Part of this study was presented at The Second International Conference on Osteoporosis, Osaka, Japan, November 1997.
This study was designed to test whether bisphosphonates disturb the process of fracture healing. Female Sprague-Dawley rats were injected with either two doses of bisphosphonate (incadronate) (10 μg/kg and 100 μg/kg) or vehicle three times a week for 2 weeks. Right femora were then fractured and fixed with intramedullary wires. Incadronate treatment was stopped in pretreatment groups (P-10 and P-100 groups), while the treatment was continued in continuous treatment groups (C-10 and C-100 groups). Animals were sacrificed at 6 and 16 weeks after surgery. Soft X-ray of all fractured femora was taken. After mechanical testing, fractured femora were stained in Villanueva bone stain and embedded in methyl methacrylate. Cross-sections near fracture line were analyzed by microradiography and histomorphometry. Radiographic study showed that bony callus was present in all the fractures and incadronate treatment led to a larger callus, especially in C-100 group at both 6 and 16 weeks. Histologic study showed that the process of fracture healing in pretreatment groups was delayed at 6 weeks, but reached control level thereafter and showed same characteristics as in control at 16 weeks. Woven bony callus could still be seen in continuous treatment groups at 16 weeks. Mechanical study indicated that the ultimate load of C-100 group was slightly higher than the other treatment groups and control. The results suggest that pretreatment with incadronate did not affect fracture healing at 16 weeks after fracture. However, continuous incadronate treatment could lead to larger callus, but it delayed remodeling process during fracture healing, especially with high-dose treatment.
Bisphosphonates have been known as inhibitors of bone resorption and used as effective therapeutic agents for a variety of bone diseases with high bone resorption, such as metastatic bone disease, Paget's disease, and osteoporosis.1-7 Clinical trials have shown that intermittent cyclical etidronate treatment or continuous treatment with alendronate increases bone mass and reduces the rate of fractures in postmenopausal women.8, 9 Both pamidronate and etidronate decrease risk of bone fractures in Paget's disease and metastatic bone diseases.10, 11 Recently much attention has been paid to the effects of bisphosphonates on fracture healing. For example, clodronate increased callus formation, delayed the remodeling, and did not affect the bending strength of fractured bone.12-14 Tiludronate did not decrease or delay the biomechanical recovery of fractured bone in dogs.15 The previous studies12-18 have been aimed mainly at revealing the possible harmful effects of bisphosphonates on fracture union. However, the effects of bisphosphonates related to treatment mode and dosage on fracture healing have not been well elucidated. Further studies are needed, especially using the new bisphosphonates with more powerful antiresorptive ability.
The purpose of the fracture healing is to restore the original histologic structure and mechanical properties of the broken bone.19-21 The process of experimental fracture healing has mainly been studied with radiographic and histologic techniques.22-24 In the present study, we also used the histomorphometrical method25 following bisphosphonate treatment to quantitatively analyze callus formation and subsequent changes. The functional recovery of the fractured limb was determined by a three-points bending test of the fractured femora.22-24
The purpose of the study was to investigate the effect of incadronate26-29 (Fig. 1), one of the new bisphosphonates, on the process of fracture healing and mechanical properties of callus.
MATERIALS AND METHODS
Female Sprague-Dawley rats (n = 120, 6 weeks old; Japan SLC Inc., Hamamatsu, Japan) were acclimated for 7 days to local vivatium conditions (24 ± 2°C and 12 h/12 h light-dark cycle). During the experimental period, animals were housed in cages (floor area 988 cm2 and height 18 cm) and allowed free access to water and pelleted commercial rodent diet (Oriental Yeast Co., Tokyo, Japan). The rats weighed 150 ± 8 g at the beginning of the study. The protocol was approved by the Kagawa Medical University animal study committee.
The animals were randomly allocated into five groups: one vehicle control group (V) and four treatment groups with the same mean body weight (24 rats per group) (Fig. 2). Treatment groups were divided based on the dose and mode of treatment. Two doses (10 μg/kg and 100 μg/kg) of incadronate (Yamanouchi Pharmaceutical Co., Tokyo, Japan) and two modes of treatment including pretreatment before fracture (P-10 and P-100 groups) and continuous treatment before and after fracture (C-10 and C-100 groups) were used. All animals were injected with either incadronate (10 μg/kg, 100 μg/kg) or vehicle (0.9% saline) three times a week for 2 weeks. Surgery was then performed on the right femora. After surgery, the treatment was stopped in P-10 and P-100 groups but was continued in C-10 and C-100 groups three times a week until sacrifice. Vehicle administration was also continued until sacrifice. The animals were sacrificed at 6 and 16 weeks after surgery.
The incadronate solutions were prepared by adding the powder form of the compound to 0.9% saline and adjusting its concentration to 10 μg/ml or 100 μg/ml. Incadronate was administered subcutaneously and two doses were used such that the injection volume was the same in all animals.
Surgery was performed under general anesthesia with pentobarbital sodium (intraperitoneal 50 mg/kg; Abbott Laboratories, North Chicago, IL, U.S.A.). All animals were prepared for surgery by shaving and cleansing the right rear leg. Through a lateral approach on the right leg, a transverse osteotomy at the midshaft of the femur was made by a fine-toothed circular saw blade mounted on an electrical drill (Kiso Power Co., Osaka, Japan). A stainless wire (diameter 1.5 mm; Zimmer, Warsaw, IN, U.S.A) was first inserted into the medullary cavity from the fracture site penetrating into the distal end of femur by a wire driver (Stryker Co., Kalamazoo, MI, U.S.A). After reduction of fracture, the wire was then retrogradely driven to the proximal part of the femur, and fracture fragments were contacted and stabilized. The wire was cut on the surface of intercondylar groove to make sure that the movement of the knee was not influenced. Unrestricted activity was allowed after recovery from anesthesia.
Before sacrifice, all animals were double fluorescent-labeled with calcein (subcutaneously 6 mg/kg; Wako, Ltd., Osaka, Japan) at the schedule of 1-6-1-1. The animals were sacrificed at 6 and 16 weeks after the fracture surgery and bilateral femora were collected. The intramedullary wires were extracted without difficulty and the femora were dissected to free of soft tissues. The anteposterior soft radiographs of all fractured femora were taken (30 Kvp, 2 mA, 15 minutes; SRO-40; Sofron, Tokyo, Japan).
Both the fractured (right) and intact left femora were tested by the three-point bending method using a mechanical testing machine (Model TK-252C; Muromachi Kikai Co., Ltd., Tokyo, Japan). The femur was placed, facing its anterior surface down, on the two lower support bars (10 mm apart) with the loading bar positioned at the fracture site, or the middle of femur (anterioposterior position). Load was applied at the strain rate of 2.5 mm/minute until break. Ultimate load and stiffness were determined from the load-deformation curve by a connected computer. The ratio of ultimate load of the fractured femur to ultimate load of the intact femur (LF/LI) was calculated.
After mechanical testing, the fractured femora were repositioned, fixed in 70% ethanol, stained in Villanueva bone stain, dehydrated in increasing concentrations of ethanol, defatted in xylene, and embedded in methyl methacrylate. Two 200-μm-thick cross-sections were cut with a band saw (Exakt; Otto Herrmann Co., Norderstedt, Germany) in an area within 500 μm of the original fracture line and then ground to 100 μm thickness for contact microradiographs (15 Kvp, 2 mA, 10 minutes; SRO-40; Sofron) and to 30 μm thickness for histomorphometry (Fig. 3).
Histomorphometrical analyses were performed with a semiautomated digitizing image analyzer. The system consisted of a light or epifluorescent microscope and a digitizing pad coupled to a computer with a histomorphometric software (System Supply Co., Nagano, Japan). Polarized light was applied to distinguish lamellar bone from woven bone. Total cross-sectional area (T.Ar), original cortical area (O.Ar), and medullary area (M.Ar) were measured at ×12.5 magnification and callus area (Ca.Ar) was calculated. Further measurement was made at ×125 magnification in four standardized quarters: anterior, posterior, medial, and lateral aspects (Fig. 3). T.Ar, bone area (B.Ar), bone surface (B.Pm), single-labeled surface (sL.Pm), double-labeled surface (dL.Pm), and interlabeling width (Ir.L.Wi) were measured. Percentage bone area (%B.Ar), percentage labeled surface (%L.Pm), mineral apposition rate (MAR), and bone formation rate (BFR/B.Pm) were calculated (Table 1).
Table Table 1. Histomorphometrical Parameters
The difference between the control and treatment groups and among treatment groups was tested by two-way analysis of variance and Fisher's protected least significant difference test. p values < 0.05 were considered significant.
Among 120 rats with fractures, 9 were excluded because of death or technical failure of the surgery. After fracture surgery, the rats resumed normal activity within a few days, and incadronate injection did not cause any side-effects.
Animal body weight increased over the study period. After administration of incadronate for 2 weeks, the weights did not differ among the groups. At 6 weeks after the surgery, only C-100 group showed significantly lower body weight than the V group (difference between mean weights = 22 g, p < 0.01). At 16 weeks after the surgery, there was no significant difference in body weight among any of the groups.
Soft X-ray observation showed external callus formation in all fractured femora (Fig. 4). At 6 weeks, the callus width was wider and the callus density higher in both incadronate pretreatment groups (P-10 and P-100 groups) and continuous treatment groups (C-10 and C-100 groups) than in the V group. At 16 weeks, similar observation was found only in continuous incadronate treatment groups, especially in C-100 group.
Cross-sectional area was slightly larger in P-10, P-100, and C-10 groups as shown by contact microradiographs and significantly larger in C-100 group as determined by the measurement of cross-sectional area (+165% at 6 weeks and +185% at 16 weeks) in comparison with the V group (Fig. 5). Morphological observation of external callus indicated that at 6 weeks new cortical shells formed around the callus area with new bone marrow between them and original cortex in 80% (8/10) of the fractured femora in the V group and 40% (5/12) of the fractured femora in the P-10 group, whereas callus only showed trabecular type bone in P-100 and continuous treatment groups. At 16 weeks, new cortical shells were observed in all fractured femora in groups V, P-10, and P-100 groups, in only 20% in the C-10 group and none in the C-100 group (Fig. 6 and Table 2).
Table Table 2. Percent of New Cortical Shell in Each Group
Histology and histomorphometry
In the V group, at 6 weeks histologic observations of callus showed that callus bone had transformed into a new cortical shell and new bone marrow was observed between the new cortical shell and original cortex (Figs. 7A and 7A′). Sequential labeling on the inner surface of the cortical shell indicated the progress of appositional bone formation, which resulted in thickening of the shell and reduction of new marrow space from 6 to 16 weeks (Figs. 8A and 8A′). Under polarized light, orientation of fibrils in the cortical shell became more regular at 16 weeks than at 6 weeks, implying that further remodeling occurred inside the cortical shell. The value of %B.Ar was the lowest at 6 weeks and not significantly different from the other groups at 16 weeks. The %L.Pm, MAR, and BFR/B.Pm were the highest at 16 weeks (Table 3).
Table Table 3. Evaluation of Histomorphometrical Parameters
In the P-10 and P-100 groups, callus showed trabecular type bone at 6 weeks (Figs. 7B, 7B′, 7C, and 7C′). Linear labeling and regular orientation of fibrils on the trabecular surface indicated that lamellar bone formation was undergoing and leading to the formation of new cortical shell at 16 weeks (Figs. 8B, 8B′, 8C, and 8C′). P-10 and P-100 groups showed delayed appearance of new cortical shell compared with the V group at 6 weeks but presented the same histologic characteristics as the V group at 16 weeks. In the P-100 group, %B.Ar was higher (+46%) and %L.Pm lower (–43%) than in the V group at 6 weeks. %L.Pm, MAR, and BFR/B.Pm in the P-10 and P-100 groups were slightly lower than those in the V group at 16 weeks (Table 3).
In the C-10 group, callus showed trabecular type bone in which irregular orientation of fibrils indicated that the bone in callus was woven at 6 weeks (Figs. 7D, and 7D′). At 16 weeks, 20% of the specimens showed another kind of cortical shell structure where there was no linear labeling on inner surface and orientation of fibrils was still irregular compared with V group. Callus mainly showed trabecular woven bone (Figs. 8D, and 8D′). Less linear labeling indicated that lamellar bone formation was not active. Although at 6 weeks the measured parameters in the C-10 group were not significantly different from the V group, at 16 weeks %L.Pm, MAR, and BFR/B.Pm were significantly lower (−47%, −30%, and −67%, respectively) than the corresponding values in the V group (Table 3).
In the C-100 group, primary trabeculae were still seen at 6 weeks (Figs. 7E, and 7E′). Dotted and diffuse labeling suggested that woven bone formation had not been completed. At 16 weeks, callus showed mainly trabecular type bone in which orientation of fibrils was random (Figs. 8E, and 8E′). The osteocytes in the callus were more circular-shaped than the normal spindle-shaped osteocytes in the V group. There was no new bone marrow, but trabecular woven bone formed a dense woven bone shell around the callus area. Less linear labeling indicated that lamellar bone formation was inactive. The %B.Ar was significantly larger than that in the V group at 6 weeks, whereas %L.Pm, MAR, and BFR/B.Pm were significantly lower at both 6 weeks (−60%, −50%, and −83%, respectively) and 16 weeks (−74%, −51%, and −89%, respectively) compared with the V group (Table 3).
In the three-point bending test, all the fractured femora failed along the original fracture line. The ultimate load and stiffness of fractured femora within the same group increased from 6 weeks to 16 weeks after fracture for all groups. Ultimate load in the C-100 was higher compared with the other groups at 16 weeks (Fig. 9A). Stiffness in both the C-10 and C-100 groups was higher than in the V, P-10, and P-100 groups (Fig. 10). However, the differences in ultimate load and stiffness were not statistically significant.
The LF/LI ratio showed no significant difference among any of the groups at 6 weeks, but it increased significantly in group C-100 as compared with the V group at 16 weeks (Fig. 9B).
Fracture fixed by an intramedullary wire undergoes the process of fracture healing under nonrigid conditions, which is characterized by the production of external callus.22, 23 Periosteal callus formation, endochondral ossification, and bone remodeling occur sequentially in the fracture site as described previously in humans19, 20 and in animals.21-24 In the present study, similar time-course changes were also identified. The present fracture model was shown to be useful for evaluation of fracture healing.
During fracture healing, osteoclasts play an important role in endochondral ossification and remodeling of woven bone to lamellar bone.20-22, 30 Bisphosphonates inhibit osteoclast activity31, 32 and their continuous long-term use further inhibits osteoclast differentiation.33 The inhibitory effect may be directly on the osteoclasts and partly mediated by other cells, especially osteoblasts.34 Influence of bisphosphonates, such as etidronate, clodronate, and pamidronate, on fracture healing has been reported previously in clinical usage8-11 and animal studies.12-18, 35 The effect of the use of incadronate on the process of fracture healing has not been investigated.
The present study demonstrated the effects of administration of incadronate on fracture healing in rats using two different doses and two treatment modes. One of the two doses, 100 μg/kg, three times a week, was 10 times that of the anticipated clinical therapeutic dosage. Thus the study is of concern in clinical situations where it may be necessary to decide whether bisphosphonate treatment should be continued or stopped when fracture occurs during the treatment.
In the present study, callus formation evaluated by radiographs showed that all fractures healed with external osseous callus. It confirms that incadronates do not inhibit mineralization of fibrocartilage.36 Furthermore, incadronate treatment led to a larger callus as evidenced by radiographs and histomorphometrical measurement, especially in a high-dose continuous treatment compared with the control group. In previous studies, clodronate treatment (10 mg/kg, subcutaneously daily) for 6 weeks led to a higher calcium concentration in callus13 and treatment for 8 weeks induced a significantly wider callus width12 compared with control. Bone mineral content increased significantly compared with control after 12-week pamidronate treatment (0.5 mg/kg, intravenously once a week).35 However, treatment with alendronate at an oral dose of 2 mg/kg/day for 9–16 weeks did not affect callus formation.18 Bony callus may not be visible under high-dose etidronate therapy because of the inhibition of mineralization.17, 37
Histologic examination in the present study further indicated how the healing process after mineralization progressed under bisphosphonates administration. The healing process in the C-100 group proceeded most slowly as evidenced by woven bone formation at 6 weeks. Trabecular woven bone in callus was mainly seen in both continuous treatment groups at 16 weeks. This reflects that remodeling of woven bone, as shown also by the values of %L.Pm, is delayed by continuous incadronate administration. Even endochondral ossification occurs at a slower rate in a high-dose continuous treatment. In previous studies, it was also reported that woven bone formation in callus was seen at 6 weeks after fracture under 6-week clodronate treatment (10 mg/kg, daily) in adult rat.14 There was greater persistence of primary callus compared with control at 12 weeks after fracture under 16-week pamidronate treatment (0.5 mg/kg, intravenously once a week) in adult sheep.35
Unlike the previous studies, pretreatment of incadronate was carried out in the present study. The healing process in both pretreatment groups progressed more slowly as evidenced by the delayed appearance of new cortical shells at 6 weeks, but caught up with the control group at 16 weeks. These new cortical shells, which originate from the external callus, will reconstruct original bone structure at the fracture site as a result of remodeling and modeling.22 This reflects that endochondral ossification and bone remodeling during fracture healing could recover or be accelerated to catch up with the control group at 16 weeks after withdrawal of the drug as observed in the present study. Pretreatment with different dosages did not show any obvious dose-dependent changes.
Bending tests are often used to determine mechanical properties because they are convenient and expeditious.38 It is possible to locate the loading bar at the fracture site to test the specific part of the bone by using the three-point bending test.39, 40 The three-point bending test was also applied in the present study. There was no significant difference in ultimate load between the treatment groups and control group. This agrees with previous studies using bending test with clodronate, pamidronate, and alendronate treatment.12, 17 18 In previous studies, using the tensile test, there was no significant difference in the ultimate tensile load between the clodronate (50 mg/kg) and placebo groups until 4 weeks after fracture in adult rats.13 However, the ultimate tensile strength in pamidronate treatment group was significantly higher compared with the control group at 12 weeks after fracture in adult sheep.35 In the present study, at 16 weeks after fracture, a significant recovery of ultimate load as shown by the LF/LI ratio in C-100 group was found compared with the other groups. As a matter of fact, when ultimate stress in the C-100 group (7.71 ± 1.54 N/mm2) was calculated by ultimate load divided by cross bone area in the fracture plane, the material property of fractured femora in the C-100 group was significantly lower (p < 0.01 vs. the V group) than that in control group, suggesting poorer quality of bone in callus compared with control. This implies that a larger volume and cross-sectional area of callus would compensate for less quality of bone in callus to achieve the mechanical recovery of fractured bone.
As shown in the present and previous studies,13-14, 35 the recruitment of periosteal cells to the fracture site, differentiation of these cells to chondrocytes and osteoblasts, and the process of mineralization of fibrocartilage were normal under new bisphosphonate treatment. In fact, the production of the mineralization matrix in callus by endochondral bone formation and in growth plate was increased by clodronate and incadronate treatment.14, 20 41 However, bone remodeling and appearance of new bone marrow were delayed under the treatment compared with the control. This is due to loss of osteoclast activity, resulting from bisphosphonate treatment.26, 27 After withdrawal of the drug, the delayed process may catch up with the control as shown by the pretreatment groups.
In the present study, under bisphosphonate treatment, bone formation was not active and immature osteocytes embedded in bone tissue could be seen as in an earlier study.14 This implies that osteoblast recruitment and differentiation may also be delayed in vivo. But a study by Giuliani et al. has showed that bisphosphonates could directly stimulate formation of osteoblast precursors and mineralized nodules in both murine and human bone marrow cultures in vitro.42 Based on the bone remodeling mechanism,43 some authors suggested that bisphosphonates could affect osteoblast differentiation directly.14 However, osteoclasts have been shown to exhibit expression of mRNAs of TGF-β and proto-oncogene c-fos during endochondral ossification processes of the human growth plate,44, 45 both of which have been shown to have important roles in the metabolism of chondrocytes and osteoblasts.46, 47 Thus, it may be assumed that osteoblasts are affected indirectly through the osteoclast recruitment mechanism and cytokine production. However, Mundy et al. have found that an increase in osteoclastic bone resorption in myeloma is usually associated with a marked impairment in osteoblast function and that the impaired osteoblast responds to the increase in bone resorption.48 Overall, the interaction between osteoblast and osteoclast may be involved in a complicated process and be different under different conditions of diseases. The interaction between osteoblast and osteoclast during fracture healing under bisphosphonate treatment remains to be clarified.
The histologic findings under continuous incadronate treatment in the present study were similar to those under clodronate and pamidronate treatment in adult animals,13-14, 35 suggesting that a similar histologic process could be seen in both young and old animals despite time-course differences. In fact, age has been shown to affect fracture repair and bone remodeling.43, 49 50 However, the accumulation of bisphosphonates in bone, including their deposition and clearance, could be affected by age and bone conditions.51, 52 Further studies using aged animals would be necessary to determine if bisphosphonates affect fracture healing in aged animals and/or under osteopenic condition.
With respect to the clinical usage, the present study suggests that the process of fracture healing would not be disturbed if the bisphosphonate treatment is stopped when fracture occurs. However, continuous bisphosphonate treatment even at a possibly clinical dose could delay callus remodeling and maybe further delay reconstruction of bone structure. Because the histologic results indicated that fracture healing has not been completed at 16 weeks after fracture in the present study, a longer study of complete fracture healing is needed to confirm the safety of continuous bisphosphonate treatment.
In conclusion, in this study, pretreatment with incadronate delayed fracture healing at 6 weeks but did not affect fracture healing at 16 weeks after fracture. However, continuous incadronate treatment could lead to larger callus, but it delayed the remodeling process during fracture healing, especially with high-dose continuous treatment. Furthermore, consolidation of fracture was achieved by larger callus volume under continuous incadronate treatment.
We thank Ryuhei Fujimoto, D.V.M. of Yamanouchi Pharmaceutical Co., Ltd., Tokyo, Japan, for providing incadronate. We also thank Miss Shimano and Miss Kawada for their assistance in processing the bone specimens. Jiliang Li is supported by a Japanese Government (Monbusho) scholarship.