A novel rat model of tibial fracture for trauma researches: a combination of different types of fractures and soft tissue injuries

The incidence of open tibial fracture as a part of isolated injury or polytrauma is on the rise due to increase in the incidence of motor vehicle accidents [1, 2]. The outcomes for severe open fractures are still a great challenge for all the trauma surgeons in the treatment [3, 4]. Especially, the soft tissue injury in open fractures still leads to various complications, such as infection, muscle necrosis, ischemia, and even limb loss [5, 6]. And it is generally accepted that the outcomes of fractures depended on not only the fracture itself but also the combined soft tissue injury. Thus, it is very critical to develop an animal model for open fracture research with appropriate consideration of soft tissue injury in fractures.

In previous studies, it was shown that the closed fracture was made by an open osteotomy through elimination of the added variable of local wound healing. Since then, a more standard closed fracture was created by making the femur “prepinned” with an intramedullary wire and subsequently fractured with a blunt guillotine [7‐9]. However, existing open tibial fracture models can only provide minimal soft tissue injury, which leaded to the limited application in basic researches. This kind of minimal soft tissue injury in the existing fracture models was made according to the anatomical characteristics step-by-step instead of being caused by high energy. These models which could not mimic the clinic fractures which always occurred with corresponding soft tissue injuries in seconds are less applicable to trauma researches. Naturally, we purpose that a more appropriate rat fracture model which can mimic different fractures combined with soft tissue injuries will be a valuable supplement to the existing fracture models for trauma researches.

To meet these requirements, we made a modification to the traditional methods and designed a simple and adjustable apparatus with the buffer disc settings which could change the crash time of the hammer onto the rat’s legs and ultimately change the energy transferred to the involved extremity. These buffer discs made it easy and repeatable to provide various typical tibial fractures combined with soft tissue injuries.

In this experimental study, we aimed to investigate a novel tibial fracture model providing different fractures combined with soft tissue injury for better application in trauma research.

Open fractures have drawn considerable attentions recently [6]. Fractures of the tibial diaphysis are the most common long bone fracture, and approximately 24% of these fractures are open [18]. The soft tissue injury is one of the most complex problems due to its diversity in diagnosis and management of open fractures [19]. The high-energy nature of most of these fractures contributes to the increased proportion of Gustilo type III. In the epidemiologic study, Court-Brown et al. found that nearly 60% of open tibial shaft fractures were Gustilo type III [20]. Thus, a more appropriate rat fracture model which could mimic different fractures combined with soft tissue injuries will be valuable in trauma researches.

There are several different open fracture rat models modified from the classical model performing transverse osteotomy combined with skin incision and resections of muscle around the fracture site [10, 21]. However, most models provided closed or open fractures with only minimal soft tissue injury which was not caused by high energy, and these models were made according to the anatomical characteristics step-by-step and could not mimic the clinic fractures which always occurred with corresponding soft tissue injuries in seconds. But the pathological changes in an open fracture are too complicated to be mimicked by this simple method because soft tissues consist of several complex structures, such as skin, muscles, and vessels. This main flaw leaded to the limited application of these models in trauma researches. So compared with these existing models, our novel approach would have been more impactful in trauma research because we could perform different fractures combined with soft tissue injuries in a rat tibial fracture model with high reproducibility.

In our experiment, we made a modification and designed a simple and adjustable apparatus with buffer disc settings. A high-speed photography system was used in the pre-experiment to explore the appropriate buffer disc setting parameters. We tried different buffer disc setting parameters including 1 mm to 15 mm, evaluated the fracture types by X-ray immediately, and assessed the associated soft tissue injury by two senior orthopedic surgeons. The results indicated that when the buffer disc setting was settled at 3 mm, the most common occurred fractures would be closed fractures with minimal soft tissue injury. And more severe fractures with moderate soft tissue injury would be expected at 10 mm. Furthermore, the most severe fracture combined with extensive lacerations, severe muscle avulsion, and even vascular injuries would be expected at 15 mm.

After analyzing the result of pre-experiment comprehensively, the 3 mm, 10 mm, and 15 mm were purposed to provide three types of fractures combined with soft tissue injuries based on concepts of Gustilo classification [13‐15]. Type I are simple closed fractures with only mild ecchymosis, type II are open and moderate comminuted fractures with skin laceration and slight muscle rupture but without obvious ischemia, and type III are more severe comminuted fractures with segmental bone loss and severe muscle avulsion and ischemia in the distal extremity [14, 15]. The types of fractures combined with soft tissue injuries in our study were evaluated by this modified fracture categories.

We evaluated the fracture patterns by obtaining radiographic image of rats’ involved limbs at 6 h after fracture. The results of X-ray and μCT indicated that different fractures combined with soft tissue injuries were successfully provided by our apparatus. The anteroposterior (AP) radiographs of tibias showed that a simple transverse fracture happened in group 2, a moderate comminuted fracture with small segments happened in group 3, and a severe comminuted fracture with large gone segments occurred in group 4 (Fig. 3). The fracture distribution showed no fractures occurred in group 1, while 36 type I fractures in group 2, 34 type II and 2 type I fractures in group 3, and 36 type III fractures in group 4(Table 1). These results confirmed that fracture distributions differed between groups which kept in line with different buffer disc settings and fracture categories.

Vascular injury was another predominate factor in outcomes of open fractures [22, 23]. However, it was rarely described in existing fracture models. CTA is a unique radiographic method that demonstrates the arterial vasculature structures through 3D volumetric reconstruction which could rapidly detect vascular injuries [24]. In our study, CTA results showed moderate and partial vascular injuries in group 3 (2/36) and obvious vascular injuries in group 4 (36/36) (Fig. 3). We contributed these results to the buffer disc settings which changed the crash range of the blade and subsequently changed the severity of injuries. What is more, rats in group 4 suffered obvious vascular injuries according to the result of CTA, and some of them had partial necrosis of toes at day 7 postoperative (Fig. 3). However, no total necrosis or ischemia of the leg was found in any groups. This suggested that in addition to fracture patterns, soft tissue injury may also have a profound effect on the outcome of trauma. It may be important to take soft tissue injury into account when we perform basic researches on trauma. Our novel model was more useful to meet these requirements and provided a new selection for trauma researches. And the results of weight monitoring showed that rats suffering the most severe fracture and soft tissue injury had the obvious malnutrition in group 4 (Additional file 1: Figure S1). As described in the previous study, there was a high prevalence of malnutrition among the trauma patients [25]. In our experiment, the characteristics of weight change in all groups showed that our model could effectively provide different fracture types which mimicked the clinical cases.

Recent investigations have demonstrated that inflammatory responses followed by infiltration of inflammatory cells and release of cytokines would occur after fractures [16]. In our study, histology and ELISA were performed to evaluate the inflammatory cell infiltration, necrosis, and serum cytokines after fracture. The results of HE stains, Masson stain, and caspase-3 demonstrated different levels and durations of infiltration between groups. Especially, the most severe inflammatory cell infiltration and the longest duration of infiltration occurred in group 4. These results indicated the severity and duration of inflammation depended on the severity of fractures which has been demonstrated in other investigations [26]. Additionally, ELISA analysis of inflammatory cytokines was performed to evaluate the systemic inflammation after fractures. Skeletal and tissue injuries may cause a hyper-inflammatory reaction of the immune system manifested by elevation in levels of pro-inflammatory cytokines, and the massive secretion of pro-inflammatory cytokines usually induces upregulation of anti-inflammatory cytokines such as IL-4 and regulatory cytokines such as transforming growth factor-beta (TGF-beta) and IL-10, which results in a decrease of the severity of the inflammatory reaction [27, 28]. In our study, compared with group 1 (control group), the levels of the pro-inflammatory cytokine TNF-α and IL-1β in groups 3 and 4 were significantly higher than those in groups 1 and 2. Additionally, the corresponding duration of these pro-inflammatory cytokines were the longest in group 4. The anti-inflammatory cytokine level in rats showed that TGF-β levels were significantly higher in surgery groups than in group 1 in the later stage, and IL-10 levels were significantly higher in surgery groups than in group 1 throughout the entire process. The increase in pro-inflammatory cytokines is evident soon after injuries (6 h) and lasts usually for 24–48 h in most cases, and the duration of this reaction depends on the severity of the trauma and the basic state of the patient [27, 28]. In our experiment, the inflammatory cytokines in rats’ blood samples showed the similar trend as that in most clinical cases. This inflammatory change also showed that we performed this novel tibial fracture model providing different fractures combined with soft tissue injury successfully.

There are several advantages in our model. Firstly, various fractures combined with soft tissue injuries could be successfully provided with indicated buffer disc settings. Our novel model could provide different soft tissue injuries which mimicked the clinical trauma cases and assess the severity of soft tissue injuries in multiple methods. Secondly, three fracture categories which derived from the concept of OTA and Gustilo classifications could be created in our model with high repeatability. These fracture types in our model would reflect a similar progression of severity as the Gustilo types, but not a radical copy of them. We contribute it to the design of the buffer disc settings which can be manipulated easily before operation and change the crash range of the blade onto the involved leg and lead to the indirect change of the fracture and soft tissue injury types. Additionally, to our knowledge, it is the first application of CTA in the evaluation of vascular injuries in open tibial fracture models. Furthermore, this novel model could be used in researches focused on the cross-talk between bone union and inflammatory microenvironments which could contribute to the nonunion of fractures.

There are several limitations of this study. First, the study is based on the short-term observation of fractures without long-term follow-up to assess bone union and functional recovery. Second, buffer disc setting parameters need more tests before its application in rats with different weight and diameter of the extremities or in other animals. What is more, vascular injuries in cases of open fracture are caused by external injury as well as internal injury due to fracture. But as it was difficult to perform a standardized and accurate assessment of internal injuries in a rat fracture model, we used CTA to assess different degrees of vascular injuries only caused by external injury in our experimental. Fourthly, body weight, blood loss, and fluids can have profound impacts both on injuries and inflammatory responses. In our experiment, we monitored the body weight preoperatively and postoperatively but missed monitoring and controlling the blood loss and fluids which were difficult to perform standardly in rats. Fifthly, wire fixation extending outside of the knee joint could have induced soft tissue injury itself. Although we harvested the tibialis anterior muscle near the fracture site for histological analysis, the wire may have an effect on levels of inflammatory cytokines in the peripheral blood. Additionally, we did not give any antibiotics to the rats after fractures. Antibiotics [29] given according to the bacteria in open fracture in clinic played a critical role in the inflammatory response.

The novel apparatus was valuable in performing different fractures combined with soft tissue injuries in a rat tibial fracture model with high reproducibility. Therefore, this novel rat model of tibial fracture is effective and provides a new selection in trauma research.