In multiply injured patients the blunt chest trauma represents one of the most critical injuries and is regarded as an important trigger of the posttraumatic systemic inflammatory response occurring after severe trauma.1–3 It was demonstrated that a thoracic trauma induces several systemic effects, such as a rapid release of pro-inflammatory cytokines [e.g., tumor necrosis factor (TNF)-α, interleukin (IL)-6] and prostanoids, as well as an activation of the coagulation and complement systems.4–6 The systemic increase of inflammatory mediators was shown to be much higher after a blunt chest trauma in comparison to other injury patterns, indicating its important impact on the systemic inflammatory response in poly-traumatic patients.7 The severe systemic inflammation can impair the function of cells in tissues not initially directly affected by the trauma. For instance, it was shown in a mouse model of blunt chest trauma that peritoneal- and splenic macrophages as well as splenocytes revealed severe immune dysfunction.8–10 An excessive inflammatory response can even lead to a multi-organ dysfunction syndrome (MODS).11
Approximately 50% of patients with a blunt thoracic trauma are additionally affected by fractures of the extremities.12 There is strong clinical evidence that fracture healing is delayed in such patients. Bhandari et al. reported that multiply injured patients suffering from tibial fractures exhibit an approximately three times higher risk of a second surgical intervention compared to patients with an isolated fracture. The main cause of reoperation were non-unions due to disturbed bone healing.13 The reasons for impaired bone regeneration after severe trauma are unknown, but it can be assumed that the early posttraumatic systemic inflammatory response with the strong increase in pro-inflammatory cytokines could influence the fine local inflammatory balance of the bone healing process. This assumption is supported by the observation that a systemic inflammation induced by the administration of lipopolysaccharide in a rat model of experimental sepsis delayed fracture healing.14 However, despite the clinical relevance, no experimental study has so far investigated the influence of a blunt chest trauma on fracture healing.
It is well known that fracture healing is significantly influenced by local biomechanical factors and that a too flexible fracture fixation leads to delayed union or even a non-union compared to a more rigid fixation.15–17 The biomechanical environment already influences the early stages of bone healing for instance by affecting vascularization18 or macrophage immigration into the fracture callus during the inflammatory phase of fracture healing.19 It remains to be investigated whether the early posttraumatic systemic inflammation interacts with the local inflammatory processes and differentially influences regular or delayed bone healing under more rigid or more flexible biomechanical conditions, respectively.
To address these questions we investigated the influence of a blunt chest trauma on fracture healing in a rat osteotomy model under proper or more flexible fixation conditions using an adjustable external fixator. We hypothesized that the systemic inflammatory response induced by the thoracic trauma would delay regular bone healing under proper biomechanical conditions. Furthermore, we postulated that the blunt thoracic trauma would further increase the delay of bone healing induced by mechanical instability.
The animal experiment was performed according to international regulations for the care and use of laboratory animals and approved by the local ethical committee (Regierungspräsidium Tübingen, Germany). Thirty-three male Wistar rats (weight 400–450 g) were randomly divided into a group receiving a blunt chest trauma and a femur osteotomy, which was stabilized with an external fixator, and a group receiving solely the stabilized osteotomy. Each group was subdivided in a more rigidly or more flexibly fixated group (n = 8–9). Separate animals, which received the flexibly fixated osteotomy with/without thoracic trauma, were used for blood withdrawals in order to measure the early inflammatory response at 0, 6, 24, and 72 h (n = 7–8).
Surgery and Blunt Chest Trauma
Surgery was performed as described previously.20 Briefly, the rats were anesthetized with 2% isoflurane (Forene®, Abbott, Wiesbaden, Germany). To avoid wound infection, the rats received daily subcutaneous injections of clindamycin-2-dihydrogenphosphate (45 mg/kg, Sobelin®, Pfizer GmbH, Karlsruhe, Germany) until the third postoperative day. A custom-made external fixator was attached to the right femur by four threaded stainless steel pins (Jagel Medizintechnik, Bad Blankenburg, Germany).21 An offset of the fixator block of either 6 mm (more rigid) or 12 mm (more flexible) was chosen resulting in an axial stiffness of either 119 or 31 N/mm, respectively. Subsequently, an osteotomy gap of 1 mm was created between the two inner pins at the mid-shaft of the femur. According to a computational musculoskeletal model,22 it can be assumed that the axial component of the internal force in the rat femur could reach up to six times body weight, probably resulting in an interfragmentary strain (IFS) of 21% with the more rigid and 82% with the more flexible fixator. Therefore, callus healing can be expected for both fixation conditions.16 Immediately after surgery half of the rats received an additional blunt chest trauma under general anesthesia using a blast wave generator as previously described in detail.5, 6 This model allows a bilateral, isolated lung contusion by the application of a standardized single blast wave centered on the middle of the thorax. An analgesic (20 mg/kg, Tramal®, Gruenenthal GmbH, Aachen, Germany) was administered subcutaneously during the operation and was diluted in the drinking water (25 mg/L) for the first 3 days following surgery. Each animal was housed in its own cage, given unrestricted access to food and monitored daily for infection and mobility.
Analysis of Inflammatory Cytokines in the Serum
Blood withdrawal was performed under general anesthesia by taking 0.2 ml blood from the lateral tail vein. Blood was collected in serum microvettes (Sarstedt AG & Co., Nümbrecht, Germany). After storing the blood for 1 h at room temperature to accelerate clotting it was centrifuged at 1 000g for 10 min. The serum was collected and frozen at −80°C until further investigation. Numerous ILs (IL-1α, IL-1β, IL-2, IL-4, IL-6, and IL-10), granulocyte-macrophage colony-stimulating factor (GM-CSF), interferon γ (IFN-γ), and TNF-α (TNF-α) were measured using an enzyme linked immunoassay (Bio-Plex Suspension Array System, Bio-Rad Laboratories, Hercules, CA) according to the manufacturer's protocols. Levels below the detection limit of the assay were set to zero.
Monitoring of Body Weight, Ground Reaction Force, and Motion
Three days prior to surgery and 2, 7, 14, 21, 28, and 34 days postoperatively, the bodyweight and the ground reaction force of the operated limb of each rat were measured. For measurement of the ground reaction force, the rats were allowed to move freely through an acrylic glass tunnel containing a force plate (HE6×6, Watertown, MA). The peak vertical ground reaction force during normal gait of the operated limb was averaged from five loads for each rat. The postoperative ground reaction forces were related to the preoperative values. On the same days, the activity of each rat was recorded during the 12 h night cycle using an infrared beam detection system fitted to each cage (ActimotMot-System, TSE Systems GmbH, Bad Homburg, Germany). Movement was quantified by registering counts when a light beam was interrupted. All values were related to preoperative measurements.
After 35 days the rats were sacrificed and the operated and contra-lateral femurs were explanted. The fixators were removed. Because of fractures in regions of the proximal pin, two animals in the more rigidly fixated group without blunt chest trauma had to be excluded. The flexural rigidity of the femurs was evaluated using a non-destructive, three-point bending test. The femurs were potted with their distal end in cylinders using polymethylmethacrylate (Technovit® 3040, Heraeus Kulzer GmbH, Wertheim, Germany). The cylinder itself was fixed in a hinge joint, serving as the proximal support for the bending test, whereas the head of the femur rested with the major trochanter on the bending support so that a 30 mm free length (l) between the bending supports for the bone remained. A quasistatic load was applied in a three-point bending mode with a materials testing machine (1454, Zwick GmbH, Ulm, Germany) using a 50 N load cell (A. S. T. Angewandte System-Technik GmbH, Dresden, Germany). The bending load F was applied on top of the callus tissue and continuously recorded versus sample deflection up to a maximum force of 10 N at a crosshead speed of 1 mm/min. Flexural rigidity EI was calculated from the slope k of the linear region of the load–deflection curve. Since the callus was not always located at the middle of the supports (l/2), the distances between the load vector and the proximal support (a) and the distal support (b), respectively, were considered for calculating the flexural rigidity according to (in N/mm2).
The femora were scanned using a microcomputed tomography (µCT) scanning device (Skyscan 1172, Kontich, Belgium) at a resolution of 30 µm, operating at a peak voltage of 50 kV and 200 µA. The total volume (TV), bone volume fraction (BV/TV), bone surface/total volume (BS/TV), and the maximum moment of inertia (Imax) were evaluated by segmentation of the former osteotomy gap using the CT analysis software (CTAnalyser, Skyscan, Kontich, Belgium). Global thresholding was performed to distinguish between mineralized and non-mineralized tissue. The gray value corresponding to 25% of X-ray attenuation of the cortical bone of each specimen was taken as the threshold.23
Values of ground reaction force, activity, and serum IL-6 values are shown as mean ± standard error. Values of bending rigidity and µCT-values are presented as median and interquartile ranges. The statistics software PASW Statistics 18.0 (SPSS, Inc., Chicago, IL) was used. The time courses of the ground reaction force and the activity were compared using the Russell's error factor.24 Flexural rigidity, µCT parameters and serum IL-6 levels were compared using a one-sided Student's t-test. The level of significance was p < 0.05.
Body Weight, Ground Reaction Force, and Motion
All rats recovered quickly from the operation. The bodyweight did not change significantly during the healing period (results not shown). The activity of the animals was slightly reduced at the second postoperative day and returned to preoperative values after 1 week (Fig. 1A). The application of the blunt chest trauma had no significant influence on the activity nor did the fixation stability. The vertical ground reaction force decreased to approximately 50% of preoperative values after 2 days. Then the ground reaction force steadily increased up to the preoperative values. Again, the groups did not differ significantly (Fig. 1B).
Inflammatory Cytokines in the Serum
The blunt chest trauma led to a significant, approximately threefold increase of IL-6 in the serum after 6 and 24 h, and to an approximately twofold increase after 72 h, all compared to preoperative values (Fig. 2). In comparison to an isolated fracture, the thoracic trauma increased IL-6 levels significantly after 6 and 24 h, whereas they no longer differed after 72 h. Rats with an isolated fracture did not show significant differences in IL-6 values at any time point compared to preoperative values. IL-10 and INF-γ serum concentration did not reveal any differences between the groups (data not shown). IL-1α, IL-1β, IL-2, IL-4, GM-CSF, and TNF-α serum values were below the detection limit of the assay in all groups.
The blunt chest trauma significantly decreased flexural rigidity in the more rigidly fixated group by 63%, whereas it had no significant influence on the more flexibly fixated group (Fig. 3). In the rats without thoracic trauma, the more flexible fixation significantly decreased flexural rigidity by 53% in comparison to the more rigid fixation.
Under the more rigid fixation the blunt chest trauma significantly reduced TV of the callus by 20% (Fig. 4A), BS/TV by 24% (Fig. 4C) and Imax by 50% (Fig. 4D). BV/TV was not significantly influenced by the thoracic trauma in the more rigidly fixated group (Fig. 4). Under the more flexible fixation the blunt chest trauma did not significantly influence any of the evaluated parameters. In animals without thoracic trauma, a more flexible fixation increased TV significantly by 34% (Fig. 4A), but decreased BV/TV significantly by 18% in comparison to the more rigid fixation (Fig. 4B).
Confirming the clinical evidence13 this study reports for the first time that a blunt chest trauma considerably delayed bone healing in an experimental trauma model. We also investigated the hypothesis, whether a blunt thoracic trauma would further increase the delay of bone healing induced by mechanical instability. As expected, the more flexible fixation impaired healing in comparison to the more rigid fixation, whereas the thoracic trauma had no additional adverse effect under more flexible experimental conditions.
Under the more rigid fixation, the blunt thoracic trauma impaired fracture healing considerably in comparison to rats with an isolated fracture. This was confirmed by a significantly reduced bending stiffness as well as a smaller callus volume, a decreased moment of inertia and a reduced relative bone surface, indicating inferior callus quality. To exclude the observed impairment of fracture healing in the severely injured rats as being a result of lower activity and reduced limb loading25 we monitored animal movement and ground reaction forces during the postoperative period. We could demonstrate that all rats recovered quickly from the thoracic trauma and loaded the operated limb properly with no significant differences to the groups with isolated fractures. Therefore, it can be implied that the observed impairment of fracture healing did not result from different mechanical stimulation of the callus but from the systemic inflammation induced by the blunt chest trauma.
In this study we used a well-established blunt chest trauma model in rats, inducing a reproducible, bilateral contusion of the lung via a single blast wave on the thorax.5, 6, 10 Confirming former experimental studies6, 9 the blunt thoracic trauma induced a considerable, rapid, and transient increase of the serum levels of the pro-inflammatory cytokine IL-6 during the first 24 h, whereas the isolated fracture did not significantly influence the serum level of this cytokine. In humans an increased IL-6 level is regarded as an early important marker for the systemic inflammatory response after chest trauma.2, 4 Furthermore, IL-6 showed the best correlation with injury severity and mortality compared to other cytokines.11, 26 The levels of other cytokines, which were shown to be increased in humans after severe trauma (e.g., TNF-α and IL-10),11 were not elevated at the investigated time points in our study. However, Flierl et al.5 reported a very early increase of IL-10 and TNF-α in the same rat model. The results might differ due to different investigation time points or evaluation procedures. Corresponding to the human situation, Flierl et al.5 also demonstrated a rapid and transient activation of the complement system, a crucial part of the innate immunity, in the rat thoracic trauma model. From these data and our own results we propose that the trauma model used induced, similar to the human situation, a complex systemic inflammatory response involving a systemic cytokine increase and complement activation.
The systemic inflammatory response could influence bone healing by multiple potential mechanisms. The rapid but transient increase of inflammatory cytokines, for example, IL-6, IL-10, and TNF-α, are likely to disturb the fine inflammatory balance in the early phase of fracture healing. The initial inflammatory phase of bone healing is mainly characterized by the invasion of macrophages, polymorphonuclear leukocytes, and lymphocytes into the fracture site. These early inflammatory processes are regulated by pro-inflammatory cytokines released by the immune cells, for example, IL-6, IL-1, and TNF-α.27, 28 These cytokines have multiple concentration dependent effects and could for example increase extracellular matrix synthesis, stimulate angiogenesis and recruit mesenchymal precursor cells to the injury site, as well as regulate osteoclasts by stimulating their resorption activity and osteoclastogenesis from hematopoetic precursors.27, 29, 30 The posttraumatic systemic increase of cytokines, such as IL-6, IL-10, and TNF-α could possibly influence these finely tuned processes by altering the local recruitment and activation of immune cells, for example, macrophages, enhancing or prolonging the inflammatory reaction at the fracture site and finally resulting in delayed healing. It has already been shown that a local activation of macrophages could induce an immature callus with reduced biomechanical characteristics.31 Increased systemic concentrations of pro-inflammatory cytokines could also enhance the number and activity of osteoclasts,30, 32 which already play an important role in the resorption process during the early phases of fracture healing.33 The blunt thoracic trauma also leads to an activation of the complement cascade.5 Its function is the opsonization of antigens, the support of phagozytosis and the induction and modulation of the inflammatory reaction of immune cells.34, 35 There is also evidence that complement might play a role in fracture healing. Recently, we could demonstrate that a key receptor, C5aR, is expressed by inflammatory cells, osteoblasts and chondrocytes in the fracture callus, indicating that these cells are target cells for activated complement during the healing process.36
Even though systemic inflammation was transiently present during the very early phase of fracture healing it was able to disturb the complex fracture repair sequence resulting in impaired healing after 35 days. Which molecular and cellular mechanisms were responsible for the observed impairment of bone regeneration remain to be clarified and need to be further investigated by the evaluation of earlier healing time points.
As expected, the more flexible fixation delayed healing in the present study, which was reflected by a larger but qualitatively inferior callus. This is well known and confirms previous work.37 However, under the mechanically induced compromised healing conditions the blunt chest trauma had no additional adverse effect. It has been shown that unstable mechanical conditions already influenced the early inflammatory phase of fracture healing by modulating the immigration and activity of macrophages.19 Possibly, the systemic inflammation did not provoke an additional effect in this already stimulated inflammatory environment. However, a limitation of our study could be that an influence of the systemic inflammation was possibly not detectable in the experimental setup used. The flexible fixator used in the present study decreased healing considerably. After 35 days the flexural rigidity of the more flexibly fixated callus was still rather low, revealing only 25% of the intact bone. It has been previously shown that small differences between treatments can barely be detected at a healing stage with low flexural rigidity.38 Possible solutions would be to slightly increase the stiffness of the more flexible fixator or to investigate later healing periods.
In conclusion, our results demonstrate that a blunt thoracic trauma significantly impaired fracture healing. Ongoing studies of our group are focusing on the underlying molecular and cellular mechanisms, and include earlier investigation periods. These findings are of interest for the development of therapeutic strategies for the treatment of poly-traumatic patients.
This study was funded by the German Research Foundation (KFO 200). The authors appreciate the technical assistance of Ursula Maile and Marion Tomo. Each author in this manuscript has not and will not receive benefits in any form from a commercial party related directly or indirectly to the content of this manuscript. None of the authors have any conflicts of interest.