage. Moreover, in a more severe model combining trauma with shock, we aimed to explore the early effects of shock on ADAMTS13 activity. Lastly, in this model we studied the effect of inhibiting plasmin with TXA on ADAMTS13 antigen, ADAMTS13 activity, VWF multimers, and platelet activation.

MATERIALS AND METHODS

Murine Models of Trauma and Shock

Animal experiments were approved by the Institutional Animal Care and Use Committee and performed in compliance with the European Parliament directive (2010/63/EU) and the Dutch national law (Wod, 2014). The study was designed according to the ARRIVE, Supplemental Digital Content 3, http://links.lww.com/TA/F440 (Animal Research: Reporting of In Vivo Experiments) criteria.¹⁵

Transgenic Mouse Strain

Cryopreserved sperm of ADAMTS13^KO mice with a B6 background was revived according to in-house procedures and embryos were obtained through in vitro fertilization using female 129S2/SvPasOrlRJ (Janvier Labs, Le Genest-Saint-Isle, France), which are unique in having a full-length ADAMTS13 similar to humans.¹⁶ Fertilized eggs were implanted in female 129S2/SvPasOrlRJ fosters. Heterozygote offspring were used for breeding to produce both wild-type and ADAMTS13^KO offspring. Genetic background was confirmed maximally 21 days after birth using polymerase chain reaction (PCR).

Animal Housing

Mice were housed on-site, 7 days before the experiment, in individually ventilated cages. Animals had access to food (Teklad global 16% protein; Envigo, Indianapolis, IN, USA) and water ad libitum and were exposed to 12-hour day-night cycles. Mice were 8–12 weeks old during the experiment.

Mice Anesthesia, Ventilation, and Cannulation

Mice were anesthetized in an induction chamber with 4.5% sevoflurane (AbbVie, Chicago, IL, USA) and 60% FiO₂, followed by an intraperitoneal injection of 0.06 μg/g fentanyl, with an additional 0.03 μg/g after 30 minutes (Hameln Pharma, Hameln, Germany). While receiving mask ventilation with continuous anesthesia (FiO2 of 50%, 4.5% sevoflurane), the trachea was cannulated, after which the mice were mechanically ventilated (VentElite; Harvard Apparatus, Holliston, MA, USA) with tidal volumes of 7.2 μL/g, a respiratory rate of 190 breaths per minute, an inspiratory-to-expiratory ratio of 1:1.5, and 40% FiO2. A lung recruitment maneuver (inspiratory sigh of 20%) was performed every 30 minutes. Mice remained under 2.5% sevoflurane anesthesia throughout the experiment, and anesthetic depth was monitored every 10 minutes by assessing the absence of pedal reflex and checking for hemodynamic response to painful stimuli.

The right carotid artery was cannulated with a two-French polyurethane catheter (Instech, Plymouth Meeting, PA, USA), which was connected to a pressure transducer and amplifier (Quad Bridge Amp; ADInstruments, Dunedin, New Zealand). Data were acquired using Powerlab version 4/30 and Labchart version 8 (ADInstruments). The jugular vein was cannulated with a one-French polyurethane catheter (Instech), and the animals received a continuous intravenous infusion of fentanyl (0.12 μg/g/h; Pump 11 Pico Plus Elite; Harvard Apparatus) in 20 μL/g/h Ringer’s lactate supplemented with 15 mM glucose and 2 mM sodium bicarbonate (Perfusor; BBraun, Melsungen, Germany). Body temperature was maintained at 37°C using a rectal thermometer, heated table, and heat lamp. Mice were randomized into 2 distinct models: a trauma with uncontrolled hemorrhage model and a trauma and shock model.

The first model included a control group and a trauma group (n=15 per group). The groups size was based on previous pilot experiments and limited by ethical approval in order to reduce animal use. Trauma consisted of bilateral tibia and fibula fractures, followed by 3 crushes on both sides of the rectus sheath and 5 individual crushes to the small intestine, located one centimeter distal to the Treitz ligament. To simulate uncontrolled hemorrhage, the left lobe of the liver was punctured using an 8 mm biopsy punch (Kai Medical, Honolulu, HI, USA), and the abdomen was left open, covered only with saline-wetted gauze. Following trauma, or at the corresponding timepoint in the control group, mice received a 400 μL intravenous infusion of Ringer’s lactate over 30 seconds. This model did not involve any blood withdrawal, as the goal was to observe the impact of ADAMTS13^KO in primary hemostasis and mortality in conditions of uncontrolled bleeding.

In the second model, mice were randomized into one of the following 2 treatment groups: vehicle and TXA (n=15 per group). Mice underwent similar trauma induction as described in the first model. However, this model introduced controlled blood withdrawals to simulate shock. Blood was withdrawn through the carotid artery at two fixed time points: before the traumatic injury and immediately after the liver puncture. In addition, incremental blood withdrawal (50 μL at a time) was performed to target a mean arterial pressure (MAP) of 20 to 25 mm Hg. Body temperature was passively lowered from 37°C to 35°C to further simulate shock conditions. Following trauma, all mice received the same 400 μL intravenous infusion of Ringer’s lactate as in the first model.

According to group allocation, following trauma and blood withdrawal, mice received an intravenous bolus of 10 mg/kg TXA (Mylan, Canonsburg, PA, USA), or corresponding vehicle (NaCl 0.9%) in a total volume of 4 µL/g. The dose of TXA is similar to the dose used in human trauma patients.¹⁴

Terminal Blood Sampling and Measurements

After thirty minutes the experiment was ended by a terminal blood withdrawal. Lungs were collected for histopathologic analysis. The time of mortality was recorded for mice that did not survive the experiment. Blood and organ measurements were only performed on mice that survived the whole experiment. Further details regarding sample processing are described in the Supplementary Methods, Supplemental Digital Content 2, http://links.lww.com/TA/F439.

Measurements

ADAMTS13 and VWF Measurements

Total ADAMTS13 antigen was determined using ELISA as previously described.¹⁷ ADAMTS13 activity was determined using cattle FRETS-VWF71.¹⁸ High molecular weight VWF multimers were determined using electrophoresis and western blot. A detailed description of ADAMTS13 and VWF measurements is described in the Supplementary Methods, Supplemental Digital Content 2, http://links.lww.com/TA/F439.

Rotational Thromboelastometry

ROTEM Delta, EXTEM was performed according to the manufacturer’s guidelines. The ROTEM minicups were used, which use 105 μL of whole blood. Clotting time is the time from activation of coagulation until the clot has reached a 2 mm amplitude. Maximum clot firmness is the maximum amplitude reached during the time of the assay (90 min). Maximum lysis is the percentage of clot lysis relative to the maximum amplitude at the end of the assay.

Platelet Activation and Platelet-leukocyte Aggregates

Details are described in the Supplementary Methods, Supplemental Digital Content 2, http://links.lww.com/TA/F439.

Plasmin Activator Inhibitor 1

Plasmin activator inhibitor 1 (PAI-1) was measured using ELISA according to manufacturer’s instructions (Innov-Research, Novi, MI, USA). This assay measures the active PAI-1 (ie, not complexed to plasminogen activators).

Organ Histopathology

After the experiment, the right lung was fixed in 4% formaldehyde, embedded in paraffin, and cut into 4-µm thick sections. Sections were stained with hematoxylin and eosin. Assessment of lung injury was done by a pathologist, using a scoring system for lung edema, interstitial inflammatory cell infiltration, endothelialitis, hemorrhage, and the presence of thrombi or coagulation. The following scale was used: 0 (absent), 1 (mild), 2 (moderate), or 3 (severe).

Statistical Analyses

All data were analyzed using SPSS version 25.0 (IBM, Armonk, NY, USA). Graphs were made using GraphPad version 9.0 (Prism, La Jolla, CA, USA). Continuous data were seen as nonparametric, presented as median with interquartile range, and analyzed with Kruskal-Wallis test with post-hoc Dunn test or with the Mann-Whitney U test. Binomial data were analyzed with the Fisher exact test or the χ² test. Survival distributions were displayed as a Kaplan-Meier curve and analyzed with the log-rank test. Linear regression was used to analyze the correlation between ADAMTS13 activity and lactate. A P-value < 0.05 was considered statistically significant.

RESULTS

ADAMTS13^KO Mice Are Protected From Traumatic Exsanguination

An overview of the model of trauma with uncontrolled hemorrhage is shown in Fig. 1A. Genetic background of wild-type and ADAMTS13^KO mice was confirmed by PCR (Fig. 1B). Baseline characteristics and arterial blood gas analysis of both groups are shown in Supplementary Table S1, Supplemental Digital Content 2, http://links.lww.com/TA/F439.

Figure 1.

ADAMTS13^KO mice are protected from traumatic exsanguination. (A) Overview of the model of trauma with uncontrolled hemorrhage. (B) Genetic background was confirmed using polymerase chain reaction. We observed a 400 bp PCR product for wild-type mice, a 300 bp PCR product for ADAMTS13^KO mice, and both 400 bp and 300 bp PCR product for heterozygote mice. (C) ADAMTS13^KO mice had a significantly improved survival compared with wild-type mice after trauma. (D) Mean arterial pressure during the experiment. Data are presented as survival curve, mean with its standard error, or boxplot with all datapoints. *P <0.05, **P <0.01.

Mortality 30 minutes after trauma was significantly higher in wild-type mice compared with ADAMTS13^KO mice (47 vs. 13%, P=0.046; Fig. 1C). Interestingly, the mean arterial pressure was lower in the wild-type mice compared with the ADAMTS13^KO mice, albeit not significantly (Fig. 1D). Lactate was increased in both wild-type and ADAMTS13^KO mice after trauma, but did not differ between these groups (Supplementary Table S1, Supplemental Digital Content 2, http://links.lww.com/TA/F439).

ADAMTS13 antigen and activity were undetectable in all ADAMTS13^KO mice (Fig. 2A and B), and did not differ between wild-type trauma and wild-type control. Platelet count was not decreased after trauma and also did not differ significantly between wild-type and ADAMTS13^KO mice in this model (Fig. 2C). High molecular weight VWF multimers were significantly higher after trauma compared with control in wild-type mice. Moreover, more high molecular weight multimers were found in ADAMTS13^KO mice compared with wild-type mice, both in the control group and after trauma (Fig. 2D). ROTEM EXTEM clotting time, maximum clot firmness, and maximum lysis did not differ between ADAMTS13^KO and wild-type mice (Fig. 3A–3C). Active PAI-1 also did not differ between groups (Fig. 3D).

Figure 2.

ADAMTS13 and VWF in wild-type and ADAMTS13^KO mice. (A, B) Total ADAMTS13 antigen and activity were undetectable in all ADAMTS13^KO mice, but did not significantly differ between trauma and control. (C) Platelet counts did not significantly differ between wild-type and ADAMTS13^KO mice. (D) Von Willebrand factor (VWF) percentage of high molecular weight multimers (HMWM) were significantly lower in wild-type compared with ADAMTS13^KO mice. Data are presented as boxplots with all datapoints. **P <0.01, ***P <0.001.

Figure 3.

Rotational thromboelastometry and fibrinolysis in wild-type and ADAMTS13^KO mice. (A–C) Rotational thromboelastometry EXTEM showed no difference between wild-type and ADAMTS13^KO mice in clotting time, maximum clot firmness, and maximum lysis. (D) Active plasma plasminogen activator inhibitor 1 (PAI-1) was also not different between wild-type and ADAMTS13^KO mice.

Trauma did not induce expression of platelet activation markers. CD62P and GPIIb/IIIa (JON/A) expression on the platelet membrane with/without agonistic stimulation did not differ between wild-type and ADAMTS13^KO mice (Fig. 4A and B). Platelet-leukocyte aggregates, which were confirmed by imaging flow cytometry, also did not differ between wild-type and ADAMTS13^KO mice (Fig. 4C). There were no significant differences in organ damage between wild-type and ADAMTS13^KO mice and no signs of microthrombosis in lung tissue were observed (Supplementary Fig. S1, Supplemental Digital Content 2, http://links.lww.com/TA/F439).

Figure 4.

Platelet activation and reactivity in wild-type and ADAMTS13^KO mice. Platelets were identified based on CD41 expression as well as side scatter characteristics. Expression of CD62P (A) and JON/A (GPIIb/IIIa; B) was assessed with and without agonists: 20 uM ADP, 4 mg/mL ristocetin, or 1.5 U/mL thrombin. Platelet-leukocyte aggregates were determine using flow cytometry and visually confirmed using ImageStream (C). No difference was found between wild-type and ADAMTS13^KO mice. Data presented as boxplot with all datapoints.

ADAMTS13 Activity and VWF Multimers Are Elevated After Trauma and Shock but Remain Unaffected by TXA

We studied whether TXA improves outcomes by limiting ADAMTS13 degradation by plasmin. Given that shock induces hyperfibrinolysis, we used a different experimental model, by adding blood withdrawals to induce shock (Fig. 5A). Baseline characteristics and arterial blood gas analysis for this model are shown in Supplementary Table S2, Supplemental Digital Content 2, http://links.lww.com/TA/F439. We found that TXA significantly decreased mortality rates compared to vehicle (13 vs. 47%, P=0.046; Fig. 5B). MAP and lactate levels did not differ between TXA and vehicle (Supplementary Fig. S2, Supplemental Digital Content 2, http://links.lww.com/TA/F439 and Fig. 5C). Trauma and shock resulted in a significant decrease of active PAI-1 activity; however, no significant difference in PAI-1 activity was observed between the TXA and vehicle groups (Fig. 5D). Platelet count and ROTEM parameters did not differ between groups (Supplementary Fig. S3A–S3D, Supplemental Digital Content 2, http://links.lww.com/TA/F439). ADAMTS13 antigen was not significantly affected by trauma and shock (Fig. 6A). However, ADAMTS13 activity was significantly higher in the trauma and shock group, when compared with sham (154 [91–239] vs. 103 [96–121], P=0.04; Fig. 6B). In addition, the percentage of high molecular weight VWF was significantly increased after trauma and shock (62 [54–67] vs. 53 [48–57], P=0.01; Fig. 6C). ADAMTS13 activity was significantly associated with shock severity, as measured by lactate concentration (Fig. 6D). There were no significant differences between ADAMTS13 antigen, ADAMTS13 activity, or VWF multimers between TXA and vehicle.

Figure 5.

Overview of the trauma and shock model, and the effects of TXA on mortality. (A) Overview of the model of trauma and shock. The sham group did not undergo trauma and blood withdrawal. (B) Tranexamic acid (TS + TXA) was associated with a significantly decreased mortality compared with vehicle (TS + VEH). (C) Lactate was significantly increased after trauma and shock. (D) Trauma and shock resulted in a significant decrease of active plasminogen activator inhibitor 1 (PAI-1). There was no significant difference in PAI-1 activity between TXA and vehicle. Data are presented as survival curve, mean with its standard error, and boxplot with all datapoints. *P <0.05, ***P <0.001.

Figure 6.

The effect of trauma and shock on ADAMTS13, which remained unaffected by tranexamic acid. (A) Total ADAMTS13 (ie, both full-length and truncated forms) was not significantly different after trauma and shock compared with sham and did not differ between TXA and vehicle. (B) ADAMTS13 activity was significantly increased after trauma and shock but was not influenced by TXA treatment. (C) After trauma and shock, the percentage of high molecular weight multimers (HMWM) von Willebrand factor (VWF) was significantly increased. VWF HMWM did not differ significantly between TXA and vehicle. (D) Both trauma and shock groups combined. After trauma and shock, ADAMTS13 activity was associated with shock severity, measured by lactate. Data are presented as survival curve, mean with its standard error, boxplot with all datapoints, or scatter dot plot with linear regression function. *P <0.05, ***P <0.001.

DISCUSSION

In this study, we explored the role of ADAMTS13 in two distinct experimental models of trauma and shock. Firstly, in a model of severe trauma with uncontrolled bleeding, we found that ADAMTS13^KO mice demonstrated significantly improved survival compared with wild-type mice, which was associated with increased circulation of high molecular weight VWF multimers. In a second model, wild-type mice were subjected to additional blood withdrawals, to further induce shock. Trauma and shock resulted in increased ADAMTS13 activity and VWF multimers. Inhibiting plasmin using TXA improved survival, but did not affect the depletion of PAI-1 and did not prevent an increase in ADAMTS13 activity and VWF.

Most prior clinical research has linked decreased ADAMTS13 antigen and activity levels with late-stage complications such as thrombosis and organ failure.^5,9,19,20 Recently, we found that in the early stage after trauma, ADAMTS13 can be truncated and hyperactivated in severely injured patients (ie, low ADAMTS13 antigen levels, but high activity levels), associated with hyperfibrinolysis and increased mortality.¹². An explanation for the seemingly contradictory findings on the role of A13 may be that TIC changes over time, in which early on (ie, during bleeding), low ADAMTS13 is beneficial by cleaving VWF multimers or even fibrin(ogen). Later in the trajectory of trauma, after which bleeding has been contained and TIC shifts to a more prothrombotic state, low ADAMTS13 is associated with (micro)thrombosis and organ failure. In this study, we hypothesized that during uncontrolled bleeding, the absence of ADAMTS13 may improve early outcomes. Indeed, our results demonstrate improved survival in the ADAMTS13^KO group, which could be attributed to a reduction in massive liver bleeding. This benefit may be explained by enhanced coagulation, potentially mediated by an increased amount of circulating high molecular weight VWF multimers. However, platelet activation and reactivity did not differ between wild-type and ADAMTS13^KO mice. Also, ROTEM parameters did not differ between wild-type and ADAMTS13^KO mice. Of note, platelet activation markers were also not different between the control group and the trauma group, which is in contrast to prior studies in trauma patients that showed increased presence of CD62P and GPIIb/IIIa, but reduced reactivity in response to agonistic stimulation.^21–23 We propose that there was likely a significant survival bias in our model, underestimating the extent of coagulopathy. In addition, we were unable to perform platelet aggregation and/or adhesion assays.

It is important to note that, within the duration of our model, ADAMTS13^KO mice did not show signs of thrombotic thrombocytopenic purpura (TTP). Platelet counts and organ histopathologic analysis were not significantly affected by trauma and no differences were observed between ADAMTS13^KO and wild-type mice within the trauma groups. Experimental studies show that ADAMTS13^KO mice generally require an exogenous stimulus (eg, by intravenous endotoxins or a challenge of recombinant ultra-large VWF multimers) to trigger TTP-like symptoms, which may be explained by genetic factors, or alternatively, by plasmin-mediated VWF cleavage.^24–27 It is unknown whether trauma would have elicited a TTP-like reaction after a longer observation period.

In a separate experimental model, using only wild-type mice, we incorporated blood withdrawal in addition to trauma, which increased shock severity. We found a strong positive correlation between shock severity and ADAMTS13 activity. The percentage of high molecular weight VWF multimers was also significantly increased after trauma and shock, which has been described in previous studies as well.^5,19

A potential mechanism behind the correlation between shock severity and ADAMTS13 activity may be fibrinolysis. It is well known that shock results in tPA release, which is driven by tissue hypoperfusion.²⁸. In vitro studies showed that plasmin can truncate ADAMTS13, resulting in increased ADAMTS13 activity.^12,13 A hyperactive form of ADAMTS13 could lose its substrate specificity, leading to unselective cleavage of other proteins, such as fibrinogen and fibrin.^12,29,30 Here, although we did not show hyperfibrinolysis in the ROTEM EXTEM parameters, we did observe a significant reduction in active PAI-1 after trauma and shock, likely through increased tPA-PAI-1 complexes, suggesting enhanced fibrinolytic activity.³¹ Fibrinolysis due to shock may explain why ADAMTS13 activity was only increased in the more severe trauma and shock model, and not in the less severe model.

TXA is widely used as an antifibrinolytic treatment and has been shown to improve survival in bleeding trauma patients.¹⁴ We studied whether TXA would protect ADAMTS13 from degradation, thereby preventing ADAMTS13 hyperactivity and hyperfibrinolysis. Although we found that TXA was associated with decreased mortality rates, we did not find a difference between ADAMTS13 antigen or activity after TXA treatment, nor did we see any difference between TXA and vehicle on PAI-1 levels and VWF multimers. One potential explanation for these findings may be related to the dosing of TXA, which was based on clinical practice, rather than its specific effects on ADAMTS13.^14,32 The reduced PAI-1 levels in the TXA group, similar to the vehicle group, suggest that the dosing may have been insufficient to fully suppress plasmin generation, allowing for the truncation of ADAMTS13 and an associated increase in ADAMTS13 activity. In addition, we observed considerable variability in ADAMTS13 antigen and activity across all groups following trauma and shock, which may have limited our ability to detect a minor effect of TXA on ADAMTS13 activity. In addition, the lack of difference between ADAMTS13 antigen may be explained by the detection method that we used. Due to lack of specific anti-murine ADAMTS13 antibodies, we were unable to differentiate between full-length ADAMTS13 and cleaved (eg, truncated) variants in the ADAMTS13 antigen assay, potentially underestimating the extent of ADAMTS13 degradation. However, the lack of an observed effect of TXA on ADAMTS13 activity may also indicate the involvement of additional pathways influencing the early changes in ADAMTS13 after trauma.

This study has several limitations. As the study was designed to investigate the early bleeding phase, the impact of an absence of ADAMTS13 on later complications such as thrombosis and organ failure could not be studied. Second, we did not directly quantify blood loss in the uncontrolled bleeding model, which would have strengthened the association between the survival benefit and improved primary hemostasis in ADAMTS13^KO mice. Also, we could not identify specific ADAMTS13 fragments, as specific anti-mouse antibodies were not developed. Finally, a lack of sample volume limited our assessment of fibrinolysis. Although reduced active PAI-1 suggests increased fibrinolytic activation, we did not measure additional markers such as plasmin–antiplasmin complexes or D-dimer.

In conclusion, knocking-out ADAMTS13 is associated with elevated high molecular weight VWF multimers and improved survival after trauma with uncontrolled bleeding. After similar trauma with more severe shock, ADAMTS13 activity is increased, which correlates with shock severity. A low dose of TXA did not significantly affect ADAMTS13 activity in our model.

Supplementary Material

ta-101-689-s001.pdf ^{(1.6MB, pdf)}

ta-101-689-s002.docx ^{(3.2MB, docx)}

ta-101-689-s003.pdf ^{(111.2KB, pdf)}

ta-101-689-s004.pdf ^{(1.3MB, pdf)}