A central epiphyseal plate injury model with a diameter of 2 mm was created, similar to Salter-Harris type IV epiphyseal plate fracture, using the above modeling method (Fig. 2). However, the articular cartilage remained intact; thus, the drug could be confined to the injured site, which is conducive to evaluating the efficacy of local drug treatment. Moreover, the silicone gel membrane on the surface of the holes further ensured that the drug was limited to the injured site. A diagram of the tibial growth plate drilling injury model is shown in Fig. 3. The evaluation methods for the above animal modeling results included digital radiography (DR), micro-computed tomography (micro-CT), hematoxylin and eosin (H&E) staining, and safranin fast green staining.
DR analysis was performed on the experimental rats under isoflurane anesthesia to evaluate the length of the operated limb and repair of epiphyseal plate injury. When NS was injected into the epiphyseal plate injury hole, no obvious growth was observed in the operated limb 8 weeks after surgery. The epiphyseal plate closed at 8 weeks, indicating that the proximal tibial epiphyseal plate injury affected the longitudinal growth of the tibia (Fig. 4A). After the injection of IL-1β into the hole, there was no obvious growth of the limb 8 weeks after the operation, the epiphyseal plate was completely closed at 4 weeks, and the proximal tibia was deformed (Fig. 4B). IL-1β promoted the formation of bone bridges after injury to the epiphyseal plate and shortened and thickened the deformities' appearance. DR could be performed without removing the skin and muscle and was used to continuously observe experimental rats at different time points in vivo. However, it could not display the local image and was not used to calculate the bone content or other values.
A previous study showed that bone bridge formation began 7 days after epiphyseal plate injury and was complete 28 days after injury [31]. At different time points (immediately after, 1 week after, and 8 weeks after the operation), the rats were euthanized using excessive carbon dioxide. The left hind legs of the rats were completely removed, the skin, subcutaneous soft tissue, and muscle were removed, and the bone tissue was immersed in paraformaldehyde for internal fixation and stored at 4°C for further use. The proximal tibia of the defect and adjacent distal femur were scanned at a 20 µm voxel size on a Skyscan1276 system (Bruker, Belgium) with a source voltage of 70 kV, source current of 200 µA, and exposure time of 426 ms. According to experimental requirements, the parameters of the growth plate in the epiphyseal plate injury area could be measured, such as tissue volume, bone volume, percent volume, trabecular thickness, trabecular number, and trabecular separation. Immediately after the operation, the proximal tibia and distal femur of the surgical side were photographed using micro-CT, and three-dimensional pictures were reconstructed, which showed that the model was successful. The hole was located in the middle of the epiphyseal plate, and the articular cartilage was intact (Fig. 5). The hole depth was measured from the micro-CT image immediately after the operation, and the volume within the hole was calculated. The hole size was compared to the length and volume of the hole calculated during the operation (Table 2) to determine the maximum volume that could be administered. The hole depth and volume measured using micro-CT were larger than those measured during the operation. The maximum drug volume was based on the data measured from micro-CT images showing no obvious bone bridge formation 1 week after the operation. The width of the epiphyseal plate in the IL-1β group was slightly lower than that in the NS group, and bone hyperplasia appeared at the edge of the injury (Fig. 6A). Micro-CT examination of the experimental rats 8 weeks after the operation showed that the bone bridge was formed in the NS injection group, the epiphyseal plate was significantly narrowed around the bone bridge, the structure of the tibial plateau was still intact, and the deformity was not obvious. In the IL-1β injection group, a bone bridge was formed, the epiphyseal plate was narrowed, a diaphyseal cavity was formed, the shape of the tibial plateau was disordered, and the proximal tibia was thick and deformed (Fig. 7A).
Table 2
Parameters of the hole (mean ± standard deviation)
|
NS
|
IL-1β
|
Hole depth (operation)(mm)
|
4.103 ± 0.078
|
4.182 ± 0.048
|
Hole volume (operation)(mm3)
|
12.880 ± 0.246
|
13.130 ± 0.152
|
Hole depth (micro-CT)(mm)
|
6.000 ± 0.010*
|
5.990 ± 0.068*
|
Hole volume (micro-CT)(mm3)
|
21.420 ± 0.646*
|
20.780 ± 1.012*
|
The hole depth and volume measured using micro−CT were larger than those measured during the operation.
*p<0.05, compared with parameters measured during the operation.
Although the angle and depth of the bone drill were limited by measurement and calculation, there were some failures in the modeling. For example, if the bone drill depth was large, the integrity of the articular cartilage would be destroyed, and the articular cartilage would be broken into the knee joint (Fig. 8A), and if the angle were too large, the bone drill would deviate to the fibular side (Fig. 8D). Although these failed modeling cases also showed formation of bone bridges, it was impossible to ensure the amount of drug left in the hole if the drilling was too deep into the knee joint. If the drilling angle was too large at the side of the fibula, the bone bridge formed was biased at one side, affecting the subsequent statistics of the limb shortening and angulation deformity.
The experimental rats were euthanized by excessive carbon dioxide inhalation at each time point, according to the experimental plan. The proximal tibia of the surgical side was harvested, and the soft tissue was removed, quickly placed in paraformaldehyde, and stored at 4°C for use. Ethylenediaminetetraacetic acid was used for decalcification 3 or 5 days later, and the decalcification time was approximately 15–30 days. When changing the decalcification solution, attention had to be paid to lightly pinch the tissue specimen with a syringe needle and ensure that the specimen was decalcified entirely and did not decalcify excessively. Decalcified tissue specimens were embedded in paraffin and sectioned (5 µm). H&E staining and safranin O-fast green staining were performed to observe structural tissue changes and evaluate the formation of bone bridges. One week after the epiphyseal plate drilling injury, it was found that the structure of the growth plate was abnormal around the injury site, the layer of epiphyseal plate chondrocytes was not prominent, bone trabeculae began to appear, and inflammatory cells infiltrated the hole. In the IL-1β injection group, there was increased infiltration of inflammatory cells, the proliferation area of the growth plate increased, the proliferation area became wider, the hypertrophic area became disordered, and the matrix structure was damaged. The bone trabeculae were more densely distributed (Fig. 6B-E). Eight weeks after the injury, staining of the proximal tibia showed that the thickness of the growth plate around the injury was decreased, the bone trabeculae ran through the whole growth plate, and there was no cartilage structure in the local area. In the IL-1β injection group, a small amount of inflammatory factor infiltration was observed in the injured area, the structure of local epiphyseal plate chondrocytes was disordered, the matrix structure was completely damaged, the bone trabeculae were dense, and the bone bridge was completely formed 8 weeks after the injury, entering the shaping phase (Fig. 8B-E). The inflammatory factor IL-1β promoted the local inflammatory response in epiphyseal plate injury, disrupted the repair of growth plate injury, destroyed the structure of the growth plate, caused damage to the matrix structure, and promoted osteogenesis. Therefore, reducing the release of inflammatory factors in the injured area is a potential therapeutic strategy for preventing the formation of bone bridges after epiphyseal plate injury.
H&E and safranin fast green staining were also performed on the proximal tibia of the failed model. The results showed that over-deep drilling of the articular cartilage could also cause inflammatory cell infiltration in the injured part of the growth plate, a structural disorder of the growth plate, abnormal structures of the hypertrophic and proliferative areas, damage to the matrix structure, insertion of bone trabeculae into the injured part, and formation of bone bridges. At the same time, it was found that the injured articular cartilage had local abnormal hyperplasia and tibial plateau deformation (Fig. 8B, C). Similarly, the above results were observed for large drill angles on the fibular side (Fig. 8E, F). However, because of the angle problem, the uniformity of the experimental results could not be guaranteed; thus, it was also considered a failed model.