Long-Term Effect of Incadronate Disodium (YM-175) on Fracture Healing of Femoral Shaft in Growing Rats

Authors


Abstract

The aim of this study was to investigate the long-term effect of incadronate on fracture healing of the femoral shaft in rats. Female Sprague-Dawley 8-week-old rats were injected subcutaneously (sc) with either vehicle (V group) or two doses of incadronate (10 μg/kg and 100 μg/kg) three times a week for 2 weeks. Right femoral diaphysis was then fractured and fixed with intramedullary stainless wire. Just after fracture, incadronate treatment was stopped in pretreatment groups (P groups: P-10 and P-100) or continued in continuous treatment groups (C groups: C-10 and C-100). All rats were killed at 25 weeks or 49 weeks after surgery. Fractured femur was evaluated radiologically and mechanically and then stained in Villanueva bone stain and embedded in methyl methacrylate. Undecalcified cross-sections from the fracture area were evaluated microradiologically and histomorphometrically. Radiographic observation showed that the fracture line disappeared in all groups. Cross-sectional area in the C-100 group was the biggest among all groups and in the C-10 group was larger than that in the V group at 25 weeks. Histological and histomorphometric observations showed that the process of fracture healing was delayed under continuous treatment with incadronate as evidenced by the delay of both lamellar cortical shell formation and resolution of original cortex in C groups. Percent linear labeling perimeter, mineral apposition rate (MAR), and bone formation rate (BFR) in C groups significantly decreased compared with the other groups, indicating that the callus remodeling was suppressed under continuous treatment, especially with a high dose. Mechanical study showed that the stiffness and ultimate load of the fractured femur in the C 100 group were the highest among all groups at both 25 weeks and 49 weeks. In conclusion, this study showed that long-term continuous treatment with incadronate delayed the process of fracture healing of femur in rats, especially under high dose but it did not impair the recovery of mechanical integrity of the fracture.

INTRODUCTION

BISPHOSPHONATES HAVE been used in the treatment of diseases involving excessive osteoclast-mediated bone resorption such as Paget's disease,(1) tumor-induced hypercalcemic and metastatic bone diseases,(2) and more recently, postmenopausal osteoporosis.(3–5) Several bisphosphonates have been approved as the therapies for osteoporosis in many countries. In particular, both etidronate and alendronate have been confirmed to increase bone mass and decrease the fracture rates to approximately half of control rates at the spine, hip, and other sites in postmenopausal women.(4)

Skeletal half-life of bisphosphonates generally is very long. For alendronate it is approximately 200 days in rats, 3 years in dogs, and 12 years in humans.(6) Long retention of bisphosphonates in bone might result in a prolonged suppression of bone remodeling thereby subsequently inducing harmful effects in bone.(7) The question has been raised whether bisphosphonate should be continued or stopped when fracture occurs during the treatment. Therefore, it is important to know whether the long-standing antiresorptive effect of bisphosphonate would affect the fracture healing and its mechanical properties. However, to date, very few studies have focused on the long-term effect of bisphosphonate on fracture healing.(8–12)

Etidronate, a first generation bisphosphonate, has been shown to either delay fracture healing or impair callus mineralization and the strength of fractured bone at a high dose.(8) However, a new bisphosphonate alendronate even at 10 times the anticipated clinical therapeutic dosage in mature dogs did not show any adverse effects on fracture healing, mineralization, and mechanical properties.(9) The studies using clodronate have shown different findings in the process of fracture healing, mineralization of bone matrix, callus remodeling, and bone strength.(10–13) These studies suggest that bisphosphonates greatly differ in their effects on bone remodeling, mineralization, fracture repair, and mechanical strength of the fractured bone.(8–13) Thus, it is necessary to investigate the effects of individual bisphosphonates on fracture healing.(14)

Incadronate, a third generation bisphosphonate, has the potential to maintain bone mineral density (BMD), structure, and mechanical properties of cancellous bone in ovariectomized (OVX) beagles.(15) Our previous study(16) showed that continuous treatment with incadronate (10 μg/kg and 100 μg/kg, three times a week) for 6 weeks or 16 weeks delayed the femoral fracture remodeling in rats. Results showed that with both doses, the callus area was larger and mainly consisted of woven bone, and the lamellar cortical shell was not observed, whereas the high-dose group showed the highest ultimate load among all groups at 16 weeks.(16) However, because fracture healing was not completed at 16 weeks after fracture, it was still unknown whether callus remodeling would progress and catch up to the control level, or whether the recovery of mechanical integrity of the fractured bone could advance in the late stage of fracture healing. To answer these questions, in this study a longer-term experiment (25 weeks and 49 weeks) was performed.

MATERIALS AND METHODS

Six-week-old female Sprague-Dawley rats (n = 160; Japan SLC, Inc., Hamamatsu, Japan) weighing 131.7 ± 1.6 g were acclimated for 2 weeks to local vivarium conditions (24°C ± 2°C and 12 h/12 h light-dark cycle). During the experimental period, three to four rats were housed in one cage (floor area, 988 cm2, and height, 18 cm) and allowed free access to water and pelleted commercial rodent diet (Oriental Yeast Co., Tokyo, Japan).

Animals at the age of 8 weeks were divided randomly into five groups including one vehicle control group (V) and four treatment groups (32 rats per group) according to the body weight. Incadronate disodium (YM-175; Yamanouchi Pharmaceutical Co., Tokyo, Japan) treatment had two modes including pretreatment (P groups: P-10 and P-100) and continuous treatment (C groups: C-10 and C-100). Each treatment mode had two different doses: 10 μg/kg and 100 μg/kg. All rats were injected subcutaneously (sc) with either vehicle (0.9% saline) or incadronate (10 μg/kg or 100 μg/kg) 3 times a week for 2 weeks. Fracture surgery was then performed on the right femoral diaphysis. Just after surgery, the treatment was stopped in P groups but continued in C groups until death at 25 weeks or 49 weeks. The experimental protocol (Fig. 1) was approved by the animal study committee of the Kagawa Medical University.

Figure FIG. 1..

Experimental protocol. V group, vehicle treatment; P-10 and P-100 groups, incadronate pretreatment for 2 weeks before fracture surgery; C-10 and C-100 groups, incadronate continuous treatment before and after surgery until death. Incadronate dose regimens: 10 μg/kg or 100 μg/kg, sc, three times a week.

Surgery was done under general anesthesia with pentobarbital sodium (50 mg/kg, intraperitoneally [ip]; Abbott Laboratories, North Chicago, IL, USA). All animals were prepared for surgery by shaving and cleansing the right rear leg. A transverse osteotomy at the midshaft of the right femur was made by a fine-toothed circular saw blade mounted on an electrical drill (Kiso Power Co., Osaka, Japan). A stainless intramedullary wire (diameter, 1.5 mm; Zimmer, Warsaw, IN, USA) was first inserted into the medullary cavity from the fracture site penetrating into the distal end of the femur by a wire driver (Stryker Co., Kalamazoo, MI, USA). After reduction of fracture, the wire was then driven retrogradely to the proximal part of the femur and the fracture fragments were connected and stabilized. The wire was cut on the surface of the intercondylar groove to make sure that the movement of the knee was not influenced. Unrestricted activity was allowed after recovery from anesthesia.

Before death, all rats were double fluorescent-labeled with doxycycline (30 mg/kg, sc; Pfizer, Tokyo, Japan) and calcein (6 mg/kg, sc; Wako, Ltd., Osaka, Japan) at the schedule of 2-10-2-5. The rats were killed at 25 weeks and 49 weeks after surgery and bilateral femurs were collected and dissected free of soft tissues. Anteposterior soft radiophotographs of all fractured femurs were taken (30 kilovolt peak [kVp], 2 mA, 15 minutes; SRO-M40; Sofron, Tokyo, Japan).

For mechanical testing, six to eight fractured femurs were selected randomly from each group after pulling out of the intramedullary wire. Then femurs were tested by the three-point bending method using a mechanical testing machine (model TK-252C; Muromachi Kikai Co., Ltd., Tokyo, Japan). The fractured femur was placed, facing 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 the femur (anteroposterior position). Load was applied at the strain rate of 2.5 mm/minute until breakage. Ultimate load and stiffness were determined from the load-deformation curve by a connected computer; ultimate stress also was calculated by ultimate load divided by cross-sectional area of the fracture plane, which reflects the material property of the fractured bone.

The other fractured femurs were fixed in 70% ethanol, stained with Villanueva bone stain, dehydrated in increasing concentrations of ethanol, defatted in xylene, and embedded in methyl methacrylate. The specimens from the fracture area were cut into 150-μm-thick cross-sections with a band saw (Exakt; Otto Herrmann Co., Norderstedt, Germany), ground to a 100-μm thickness for contact microradiographs (15 kVp, 2 mA, 10 minutes; SRO-M40; Sofron) and further ground to a 25-μm thickness for histomorphometry.

Histomorphometric analysis was performed with a semiautomated digitizing image analyzer. The system consists of a light or epifluorescent microscope and a digitizing pad coupled to a computer with histomorphometric software (System Supply Co., Nagano, Japan). Polarized light was applied to distinguish lamellar bone from woven bone. Total area of the cross-section (C.T.Ar, mm2) and periosteal perimeter (P.Pm, mm) were measured under light microscope. Single-labeled surface (P-sL.Pm, mm), double-labeled surface (P-dL.Pm, mm), and interlabeled width (Ir.L.Wi, mm) on periosteal surface were measured under epifluorescent light. Lamellar cortical shell perimeter (Ls.Pm, mm) and lamellar cortical shell width (Ls.Wi, mm) were measured under polarized light. All of these parameters except C.T.Ar were measured at ×125 magnification. C.T.Ar was measured at ×12.5 magnification. Percent labeled perimeter (%P-L.Pm, %), mineral apposition rate (MAR, mm/day), bone formation rate/bone surface referent (BFR/BS, mm/day × 100), and percent lamellar cortical shell (%Ls.Pm, %) were calculated. Original cortex was identified by the different density or presence of a border from the callus under light and polarized light microscope. The differences between the control and treatment groups and among treatment groups were tested by two-way analysis of variance (ANOVA) and Fisher's protected least significant difference test using the StatView J 4.02 package (Macintosh, Abacus Concepts, Berkeley, CA, USA). The differences were considered significant at p < 0.05 level.

RESULTS

Of the total 160 rats, 15 were excluded because of death during the surgical procedure, inflammation, or technical failure of surgery. After fracture surgery, the rats resumed normal activity within a few days. Incadronate injection did not cause any side effects. All animals gained body weight during the experiment, but the weight in the C-100 group was significantly lower than that in the V group. The length of both fractured and intact femurs was significantly lower in the P-100, C-10, and C-100 groups than in the V group.

Radiographs

Soft X-ray observation showed that the fracture line disappeared in all groups (Fig. 2). At 25 weeks, callus size and X-ray density in C groups were higher than those in V and P groups. At 49 weeks, there were similar findings with regard to X-ray density, but callus size was larger only in the C-100 group as compared with the V group.

Figure FIG. 2..

Radiographs of fractured femurs at (A-E) 25 weeks and (A′-E′) 49 weeks [(A and A′) V groups; (B and B′) P-10 groups; (C and C′) P-100 groups; (D and D′) C-10 groups; (E and E′) C-100 groups]. Fracture line disappeared in all groups. X-ray density was higher in incadronate treatment groups than in the V group and higher in C groups than in Pm groups. Callus width in C groups (arrow) was larger than the other groups at 25 weeks but only in the C-100 group (arrow) at 49 weeks. The length of the femurs was significantly lower in P-100, C-10, and C-100 groups than that in the V group.

Contact microradiographs (Fig. 3) showed that the C.T.Ar in P groups did not differ from that in the V group. In both C-10 and C-100 groups the area was significantly larger than that in the V group at 25 weeks but only the C-100 group showed a bigger area than the V group at 49 weeks. Morphological observation of the cross-sections showed that new cortical shells appeared in all groups at both 25 weeks and 49 weeks; all the groups except C-100 group showed the dense new cortical shell, but the C-100 group kept a highly porous shell in which endocortical border was unclear; and the remnant of original cortex was observed in all groups at 25 weeks, while it mostly disappeared in the V, P-10, and C-10 groups at 49 weeks.

Figure FIG. 3..

Contact microradiographs of cross-sections from the fracture areas at (A-E) 25 weeks and (A′-E′) 49 weeks [(A and A′) V group; (B and B′) P-10 group; (C and C′) P-100 group; (D and D′) C-10 group; (E and E′) C-100 group]. The C groups showed the larger cross-sectional area among the groups at 25 weeks but only in the C-100 group at 49 weeks. New cortical shells appeared in all groups at both 25 weeks and 49 weeks; all the groups except the C-100 group showed the dense new cortical shell, but the C-100 group kept highly porous shell in which endocortical border was unclear. The remnant of original cortex was observed in all groups at 25 weeks, while it mostly disappeared in the V, P-10, and C-10 groups at 49 weeks (original magnification ×2.5).

Histology and histomorphometry

Measurement of C.T.Ar showed no significant difference between the V and P groups at both time points. The areas in the C-10 and C-100 groups were significantly larger than in the V group at 25 weeks (+4% and +8%, respectively, p < 0.001), but only the C-100 group showed a bigger area than the V group at 49 weeks (+6%, p < 0.001; Table 1).

Table Table 1.. Histomorphometric Evaluation at Femoral Fracture Area
original image

Under polarized light (Figs. 4A–4E and 5A–5E), the lamellar cortical shell (Ls) was observed on the periosteal surface in all groups as identified by the regularly arranged fibrils. The %Ls.Pm and Ls.Wi increased in all groups from 25 to 49 weeks. Both %Ls.Pm and Ls.Wi showed no apparent difference between V and P groups but were significantly lower in C-10 and C-100 groups than those in the V group (at 25 weeks: %Ls.Pm, −22.1% and −29%; Ls.Wi, −17.5% and −22%; p < 0.01−0.001; at 49 weeks: %Ls.Pm, −39.6% and −67%; Ls.Wi, −24.1% and −38%; p < 0.001). Much woven bone remained in C groups at 25 weeks, but only little remained in the C-10 group at 49 weeks.

Figure FIG. 4..

Microphotographs of cross-sections from the fracture areas at 25 weeks (A-E) under polarized light and (A′-E′) under epifluorescent light [(A and A′) V group; (B and B′) P-10 group; (C and C′) P-100 group; (D and D′) C-10 group; (E and D′) C-100 group]. Lamellar cortical shell (black arrow) was observed on periosteal surface in all groups as identified by the regularly arranged fibrils, which is thinner in C groups than V or P groups. Periosteal linear labeling (white arrow) was less in C groups than the other groups (original magnification ×50).

Figure FIG. 5..

Microphotographs of cross-sections from the fracture areas at 49 weeks under (A-E) polarized light and (′-E′) epifluorescent light [(A and A′) V group; (B and B′) P-10 group; (C and C′) P-100 group; (D and D′) C-10 group; (E and E′) C-100 group]. C-groups showed thinner lamellar cortical shell (black arrow) than the other groups; the difference was more obvious at 49 weeks than that at 25 weeks. Periosteal linear labeling (white arrow) was less in C groups than the other groups (original magnification ×50).

Under epifluorescent light (Figs. 4A′–4E′ and 5A′–5E′), less linear labeling on periosteal surface in C groups indicated that periosteal lamellar bone formation was not active under continuous treatment with incadronate. The %P-L.Pm, MAR, and BFR/BS did not significantly differ between V and P groups at both 25 weeks and 49 weeks (Table 1). However, these parameters in the C-10 and C-100 groups were significantly lower than the corresponding values in the V group (at 25 weeks: %P-L.Pm, −8.3% and −11.9%; MAR, −0.45% and −0.75%; BFR/BS, −11.6% and −15.1%; p < 0.01−0.001; at 49 weeks: %P-L.Pm, −8.2% and −10.4%; p < 0.001; MAR, −0.35% and −0.82%; BFR/BS, −12.1% and −15.3%; p < 0.05 and p < 0.001, respectively). Remodeling of the fracture site in the V group appeared to be still ongoing, indicating that the fractured bone had not been repaired completely at the end of the experiment.

Mechanical testing

In the fractured femur, stiffness and ultimate load in the C-100 group were the highest among all groups, but in the C-10 group stiffness and ultimate load were only slightly and not significantly higher than those in the V group. Ultimate stress in the C-100 group did not significantly increase compared with that in the V group at both time points (C-100 group vs. V group: at 25 weeks, 7.2 ± 0.4 N/mm2 vs. 8.2 ± 0.9 N/mm2; at 49 weeks, 10.9 ± 1.0 N/mm2 vs. 10.5 ± 1.2 N/mm2). There was a trend that stiffness and ultimate load as well as ultimate stress in all groups increased from 25 to 49 weeks after fracture (Fig. 6).

Figure FIG. 6..

(A) Stiffness and (B) ultimate load of the fracture femurs. Stiffness and ultimate load in the C-100 group were the highest among all groups and in the C-10 group was slightly but not significantly higher than that in V, P-10, and P-100 groups at both 25 weeks and 49 weeks after fracture (ap < 0.05,bp < 0.01,cp < 0.001 versus the V group;dp < 0.05,ep < 0.01,fp < 0.001 versus the P-10 group;gp < 0.05,hp < 0.01,ip < 0.001 versus the P-100 group;jp < 0.05,kp < 0.01,lp < 0.001 versus the C-10 group).

DISCUSSION

Many studies have confirmed that the rat femoral diaphyseal fracture model is adequate to be used to investigate the fracture repair mechanism and the intervention factors on fracture healing.(12,16–19) Fracture healing is a closely regulated process but also can be interfered by exogenous factors including bisphosphonates(8–13,16,20–22) and other agents.(19,22–24)

The fracture-healing process includes various stages such as endochondral ossification, woven bone production, and callus remodeling to lamellar bone, and fracture callus is heterogeneous with respect to the tissue composition, especially in the early stage. These situations make histological evaluation of fracture callus very hard. Conventional histomorphometric parameters, usually used in cortical and trabecular bone measurements, can be applied to the quantitative and dynamic evaluation of callus remodeling to lamellar bone at the late stage of fracture healing.(25) The present study showed that the process of fracture healing progressed not only in V group, but also in the P group and the C group treatment with incadronate groups with both doses (10 μg/kg and 100 μg/kg) as evidenced by histological observations and histomorphometric measurement. One of the two doses of 100 mg/kg, three times a week, was 10 times that of the anticipated clinical therapeutics dosage. In particular, the lamellar cortical shell was observed in all groups but the C group at 6 weeks and 16 weeks(16); and percent lamellar cortical shell and its width in all groups increased from 25 to 49 weeks after fracture. However, the healing process was still delayed under continuous treatment with incadronate as shown by callus remodeling, lamellar cortical shell formation, and resolution of original cortex, which all underwent a slower rate compared with the V group.

As shown in our previous study,(16) callus remodeling under pretreatment with incadronate was delayed at 6 weeks but caught up with the V group level at 16 weeks. In the present study, both histomorphometric observation of the fracture site and the examination of mechanical strength of fractured bone showed no significant difference between V and P groups. These findings indicate that pretreatment with incadronate only delayed the early stage of fracture healing and did not affect the strength of fractured bone. Pretreatment with alendronate(9) at a dose of 2 mg/kg per day for 9 weeks before fracture in dogs was shown to lead to a larger callus but did not disturb mechanical strength of fractured bone. Moreover, Madsen et al.(21) reported that pretreatment with clodronate at a dose of 10 mg/kg per day for 4 weeks before fracture affected neither callus remodeling nor mechanical property of fractured bone at 4 weeks after fracture. These studies suggest that short-term pretreatment with bisphosphonates would not markedly disturb the fracture healing as long as the treatment is stopped when the fracture occurs.

In contrast, continuous treatment with incadronate appears to delay the fracture healing at any time point after fracture (6 weeks and 16 weeks as in our previous study,(16) and 25 weeks and 49 weeks in the present study), as evidenced by a bigger remaining cross-sectional area (except in the C-10 group at 49 weeks) and delay of both new cortical shell formation and resolution of original cortex, compared with the V group. These observations indicate that callus remodeling is suppressed under continuous treatment with incadronate, especially at a high dose.

Studies of effects of other bisphosphonates on fracture healing have shown that both alendronate at a dose of 2 mg/kg daily for 25 weeks in dogs(9) and pamidronate at a dose of 0.5 mg/kg weekly for 12 weeks in sheep(22) also delays callus remodeling. However, tiludronate(26) at doses of 5 mg/kg and 10 mg/kg daily for 6 weeks in dogs did not affect the process of fracture healing. With respect to the delay in fracture healing relative to the treatment duration, clodronate, at a dose of 10 mg/kg daily in rats, has been shown to alter neither the callus histology nor delay the fracture healing at 4 weeks,(21) but it delayed mineralization of bone matrix at 6 weeks(13) and delayed callus remodeling at 8 weeks.(12) Moreover, clodronate, even at different doses of 3, 10, or 30 mg/kg daily for 22 weeks, did not alter the callus histology or delay the fracture healing,(10) and all fracture calluses eventually were remodeled to lamellar bone, despite the clodronate treatment.(27) With respect to the delay in fracture healing relative to the dose, etidronate(8) in dogs for 20 weeks daily, has shown that at a dose of 0.1 mg/kg the fracture healed normally, at a dose of 0.5 mg/kg markedly delayed X-ray union of the fractured bone, and at a dose of 5 mg/kg totally disrupted callus mineralization activity.

Analysis of the fractured bone strength has been suggested as a crucial step for evaluation of fracture repair.(8–9,11,16,25) In the present study, the length of femurs in the P-100, C-10, and C-100 groups was lower than that in the V group, but we considered that it does not affect the results of mechanical test, because three-point bending test was performed at a relatively small and consistent distance of midshaft of the femur. As shown in our previous study,(16) in the C-100 group ultimate load of fractured bone was the highest among all groups, but ultimate stress was significantly lower than that in the V group, indicating that the bone quality in callus was poorer under high-dose continuous treatment with incadronate at 16 weeks after fracture.(16) The present study showed that in the C-100 group ultimate load and stiffness of fractured bone were the highest among all groups, but ultimate stress did not significantly differ from the V group, indicating that the material property of the fractured bone in the C-100 group might catch up to the V group level during fracture repair at 25 weeks and thereafter. This is consistent with the advance of fracture healing in the C-100 group as shown by histomorphometric observation. The increase in the strength values of fractured bone geometrically could be explained by the bigger cross-sectional area in the C-100 group. This result is in agreement with the finding using pamidronate, which indicated that the ultimate tensile strength in the treated group was significantly higher than the control group at 12 weeks after fracture in adult sheep.(22) However, mechanical strength of the fractured bone might be unaffected under treatment with alendronate,(9,20) tiludronate,(26) and clodronate(11,21) but affected use with etidronate.(8,10) Taken together, the previously mentioned studies further confirm that whether bisphosphonates interfere with fracture repair and mechanical strength of the fractured bone varies based on their chemical structure, potency, dosage, and duration.

Fracture repair is a complex morphogenetic process involving coordinated participation of several types of osteoblastic and osteoclastic cells.(28) The in vitro studies indicate that the differentiation of osteoclast progenitors into osteoclasts is required for the cell-to-cell contact of osteoblastic cells and osteoclast progenitors(29); and an increase of osteoclastic bone resorption in myeloma usually is associated with a marked impairment of osteoblast function and the impaired osteoblasts respond to the increase of bone resorption.(30) Resorption of mineralized bone matrix by osteoclasts is preceded by osteoblast collagenase activities.(31) However, bisphosphonates inhibit osteoclast activity(32) and their long-term continuous administration may further suppress osteoclast differentiation.(33) Also, osteoblast recruitment and differentiation might be affected under bisphosphonate treatment.(13,16) A study by Nii et al.(34) has shown that under incadronate treatment not only osteoclasts but also active osteoblasts were rarely observed on the newly formed trabecular bone surface. The present study showed delay of callus remodeling and lamellar cortical shell formation, suggesting that both osteoclast and osteoblast function could be inhibited significantly by incadronate.

We have observed, in our 2-week and 4 week study,(35) a time-course decrease of bisphosphonate concentration in pretreatment groups while it increased in the C group and also that bisphosphonate concentration in P group decreased more rapidly in the fractured side than in the intact side. Our previous study of 6 weeks and 16 weeks showed that the histological fracture-healing process was delayed in the P group compared with control at 6 weeks but reached a control level at 16 weeks, suggesting that retained bisphosphonate even after withdrawal keeps suppressing bone resorption at a certain time period and duration of suppressed bone resorption depends on the rate of bone turnover, which would determine the rate of release of retained bisphosphonate from the bone.

During fracture repair, callus undergoes remodeling, continuously being resorbed and formed until the normal architecture and mechanical integrity of bone is restored.(25,28) The present study showed that the fracture repair had not yet been completed at the end of the experiment, based on the remodeling of fracture site in the V group, which appeared to be still ongoing as shown by histomorphometric observation. Chao et al.(25) showed that the period of fracture repair was the longest in the late stage of fracture healing. Fracture repair mainly depends on bone turnover rate and the amount of remodeling units in different species.(23,36–38) Generally, in humans, it will take 1–4 years to complete replacement of the callus with functionally competent lamellar bone(37) and 32 weeks to complete the remodeling of the fracture site in the dog tibial diaphysis fracture model.(23) However, how long it may take to complete repair of femoral fracture in rats is still not clear.

A recent study in the mouse femoral fracture model(20) reported that the process of fracture healing eventually will proceed to completion as identified by X-ray union at 4 weeks after fracture. In fact, during fracture healing, radiological union of the fracture occurs earlier than the complete histological healing and recovery of bone strength in both animal fracture models(12,16,23,38) and humans.(37) To evaluate objectively the process of fracture healing, not only radiological but also histological and histomorphometrical examinations, as well as studies of mechanical properties of the fractured bone, should be carried out. Clinically, restoration of mechanical integrity of the fracture is the most critical issue. Therefore, many investigators have proposed that the healing of the fractured bone may be considered complete when it reaches(23,25,37): (1) the fractured bone fragments reunion as evidenced by disappearance of the fracture line radiologically, (2) restoration of anatomical architecture histologically, and (3) recovery of bone strength mechanically.

In conclusion, the present study showed that long-term continuous treatment with incadronate delayed the process of fracture healing of the femur in rats, especially under a high dose but did not impair the strength of the fractured bone.

Acknowledgements

We thank Ryuhei Fujimoto, D.V.M. of Yamanouchi Pharmaceutical Co., Ltd. for providing incadronate. We also thank Miss Shimano and Miss Kawada for their assistance in processing the bone specimens.

Ancillary