A total of 45 New Zealand white rabbits (age, 2 months; 24 females and 21 males; weight, 2.5±3.0 kg) were obtained from the Zunyi Medical College Animal Experiment Center (Zunyi, China). Animals were housed at 15°C and 45% humidity, with ad libitum access to standard rabbit food and water. The experiments were conducted at the Central Laboratory of Zunyi Medical College. The study design involved randomized grouping of animals and in vivo experiments. The study was approved by the Ethics Committee of Zunyi Medical College.

The SF/CS/nHA scafford was prepared as previously described (10). SF, CS and nHA were prepared into 2% solutions and then combined in three proportions, namely 1:1:0.5, 1:1:1 and 1:1:1.5. The mixture was stirred at 55°C using a magnetic stirrer until the bubbles disappeared and the mixing was uniform. Next, the mixture was transferred to a 9-well plate using a 20 ml syringe (~50 ml/well) and frozen at −80°C for 24 h. The frozen scaffolds transformed from fluid into solid state but still contained ice, so were sealed with parafilm. A pattern of holes was made on the parafilm so that the water could be absorbed. The scaffolds were placed in a vacuum drier for 36 h. The scaffolds were then removed and soaked in a mixture of 75% methanol and 1 mol/l sodium hydroxide for 15 h. Subsequently, the scaffolds were repeatedly washed with ultra-pure water and dried again in the vacuum drier. The scaffolds were then placed for 10 h in a solution of crosslinking agents, containing 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide and N-hydroxysuccinimide. After rinsing, the wet composite scaffolds were frozen at −80°C for 24 h, followed by vacuum drying for 36 h. Finally, the prepared composite scaffolds were sealed for later use (11,12).

A total of 45 New Zealand white rabbits were fasted from food and water for 6 h before surgery. Anesthesia was induced by intravenous injection of 2.5% sodium pentobarbital via the ear margin. The rabbits were immobilized in the four limbs to prevent damage to the vessels or nerves caused by body twitching during surgery. Next, the right forelimb was exposed, the skin was disinfected conventionally with iodophor and alcohol, and incised from the midpoint of the radial bone. Dissection was performed layer by layer with caution in order not to damage the vessels or nerves. In the exposed radial bone, a defect of approximately 2 cm in length was performed using a sterilized saw blade, as previously described (13–15). After the radial bone defect was performed, rabbits were divided into three groups (n=15 each) and treated as follows: Group A, in which SF/CS/nHA composite scaffold was implanted; group B, in which SF/CS scaffold was implanted, serving as the negative control group; and group C, in which no scaffold was implanted, serving as the blank control group (16). The incision was sutured layer by layer and disinfected with iodophor following surgery, and wrapped with gauze without applying external fixation. To prevent postoperative infection, penicillin (500,000 U) and streptomycin (0.1 g) were injected intramuscularly. The rabbits were kept in separate cages subsequent to surgery.

The diet and mobility of rabbits were observed regularly subsequent to surgery, and the incisions were checked for reddening and infection. At 4, 6, 8 and 12 weeks after surgery, the rabbits were immobilized to obtain anteroposterior X-ray scans of the radial bone. Bony calluses and replacement of the scaffold by bone tissues were observed in X-ray scans. Rabbits were sacrificed and bone tissues were harvested at different times postoperatively (4, 6, 8 and 12 weeks after surgery, n=3 at each time point) and made into pathological specimens to evaluate the repair of bone defects, the scaffold association with the surrounding tissues and the absorption of the scaffold.

Vertical and transverse sections were prepared for a comprehensive observation. Briefly, the specimens were fixed with 4% formaldehyde, decalcified and sliced into 4-µm sections. Next, specimens were dewaxed in xylene, dehydrated through an ethanol gradient, washed with water and air dried for 24 h. Pathological observation was based on hematoxylin-eosin (HE) and toluidine blue staining. The sections were dehydrated, sealed and observed under an inverted microscope to evaluate the growth of normal bone tissues, osteogenic capacity and degradation of the scaffold. The lesions in the surrounding tissues were also observed. Lane-Sandhu histologic criteria were used (17) i) Healing: None, 0; fiber, 1; fibrocartilage, 2; bone, 3; trabecular and cortical bone, 4. ii) Callus: None, 0; small amount of callus, 1; medium callus, 2; large amount of callus, 3; fully callus, 4. iii) Bone marrow: No bone marrow, 0, bone marrow started, 1; >50% of the defect area has bone marrow, 2; complete marrow, 3; mature marrow, 4.

The results of rough observation, X-ray scan and pathological observation were analyzed using SPSS version 17.0 software (SPSS, Inc., Chicago, IL, USA). F-test was conducted to compare the mean values of multiple samples. Differences with P<0.05 were considered to be statistically significant.

Rabbits were provided with food 12 h after surgery in each group. The incisions healed within 1 week without severe infection, reddening or festering. At 1 week after surgery, the rabbits in groups A and B appeared to be more active compared with those in group C, while the mobility was moderate in all groups 3 weeks later. As indicated by pathological observation, the scaffolds in groups A and B at 4 weeks after surgery were enclosed in fibrous connective tissues. The scaffolds under the periosteum presented signs of calcification and a grayish white color. Little difference was observed between groups A and B with the naked eye. At 8 weeks, the scaffold was gradually replaced by bony calluses in groups A and B. The scaffolds of group B had a light yellow color. Between weeks 12 and 16, osseointegration with a similar color and texture as that of the normal bone tissues was observed in group A. In group B, the scaffolds had a light yellow color, similar to that of the cartilage. Although the broken ends of the bone presented slight elongation at the early stage in group C, no osseointegration was observed and the defect was filled with soft tissues during the entire observational period. Fig. 1 shows the radial bone defect of each group at 16 weeks after surgery.

The results of X-ray examination conducted in group A are shown in Fig. 2. At 4 weeks after surgery, the scaffolds in the defect region presented with a grayish white high-density shadow. At 8 weeks after surgery, the calcified shadow was darkened with distinct boundary from the surrounding soft tissues. At 12 weeks, the marrow cavity was unobstructed partially between the two broken ends. Finally, at 16 weeks, no evident disparity was observed on the X-ray scans between the defect region and the normal bone tissue, while the marrow cavity was completely unobstructed.

As shown in Fig. 3, sporadic calcified shadows were observed in the scaffolds in group B at 4 weeks after surgery. These were clearly darker compared with the adjacent soft tissue shadows, though their morphology was less distinct. At 8 weeks, the calcified shadows were enhanced, although they remained lighter in color compared with normal bone tissues and were more distinct in morphology in comparison with the earlier X-ray scan. At 12 and 16 weeks after surgery, X-ray scanning revealed slightly lighter bone density shadows compared with the normal bone tissues, and the marrow cavity was partially unobstructed.

As demonstrated in Fig. 4, new bone tissues grew from the two broken ends at 4 weeks in group C, however, the length of growth was small. At 8 weeks, the calcified shadows at the two broken ends were enhanced, and the growth was insignificant. In the blank region (defect), soft tissue shadows were observed. At 12 and 16 weeks after surgery, the elongation at the two broken ends appeared to have stopped, and the calcified shadows were enhanced to the extent of resembling the normal bone tissues. As the two broken ends were closed separately, bone nonunion occurred. Groups A and B both indicated osteogenesis in early stages, but in group A, higher X-ray bone density was observed, indicating that group A had a better capacity for osteogenesis. Group C finally became bone nonunion.

HE staining was performed in each group at 16 weeks. As seen from Fig. 5A, new bone tissues were observed in group A. Trabecular bones and long spindle-shaped osteocytes with deeply stained nuclei were observed under ×100 magnification. The bone plate was distinct and the central canal, while bone matrix in lamellar arrangement was detected on the outside and the osteocytes. Under higher magnification, long spindle-shaped osteocytes were observed. In group B, the scaffolds were partially replaced by bone tissues and the osteocytes were surrounded by a considerable amount of chondrocytes (Fig. 5B). However, no trabecular bones or bone plates were observed, and large amounts of chondrocytes were observed under high magnification (Fig. 5B). Furthermore, in group C, large amounts of fibrous connective tissues were detected in contrast to a few bone-like tissues. Round and deeply-stained nuclei of myofibroblasts were observed under high magnification, which were larger in size compared with the osteocytes and vacuolated adipocytes (Fig. 5C). The Lane-Sandhu histological scores of each group are listed in Table I. Group A exhibited significantly higher Lane-Sandhu scores at each time point compared with groups B and C, indicating that the scaffold in group A achieved better osteogenesis.