Spontaneous correction of fracture deformity : a study in the rat

Sammanfattning: The aim of the present study was to elaborate a model in the rat permitting analysis of bone modeling to correct fracture deformity. Among conceivable regulators of this process, the study focused on the peripheral nervous system by analyzing markers of regenerating (GAP-43), mature (PGP 9.5) and sensory (CGRP) nerve fibers at the fracture site. CGRP has been suggested to be a potent promotor of bone formation through stimulation of osteoblasts and inhibition of osteoclasts. A total of 113 male Sprague-Dawley rats were divided into three groups with tibial fracture and one control group with intact tibia. Mid-shaft fracture was produced by three-point manual bending and fixed with an intramedullary nail in either straight alignment, posterior (31 degrees) or anterior angulation (27 degrees), thereby creating a convex and a concave side of the diaphyseal fracture in the two latter groups. Fracture heating and modeling were assessed by serial radiography, histology, immunohistochemistry and bone mineral uptake studies. During the experimental period of 98 days the rats were allowed free weight bearing. Radiography showed that the fracture callus rapidly became greater on the concave side compared to the convex in both the anterior and posterior angulation group. The callus thickness reached a peak at day 21 post fracture and from day 21 to 35 it decreased significantly on the convex side reflecting resorption. Complete fracture healing was observed at day 35. Analysis of the fracture angulation over time showed that it was corrected from 27 to 11 degreesmost of which was achieved already during the first 21 days after fracture. The adaptive changes in the proximal growth plate resulted in an 8 degrees correctional tilt of the epiphysis. In contrast to previous reports, our findings clearly reflect that most spontaneous correction of angular fracture deformity occurs in the diaphysis instead of the growth plate and notably, quite early during healing, not afterwards. This was confirmed by bone mineral uptake studies showing intense mineralization, i.e., bone formation, from day 7 to 35 on the concave side of the fracture, while on the convex side bone resorption predominated. Site-specific differences in local bone turnover, e.g., to correct angular fracture deformity, may prove to be orchestrated by the peripheral nervous system. Therefore, neuronal immunohistochemistry was applied to the tissues. Fracture of tibia was found to elicit a tremendous increase in nerve immunoreactivity. GAP-43 and CGRP positive nerve fibers were detected already on day 3 in the fracture hematoma, between day 7 and 14 also in avascular fibrocartilaginous callus and woven bone. The first PGP 9.5 fibers were seen on day 7. Semi-quantitative analysis showed a peak in neuronal CGRP immunoreactivity on day 21 representing a 38 fold increase in anterior angulation fracture and 27 fold increase in straight alignment fracture as compared to intact tibia. Comparison of the two opposing sides of anterior angulation fracture, disclosed a significant increase in CGRP immunoreactivity on the concave side coinciding with the inflammatory and early modeling phase of healing. In conclusion, angular tibial fracture in the rat elicits site-specific differences in bone formation and bone resorption within the same bone. Correction of fracture deformity by modeling occurs early, concomitantly with healing. The model proposed offers new possibilities to analyse pertinent regulators of local bone turnover. The occurrence of the "bone promoting" neuropeptide CGRP is strongly increased in fracture healing and modeling. The overwhelming predominance of CGRP nerve fibers on the concave side compared to the convex side in angular fracture suggests an important role of the peripheral nervous system in correcting diaphyseal fracture deformity by modeling.