Damage Control Orthopaedics Induced Less Trauma-Induced Coagulopathy than Early Total Care in a Porcine Polytrauma Model

Disorders of the coagulopathic system are well-known in polytraumatised patients. The severity of trauma can cause immediate activation of the coagulation and fibrinolysis systems, which significantly impact injured patients’ outcomes [1]. Trauma-induced coagulopathy has been described in up to 25% of severely injured patients, leading to a fivefold increase in mortality in these patients [1‒3]. After the initial trauma itself, surgical interventions are also known to induce and/or exacerbate coagulopathic disorders (CDs). However, the impact of differences in surgical invasiveness, i.e., in damage control orthopaedics (DCO) versus early total care (ETC), on the development of CD has not been fully understood but is of clinical relevance. ETC and DCO primarily differ in their level of surgical invasiveness. The fundamental principle of ETC involves the definitive fixation of long-bone fractures during the initial fracture surgery, thereby facilitating rapid patient mobilisation. Conversely, DCO focuses on temporary fracture stabilisation to maintain patient stability before performing definitive surgical fixation, which necessitates follow-up surgeries and extends the duration of hospitalisation [4, 5].

Among the parameters that contribute to the onset of CD, plasminogen activator inhibitor-1 (PAI-1) and thrombin-antithrombin (TAT) are of special interest. Moore et al. [6‒8] reported on the “fibrinolytic shutdown,” a term describing the neutralising effects of pro-coagulant plasma proteins such as PAI-1, by high concentrations of fibrinolytic enzymes like tissue plasminogen activator (tPA) or urokinase-type plasminogen activator (uPA). Subsequently, increased production of plasmin and hyperfibrinolysis in severely injured patients contributes to high mortality rates by consuming clotting factors and causing acidosis and hypothermia [1]. While some authors suggest that coagulopathy is caused by high PAI-1 concentrations that peak shortly after trauma and lead to a significant consumption of the protein [1, 9], others argue that PAI-1 is only a bystander to coagulopathic and fibrinolytic cascade perturbations [10]. Furthermore, the occurrence of postoperative resistance to fibrinolysis has also been observed in various surgical specialties and has been linked to a significant rise in PAI-1 levels [11].

In contrast, TAT complexes are well-established parameters for detecting disorders in the coagulopathic cascade. The TAT complex is a molecular complex consisting of thrombin and antithrombin (AT), the inhibitor of thrombin [12, 13]. Quantification assays for TAT allow the rapid and specific determination of coagulation, even in the absence of thrombosis. These assays can detect a pre-thrombotic event [14, 15]. TAT measurements have been used for the diagnosis and the assessment of treatment for disseminated intravascular coagulation, deep vein thrombosis, and thromboembolism due to their early detection capabilities [12, 16].

Polytrauma (PT) and haemorrhagic shock (HS) are leading causes of prehospital as well as in-hospital mortality, and CD complicates the management of polytraumatised patients. Detailed knowledge of the influence of surgical treatment strategies on CD is necessary to develop effective and individualised treatment algorithms. In particular, investigating the effect of surgical treatment approaches on the circulating expression of the previously mentioned pro-coagulant markers PAI-1 and TAT is of interest. The present study focuses on a possible association between fracture fixation strategy and these key coagulant proteins in a clinically relevant large-animal PT model.

A well-established porcine PT and HS model was used [17]. The experimental design was constructed and performed in accordance with the principles of laboratory animal care, the ARRIVE guidelines, and the Federation of European Laboratory Animal Science Association (FELASA). To validate compliance with the ARRIVE guidelines, appropriate documentation was made, meticulously prepared, and maintained. Legal approval was obtained (Landesamt für Natur, Umwelt und Verbraucherschutz Nordrhein-Westfalen, Recklinghausen; AZ 81.02.04.2020.A215) [18]. The data presented in this paper were collected in the context of a larger study for the benefit of the principles of the 3Rs (Replacement, Refinement, and Reduction) [19]. Nevertheless, the present study has been conceived as an independent investigation.

Male German Landrace pigs (Sus scrofa domesticus) with a body weight of 35 ± 5 kg and aged between 12 and 16 weeks were housed in ventilated rooms. After arrival, all animals were clinically examined by a veterinarian and allowed to acclimatise for a minimum of 7 days prior to the experiments. The animals were maintained in a controlled environment at 21 ± 3°C with approximately 30% humidity, under a 12-h light/dark cycle.

The animals were fasted for 12 h before PT induction with water ad libitum, followed by premedication with azaperone (Stresnil^™, Janssen, Germany) and ketamine (Ketanest, Pfizer, NY, USA). The general instrumentation and anaesthesia were performed similarly to the descriptions in Horst et al. [17] and Lupu et al. [19]. After intravenous injection of an induction dose of propofol 2% (Fresenius SE & Co. KGaA, Homburg, Germany) and midazolam (PANPHARMA GmbH, Trittau, Germany), the animals underwent endotracheal intubation. Fentanyl (PANPHARMA GmbH, Trittau, Germany) was applied as an analgesic. General anaesthesia was maintained with midazolam (PANPHARMA GmbH, Germany) and propofol (Fresenius SE & Co. KGaA, Homburg, Germany). Volume-controlled mechanical ventilation (8–12 mL/kg) was performed using a closed anaesthesia device (Draeger Evita 4, Draeger Safety AG & Co. KGaA, Lübeck, Germany). A central venous catheter (Four-Lumen Catheter, 8.5 Fr., Arrow Catheter, Teleflex Medical GmbH, Fellbach, Germany) was introduced in the external jugular vein, and a three-lumen haemodialysis catheter (12.0 Fr., Arrow Catheter, Teleflex Medical GmbH, Fellbach, Germany) was placed in the right femoral vein. Vital parameters were monitored after installing a femoral arterial line (Vygon GmbH & Co. KG, Aachen, Germany) by means of a Philips patient monitor (MP50, Philips Health Systems, Hamburg, Germany). Additionally, a suprapubic catheter (12.0 Fr, Cystofix^®, B. Braun AG, Melsungen, Germany) was prepared. Throughout the duration of the experiment (72 h), general anaesthesia was maintained, and animals were continuously infused with crystalloid solution (Sterofundin ISO^®, B. Braun AG, Melsungen, Germany) and received parental nutrition (Aminoven, Fresenius Kabi, Germany). Furthermore, haemoglobin, platelet (PLT), haematocrit, lactate, and base excess levels were measured at different time points (Table 1).

After achieving and maintaining physiological vital parameters, the animals were subjected to the combined PT. Prior to PT induction, fluid substitution was reduced to 0.1 mL/kgBW/h to keep the infusion lines open, FiO₂ was reduced to 0.21 to mimic ambient air, and hypothermia was not prevented. Blunt-chest trauma was induced by using a bolt gun machine (Blitz-Kerner, turbocut Jopp GmbH, Bad Neustadt an der Saale, Germany) with cattle killing cartridges (Dynamit-Nobel, cartridge 9 × 17; Vienna, Austria) placed on a pair of custom-made lead (10 mm) and steel (8 mm) panels located at the right dorsal lower chest. The bolt was shot while the lungs were inflated. Next, a bilateral femoral fracture was applied by means of a bolt shot using a T-shaped punch positioned on the middle third of the femur. Afterwards, a midline laparotomy was performed, the left liver lobe was explored, and a crosswise incision was made (4.5 cm × 4.5 cm, half the depth of the tissue), representing a penetrating liver injury. After repositioning of the liver, a pressure-controlled HS was induced by withdrawing blood until a mean arterial pressure of 40 ± 5 mm Hg was achieved. Subsequently, the animals were left untreated for 1.5 h, imitating a realistic interval between trauma and hospital admission [20].

At the end of the 1.5-h shock phase, the animals were resuscitated according to the trauma guidelines (ATLS^® and S3-Polytrauma-Guideline); the FiO₂ was adjusted to baseline, warmed blood (Citrate Phosphate Dextrose Adenine DONOpacks, Lmb Technologie GmbH, Oberding, Germany) was re-infused together with crystalloid solution, and normothermia (38.7–39.8°C) was maintained using a forced-air warming system [21].

Finally, the first operative phase started, and both femur fractures were stabilised by installation of an external fixator (RadioLucent Fixator, Orthofix, TX, USA; DCO group: n = 8) or by intramedullary nailing (IMN) (T2-System; Stryker GmbH & Co. KG, Duisburg, Germany; ETC group, n = 8). The control group (CG: n = 6) was not traumatised but received the same monitoring instrumentation, anaesthesia, mechanical ventilation, and nutrition. The animals were turned every 4–6 h to support respiration and prevent decubiti. Mean arterial pressure was kept at >60 mm Hg.

Venous blood samples were collected at the beginning of the experiment (0 h) and after 1.5, 2.5, 24, 48, and 72 h. After centrifugation, the EDTA-plasma was snap-frozen and stored according to the immunoassay manufacturers’ instructions. ELISA kits were purchased for PAI-1 (anti-porcine; Nordic Biosite AB, Taby, Sweden) and TAT complex (anti-human; Siemens Healthineers, Eschborn, Germany). The anti-human ELISA kit for the determination of TAT levels has previously been described as successfully cross-reacting with pigs [22].

GraphPad Prism version 9.2.0 (California, USA) was used for data visualisation and statistical analysis. The Shapiro-Wilk test was used to test for normality, after which either a two-way ANOVA followed by Tukey’s post hoc test, or a Kruskal-Wallis test, was performed as appropriate. Data are presented as mean or median, accompanied by standard error of the mean or interquartile range as applicable.

For each individual trajectory, an area under the curve (AUC) was calculated. These individual AUCs reflected the development of each particular outcome parameter over time. All individual AUC values were then summarised using the median and interquartile range for each of the experimental groups. The Mann-Whitney-U test was used to compare the median AUCs. The percentage change in the AUC represents the difference between the AUC median values of the CG, ETC, and DCO groups. For all analyses, the significance limit was set at p < 0.05.

The present study included a total of 22 animals. One animal in the ETC group died prematurely at 60 h, most likely due to cardiorespiratory failure, and was therefore excluded from further analyses. Table 1 presents the following laboratory data at different time points, including haemoglobin, PLT, haematocrit, lactate, and base excess levels.

Compared to the CG, the plasma concentration of PAI-1 was significantly raised in the ETC group after 2.5 h (p = 0.005) and 24 h (p = 0.0002). Furthermore, a significant increase in PAI-1 concentration was observed after 24 h in the ETC group compared to the DCO group (p = 0.0004). In the DCO group, the plasma PAI-1 concentration was significantly deregulated compared to the CG at 2.5 h after the trauma (p = 0.0016). At 48 h and 72 h after the trauma, no further significant differences were observed between the groups (Fig. 1a). The AUC summary supported the significantly increased production of PAI-1 in the ETC group compared to the CG (p = 0.016) (Fig. 1b).

While kinetics of TAT were similar between both treatment groups and the CG in the early phase after the trauma, the plasma TAT concentration in the ETC group had increased significantly by the end of the observation period (after 72 h) compared to both the CG and DCO groups (CG vs. ETC: p = 0.029; DCO vs. ETC: p = 0.032). Furthermore, the AUC also showed significantly increased plasma concentrations of TAT in the ETC compared to the DCO group (p = 0.026) (Fig. 2b).

Acute and ongoing blood loss, as well as CD, are responsible for high mortality rates in polytraumatised patients and represent a serious problem in acute treatment strategies and the development of complications during the further clinical course. The pathology of trauma-induced CD is complex and therefore the topic of ongoing research. Accordingly, this study focuses on key coagulant proteins (i.e., PAI-1 and TAT) that have not been previously well-investigated in relation to the surgical strategies of DCO versus ETC. The findings can be summarised as follows:

• a.

  ETC was associated with a significant increase in PAI-1 lasting for 24 h after trauma compared to the CG and DCO groups, while increased values of PAI-1 in the DCO group were only significantly elevated after 2.5 h compared to the CG.

• b.

  By the end of the observation period, the ETC group showed a significant increase in TAT protein levels compared to the CG and the DCO groups, while the DCO group showed comparable values to the CG.

Previous studies have investigated the imbalance of pro- and anticoagulatory parameters, such as PAI-1 and tPA, in polytraumatised patients that were strongly associated with unfavourable outcomes [23, 24]. However, this study reports on the direct influence of surgical treatment strategies on coagulation activity, specifically on the expression of the two pro-coagulant markers PAI-1 and TAT. Against this background, PAI-1, a pro-coagulant plasma protein, showed the association of pathologic fibrinolysis with mortality after trauma [7]. It has been shown that upregulation of PAI-1 after trauma leads to increased lung injury, pulmonary fibrosis, and thrombotic diathesis [25, 26]. Accordingly, this study has demonstrated increased PAI-1 plasma concentrations in both treatment groups. However, compared to the CG, this elevation was more pronounced and lasted longer in the ETC group compared to the DCO group. These findings support the clinical observations of ETC being associated with a higher degree of coagulopathy after PT [27]. Furthermore, it has previously been shown that in patients with severe injury and HS, circulating PAI-1 concentrations were elevated and corresponded well with high injury severity scores. Therefore, PAI-1 may be representative of the severity of endothelial damage and PLT activation after injury [28]. Moreover, Kang et al. [29] showed that trauma-induced elevation of the pro-inflammatory cytokine IL-6 induces PAI-1 production by vascular endothelial cells, contributing to thromboembolic events. Taking this mechanism into account, the increased plasma PAI-1 concentrations in the ETC group may reflect the synergistic effects of both trauma and the subsequent surgical burden by prompt definitive surgical stabilisation via IMN. Additionally, Renckens et al. [30] described the inhibitory effect of elevated plasma PAI-1 levels on fibrinolysis in trauma patients, and Giannoudis et al. [31] found that the coagulation system was negatively affected by the reduction in coagulation parameters such as AT and fibrinogen after IMN. Furthermore, Hildebrand et al. [27] showed that IMN was associated with a significantly increased fibrinolytic response. According to the literature, results from the current study show that IMN should be used with caution, especially in patients with pre-existing PT, as prompt definitive surgical stabilisation may aggravate CD in this already vulnerable population.

In vivo research has already shown that PT is associated with increased TAT concentrations [32, 33]. However, the potential correlation between the degree of surgical trauma and circulating TAT levels has not yet been fully investigated. Against this background, Suzuki et al. [34] reported that plasma TAT concentrations were increased after severe head trauma. Also, Caspers et al. [35] described a systemic upregulation of TAT after PT and demonstrated a positive association between TAT concentrations and trauma severity. Interestingly, systemic plasma TAT levels have also been shown to be increased in patients with acute respiratory distress syndrome, which is of particular relevance in PT, as blunt-chest trauma can be found in more than 50% of all polytraumatised patients and influences the surgical stabilisation strategy [17, 36]. Pape et al. [37] and Volpin et al. [38] described the invasive character of ETC, which can contribute to the occurrence and progression of pulmonary pathology, with potential pulmonary complication. Accordingly, it was shown that systemic levels of TAT are crucial in pulmonary coagulopathy, further demonstrating the widespread role of these molecules in the field of PT [36]. Also, the surgical burden has been shown to enhance pro-coagulant activity, although this was not specifically investigated in polytraumatised patients [39, 40]. In line with the cited literature, it therefore may be assumed that elevated TAT values reflect both the traumatic and the surgical burden in the ETC group, comparable to PAI-1 concentration findings. Moreover, TAT could also be used as a biomarker for the prediction of post-traumatic complications. Against this background, data presented by Lesbo et al. [41] showed a marked increase in systemic TAT concentrations 72 h after trauma. The authors described a strong association between these increased concentrations, mortality, and thromboembolic events within the first 30 days after trauma, which could help to predict the further clinical course in polytraumatised patients.

Systemic levels of PAI-1 and TAT showed distinct patterns over time, indicating associations with both trauma severity and surgical trauma. The results from the present study emphasise the importance of a thoughtful choice between the application of ETC or DCO in polytraumatised patients to prevent CD-associated, post-traumatic complications.

The applied large animal model is well-established, standardised, and realistically mimics the clinical situation. Moreover, performing the experiments on pigs allowed the best possible comparison to the human situation [42]. Furthermore, all experiments were performed by the same main investigator (Ü.M.) following established trauma guidelines (ATLS^® and S3-Polytrauma-Guideline). However, no model can perfectly imitate the human situation as trauma patients vary in age, gender, ethnicity, and pre-existing medical conditions, as well as in the severity and pattern of trauma which influences the further clinical course. Although an observation time of 72 h in polytraumatised large animal models can hardly be found in the literature, long-term observation periods must be performed to further illuminate post-traumatic changes in PT.

To the best of our knowledge, this is the first study to characterise and compare chronological data on the pro-coagulant molecules PAI-1 and TAT in relation to PT and two different surgical treatment strategies. The results of this clinically realistic porcine model present new insights into early post-traumatic coagulopathic reactions. Both PAI-1 and TAT showed a significant systemic response to ETC in PT, suggesting that ETC may worsen the coagulation status of the severely injured patient. The surgical strategy used should therefore be chosen thoughtfully, and ongoing studies should focus on PAI-1 and TAT as potential contributors to CD in severely injured patients.

The authors thank Thaddeus Stopinski, Benedikt Schopf, and Anna‐Lena Hauser for their assistance in conducting the experiments and animal care.

The animal study protocol was approved by the Institutional Review Board of Ministry for Nature, Environment and Consumer Protection in North Rhine-Westphalia, Recklinghausen, Germany (AZ 81.02.04. 2020.A215).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

This research was funded by the German Research Foundation (Grant No. 429837092).

Ü.M., M.H.-L., M.G., K.H., and F.H. designed the animal study. Ü.M., R.V.M.G., and K.H. performed the animal study. Ü.M., M.A.M., Z.H., and J.G. collected the data. Ü.M., Z.H., and T.E.M. performed the experiments. Ü.M., K.H., E.R.B., and F.H. analysed and interpreted the data and drafted the manuscript. All the authors critically revised the manuscript and approved its final version.

All data generated during this study are included in this article. Further enquiries can be directed to the corresponding author.