JOURNAL OF BONE AND MINERAL RESEARCH Volume 19, Number 10, 2004 Published online on July 21, 2004; doi: 10.1359/JBMR.040709 © 2004 American Society for Bone and Mineral Research

Transient Retention of Endochondral Cartilaginous Matrix With Bisphosphonate Treatment in a Long-Term Rabbit Model of Distraction Osteogenesis Elisabeth J Smith,1,2 Anthony McEvoy,1 David G Little,1 Paul A Baldock,2 John A Eisman,2 and Edith M Gardiner2

ABSTRACT: Bisphosphonates induce major increases in strength of callus in distraction osteogenesis in the short term. Poor understanding of the underlying mechanism, however, raises concerns about long-term consequences. In this long-term study in 32 rabbits, zoledronic acid transiently increased trabeculae by delayed temporal progression of endochondral bone remodeling but did not prevent radiographic completion of bone repair. Introduction: We hypothesized that bisphosphonate inhibition of osteoclast-mediated resorption would retain bone during repair, producing a larger callus in the short term. However, if remodeling was not restored, completion of the bone repair process in the long term could be jeopardized. Materials and Methods: Juvenile rabbits underwent right tibial osteotomy and 2 weeks of distraction, followed by a period of consolidation. Animals received saline (controls) or zoledronic acid (ZA; 0.1 mg/kg at surgery and again 2 weeks later), and distracted tibias were examined by radiograph, DXA, histology, and histomorphometry at 2, 4, 6, 18, and 44 weeks after surgery. Results: Regenerated bone in ZA-treated animals was denser than controls on radiographs at 6 weeks and had more distinct radiodense trabeculae and retention of original cortices at 18 weeks. By 44 weeks, controls and ZA-treated animals were radiographically healed and indistinguishable. Regenerate BMD and BMC increased between 2 and 4 weeks in all animals, with a greater effect in ZA. At 6 weeks, BMD and BMC in ZA-treated animals were 1.6- and 2-fold greater, respectively, than controls (p ⬍ 0.01). From 6 to 44 weeks, the control values gradually increased and approached the ZA-treated values. Regenerate bone volume and trabecular number by histomorphometry were from 1.6- to 2-fold greater in ZA-treated animals at 6 and 18 weeks (p ⬍ 0.05). Endochondral cartilaginous matrix volume was up to 2.4-fold greater in ZA-treated animals at 2 and 4 weeks (p ⬍ 0.05). TRACP⫹ cells in ZA-treated animals were larger with more nuclei. Mineral apposition rate and osteoblast number and surface were lower in ZA-treated animals at 6 weeks (p ⬍ 0.01) but not at later times. Conclusions: Disruption of TRACP⫹ cell function by ZA during bone regeneration seems to lead to an accretion of cancellous bone built on a larger endochondral cartilaginous matrix and increased bone mass, consistent with reported increases in short-term callus strength. This increase in bone mass, caused by a delay in remodeling, provided a transient advantage without preventing radiographic completion of the bone repair process in the long term. Noncontinuous treatment with nitrogen-containing bisphosphonates thus can have short-term beneficial effects without preventing long-term bone repair. J Bone Miner Res 2004;19:1698 —1705. Published online on July 21, 2004; doi: 10.1359/JBMR.040709 Key words:

distraction osteogenesis, bisphosphonate, endochondral cartilaginous matrix, rabbit INTRODUCTION

resorption through the disruption of osteoclast function and are routinely used for the treatment of postmenopausal osteoporosis and other skeletal disorders of excessive osteoclast-mediated bone resorption.(1–5) Recently, the focus has turned to their effect on fracture repair. No adverse effects on bone repair have

B

ISPHOSPHONATES INHIBIT BONE

Dr Eisman has corporate appointments with and research support from Merck Sharp & Dohme, Lilly, Aventis, Servier, NPS Pharmaceuticals, Roche, and Organon. All other authors have no conflict of interest.

been recorded in bisphosphonate-treated patients sustaining fracture, and increases in calcium content, mass, and strength of callus have been reported in animal models of fracture healing.(6 –13) We previously reported that the bisphosphonates pamidronate and zoledronic acid increase bone volume, BMC, and strength of regenerating bone (regenerate) in a rabbit model of distraction osteogenesis at 6 weeks after surgery.(14 –16) In distraction osteogenesis, which is used predominantly in children requiring limb-lengthening, fractures in and around the regenerate occur in 38% of cases after removal

1 Orthopaedic Research Unit, Children’s Hospital at Westmead, Sydney, Australia; 2Bone and Mineral Research Program, Garvan Institute of Medical Research, Sydney, Australia.

1698

MODULATION OF ENDOCHONDRAL BONE REMODELING

of the fixator frame.(17,18) Use of bisphosphonates to increase bone strength in the short term may reduce such fractures. However, knowledge of the underlying cellular mechanism and long-term consequences is necessary for full evaluation of the clinical benefit of bisphosphonate treatment in such cases, because prolonged inhibition of remodeling of the fracture callus could have long-term deleterious effects on the reconstruction of normal bone and function in the distracted limb. In this study we used DXA, radiography, histology, and histomorphometry in the rabbit model over a 44-week period to examine the mechanisms and long-term outcomes of zoledronic acid (ZA) treatment on bone repair. Use of ZA in this model, in which the rabbits reach adulthood during the time-course of the study, allowed the consequences of nitrogen-containing bisphosphonate treatment on the completion of the bone repair process to be studied in growing bone. The previously observed increases in callus size may be explained by inhibition of bone resorption and retention of newly deposited bone; however, bisphosphonates can also prevent the removal of mineralized cartilage,(19,20) which may therefore also have contributed to increased callus size and BMC. Furthermore, in vitro studies have suggested that the deposition of more bone may also occur as a result of increased osteoblast proliferation or differentiation.(21–23) Each of these mechanisms could have beneficial effects in the short term, but the antiresorptive effects could be a long-term concern if they prevent full callus remodeling. This study addresses these issues.

MATERIALS AND METHODS Experimental design Thirty-six, 8-week-old male New Zealand White rabbits underwent distraction osteogenesis. Animals were randomized to be infused with either saline (controls) or 0.1 mg/kg ZA at the time of surgery and again on day 14. ZA was prepared in saline from the hydrated disodium salt (MW, 399.5; Novartis Pharma AG, Basel, Switzerland). For each rabbit, an open mid-tibial osteotomy was performed on the right tibia, followed by monolateral external fixation (Orthofix M-100; Orthofix, Bussolengo, Italy) and distraction of 0.375 mm every 12 h for 14 days, producing a total lengthening of 10 mm, as previously described.(14 –16) Buprenorphine (0.05 mg/kg) was administered at the end of surgery and again 12 h postoperatively to all animals. The animals were supplied with rabbit pellet and water ad libitum. The fixator was left in situ until week 6 or death. Four animals in each group were killed at 2, 4, 6, and 18 weeks after surgery, and two in each group were killed at 44 weeks. These studies were conducted with the approval of the institutional animal ethics committee (WAEC 177.08 – 01).

using digital luminescent cassettes with a 50-kV and 4-mA exposure.

BMD and BMC BMD and BMC measurements were made at 2, 4, 6, 12, 18, 26, 32, and 44 weeks using a total body densitometer (DPX; LUNAR, Madison, WI, USA). DXA scans were performed with the tibia oriented in a jig in the AP projection, using software specifically designed for measuring small animals (Small Animal Software, 1.0c; Lunar). The HiRes ⬍0.5-kg slow-scan mode was used (fine collimation, sample size of 0.6 ⫻ 1.2 mm, and sample interval of 1⁄16 s). Regional BMD and BMC measurements were obtained by positioning the region of interest (ROI) over the regenerate on the scan image by an operator blinded to the treatment group.

Histology and histomorphometry For bone histomorphometry, animals were given subcutaneous calcein (6 mg/kg) at 4 days and demeclocycline (10 mg/kg) 1 day before death at 6 weeks after surgery, or for the later time-points, animals were given calcein (6 mg/kg) at 14 days and demeclocycline (30 mg/kg) at 7 days before death at 18 and 44 weeks. After death, tibias were removed, bisected longitudinally in the medio-lateral plane, and fixed in 4% paraformaldehyde (pH 7.4) for 48 h. One-half was placed in 70% ethanol for further dehydration in graded ethanol and embedded undecalcified in methylmethacrylate. Sagittal sections (7 ␮m) were cut and stained with von Kossa or left unstained and examined by fluorescent microscopy. The other one-half and entire bones from 2- and 4-week time-points were decalcified in 14.5% EDTA/0.5% paraformaldehyde (pH 8.0) for 5 weeks at room temperature. Decalcification was confirmed by radiography, and the decalcified specimens were dehydrated and embedded in paraffin. Consecutive sagittal sections (5 ␮m) from paraffin blocks were stained with hematoxylin and eosin or TRACP for cellular histology and Safranin O with fast green or Van Geisons with picric acid for staining of cartilage and bone, respectively. Sections were analyzed using a digitizing tablet and BioQuant software (BIOQUANT; R&M Biometrics, Nashville, TN, USA). The region of the regenerate bone was examined in two sections to determine histomorphometric parameters.(24) Lamellar bone was visualized on hematoxylin and eosin– stained sections under polarized light (Leica DMRB; Leica Microsystems, Wetzlar, Germany).

Statistical methods Statistical analyses were performed by one-way ANOVA within groups or two-way ANOVA between groups with linear contrasts. All data are presented as mean ⫾ SD. The level of statistical significance was set at p ⬍ 0.05.

Radiography After death, hind limbs were removed with soft tissues intact. The limbs were oriented in standard anteriorposterior (AP), and lateral projections and plain radiographs were taken with a Siemens Multix H/UPH configuration

1699

RESULTS Radiography Regenerate radiographic density including larger external callus was greater in ZA-treated animals than controls at 6

1700

SMITH ET AL.

FIG. 1. Radiographic images of distracted tibias in AP plane: (A, C, and E) saline group and (B, D, and F) ZA-treated group. (A and B) At 6 weeks after surgery, ZA samples show denser regenerate regions (brackets). Original cortical bone ends resulting from the osteotomy (arrows) and pin sites were still apparent at 18 weeks after surgery in (C) the saline group and more clearly in (D) the ZA-treated group. (E and F) At 44 weeks after surgery, remodeling had progressed such that regenerate density and medulla formation are similar in saline- and ZAtreated groups. The two radiodense bands beneath the proximal physis (dashed arrows), still apparent at 18 and 44 weeks, correspond to doses of ZA.

weeks after surgery (Fig. 1). Regenerate trabeculae and the osteotomy-related cortical bone ends were still apparent in radiographs of ZA-treated animals at 18 weeks. However, by 44 weeks, remodeling of the medulla had progressed, and osteotomy-related bone ends were no longer visible, such that the mid-diaphyseal regions of the ZA-treated and control groups were indistinguishable.

BMD and BMC DXA measurements of the regenerate region revealed that BMD (g/cm2) increased from 2 to 4 weeks in both groups (61% in saline, 106% in ZA; p ⬍ 0.01), such that, at 4 weeks, ZA-treated BMD was 1.3-fold (p ⬍ 0.01) greater than those of saline animals. From 4 to 6 weeks, BMD decreased by 20% (p ⬍ 0.05) in the saline group but not in the ZA-treated group, such that, at 6 weeks, regenerate BMD was 1.6-fold (p ⬍ 0.01) greater in the ZA-treated group than the saline group. From 6 to 44 weeks, the difference between the ZA and saline groups gradually decreased, reducing the final difference between groups to 0.02 g/cm2 (Fig. 2). BMC (g) showed similar results to that of regenerate BMD, with the difference between groups being greater (1.3-fold, p ⬍ 0.05) at 18 weeks.

Histology At 2 weeks after surgery, there was a central region of fibrous tissue in all regenerate bone extending from the medial to lateral periosteum flanked proximally and distally by trabeculae and oriented in the line of tension and extending out to the cortical bone ends, which had been generated during the osteotomy (Figs. 3A and 3B). At 2, 4, and 6 weeks, the external callus tissue was more prominent on the lateral side, away from the monolateral external fixator. In all animals, cartilaginous matrix formed by 2 weeks on either side of the fibrous tissue in the external callus region. This was subsequently reduced to produce the endochondral

FIG. 2. BMD (g/cm3) and BMC (g) were significantly greater in the ZA-treated group than control animals up to 18 weeks. Convergence of values in the saline and ZA groups was such that, by 25 weeks and thereafter, there was no significant difference between the groups for BMD or BMC. **p ⬍ 0.01 and *p ⬍ 0.05 vs. saline. ##p ⬍ 0.01 and #p ⬍ 0.05 vs. previous time-point.

cartilaginous matrix, which by 4 and 6 weeks was gradually replaced by bone through endochondral ossification and further remodeled into trabeculae (Fig. 3). Endochondral ossification was indicated by the presence of cartilage, chondrocytes, and bone deposited within lacunae of the cartilaginous matrix (Figs. 3C and 3D). Removal of endochondral cartilaginous matrix was delayed in ZA-treated animals, evidenced by minimally reduced cartilage at 2 weeks and more extensive regions remaining at 4, 6, and 18

MODULATION OF ENDOCHONDRAL BONE REMODELING

1701

FIG. 3. Bone and cartilage histology. Safranin O and fast green staining. Macroscopic view of regenerate tissue in (A, E, I, M, and Q) saline and (B, F, J, N, and R) ZA-treated groups at 2, 4, 6, 18, and 44 weeks after surgery. Orientation is medial side uppermost, proximal end left. FT, fibrous tissue; TB, trabecular bone; CB, cortical bone; Cg, cartilage. Direction of distraction shown in A (doubleheaded arrow). (C and D) Endochondral ossification was evidenced by the presence of cartilage, chondrocytes, and lacunae bone deposition (arrows). Trabeculae seemed more remodeled in the proximal and distal regions of saline regenerates compared with corresponding region of ZA-treated samples (I vs. J; arrows) and trabeculae remained at 18 weeks in ZA-treated animals (N; arrow). Magnification, ⫻1.5; scale bar, 10 mm. High magnification images from boxed area in corresponding macroscopic view showing regenerate bone (blue-green) and cartilage (red) histology in (C, G, K, O, and S) saline and (D, H, L, P, and T) ZA-treated animals at 2, 4, 6, 18, and 44 weeks. Hematoxylin and eosin–stained sections under polarized light showing lamellae in both old cortical bone (arrow) and new bone (arrowheads) in (U) saline and (V) ZA-treated groups at 44 weeks after surgery. Magnification, ⫻200; scale bar, 100 ␮m. TRACP⫹ cells (arrows) at 2 weeks in (W) saline and (X) ZA-treated groups. Bone (b) and marrow spaces (m). Original magnification, ⫻200; scale bar, 100 ␮m.

weeks (Figs. 3A–3P). At 44 weeks, remnants of the endochondral cartilaginous matrix were still present in the new cortical bone of ZA-treated but not saline animals (Figs. 3S and 3T). Nevertheless, the regenerated bone of ZA-treated animals still had matured into lamellar bone at 44 weeks (Figs. 3U and 3V). Regions within the distraction gap that were not part of the external callus showed limited cartilage formation, suggesting bone formation was predominantly through direct intramembranous ossification in this area (Figs. 3A and 3B). Formation of the medulla through removal of trabeculae from the distraction gap was more advanced in saline than ZA-treated regenerates at 4 weeks and thereafter (Figs. 3E, 3F, 3I, and 3J). By 18 weeks, remnants of trabeculae remained in ZA-treated animals but were virtually absent in controls (Figs. 3M and 3N). At the last time-point studied (44 weeks), the overall regenerate histology in ZA-treated animals was comparable with controls, with the exception of cortical cartilaginous remnants (Figs. 3Q–3V).

TRACP⫹ cells in ZA-treated regenerates at 2 weeks were larger and contained more nuclei than osteoclastic cells in the regenerates of saline animals (Figs. 3W and 3X). This distinction was less marked at the later time-points studied.

Histomorphometry Bone volume (BV/TV) in saline animals transiently increased between 2 and 4 weeks and subsequently decreased to a value of 33% at 6 weeks (Table 1), consistent with the BMD and BMC patterns (Fig. 2). A further decrease led to a BV/TV at 18 weeks of 16% that was maintained at 44 weeks. In the ZA-treated group, BV/TV also increased from 2 to 4 weeks but was not significantly reduced between 4 and 6 weeks, leading to a BV/TV (53%) that was greater than the saline group. From 6 to 18 weeks, ZA-treated regenerate BV/TV decreased significantly, but remained 2-fold greater than the saline group at 18 weeks (31%). A further decrease brought the treated group to a final BV/TV

1702

SMITH ET AL.

TABLE 1. HISTOMORPHOMETRIC ANALYSIS Parameters Bone BV/TV (%)

Tb.N/mm

Tb.Th (␮m)

Cartilage Cg.V/TV (%)

Osteoclast Oc.S (%)

Oc.N/mm

Osteoblast Ob.S (%)

Ob.N/mm

MAR (␮m/day)

OF

REGENERATED BONE

Weeks (postsurgery)

Saline

Zoledronic acid

2 4 6 18 44 2 4 6 18 44 2 4 6 18 44

40.6 ⫾ 6.1 48.0 ⫾ 5.0 33.3 ⫾ 3.4‡ 15.5 ⫾ 9.6‡ 16.6 ⫾ 1.2 8.7 ⫾ 2.9 7.3 ⫾ 0.2 4.3 ⫾ 0.7§ 0.5 ⫾ 0.9‡ ND 51.0 ⫾ 20.8 65.5 ⫾ 6.6 79.2 ⫾ 10.0 17.5 ⫾ 16.4‡ ND

42.4 ⫾ 6.9 57.0 ⫾ 12.5 53.8 ⫾ 14.8* 30.6 ⫾ 5.3†‡ 22.9 ⫾ 8.4 10.2 ⫾ 0.8 7.3 ⫾ 0.3‡ 7.4 ⫾ 0.7* 2.0 ⫾ 0.4‡ 0.5 ⫾ 0.2 42.2 ⫾ 9.1 77.9 ⫾ 17.5 72.9 ⫾ 18.1 199.9 ⫾ 46.6*‡ 177.9 ⫾ 33.2

2 4 6 18 44

3.1 ⫾ 1.2 4.7 ⫾ 0.7§ 1.8 ⫾ 0.6‡ 0.1 ⫾ 0.1§ 0.1 ⫾ 0.0

6.1 ⫾ 1.8† 11.9 ⫾ 4.0*‡ 3.9 ⫾ 1.7‡ 3.3 ⫾ 1.0† 1.8 ⫾ 0.8

2 4 6 18 44 2 4 6 18 44

7.0 ⫾ 0.8 2.8 ⫾ 0.2§ 6.9 ⫾ 2.4§ 2.4 ⫾ 2.1‡ 4.3 ⫾ 4.9 2.1 ⫾ 0.4 1.1 ⫾ 0.1‡ 1.8 ⫾ 0.7§ 0.3 ⫾ 0.2‡ 0.9 ⫾ 0.3

6.2 ⫾ 0.9 1.8 ⫾ 0.8 5.6 ⫾ 1.6 8.1 ⫾ 5.6† 6.6 ⫾ 3.9 1.8 ⫾ 0.2 0.7 ⫾ 0.3‡ 1.4 ⫾ 0.3§ 1.0 ⫾ 0.5† 0.9 ⫾ 0.1

2 4 6 18 44 2 4 6 18 44 6 18 44

42.9 ⫾ 8.2 15.1 ⫾ 3.5‡ 25.8 ⫾ 15.0 ND ND 92.9 ⫾ 14.0 76.5 ⫾ 10.1 96.0 ⫾ 1.7 ND ND 1.9 ⫾ 0.2 2.0 ⫾ 0.4 1.8 ⫾ 0.2

46.4 ⫾ 9.2 8.0 ⫾ 3.6‡ 2.5 ⫾ 2.2* ND ND 74.6 ⫾ 4.0† 68.9 ⫾ 16.6 58.8 ⫾ 6.6* ND ND 1.3 ⫾ 0.2† 1.9 ⫾ 0.5 1.9 ⫾ 0.2

Values are mean ⫾ SD. * p ⬍ 0.01 vs. saline. † p ⬍ 0.05 vs. saline. ‡ p ⬍ 0.01 vs. previous time-point. § p ⬍ 0.05 vs. previous time-point. n ⫽ 4 for each group except n ⫽ 2 in each group at 44 weeks. ND, not detected.

of 22%, which was not significantly different from the saline group (Table 1). Trabecular number in both the saline and ZA-treated groups decreased from 6 to 44 weeks after surgery. How-

ever, the trabecular loss was less rapid in the ZA-treated group, which had significantly more trabeculae at 6 weeks (Table 1). Changes in trabecular thickness in the saline group paralleled BV/TV, with a trend to increase between 2 and 6 weeks and then decrease. There was a similar trend to increase at the early time-points in the ZA-treated animals, with no difference from the saline values until the 18- and 44-week time-points, when the trabeculae remaining in the ZA-treated regenerates were significantly thicker than the sparse trabecular remnants in the saline regenerates (Table 1). Endochondral cartilaginous matrix volume (CgV/TV) within the regenerate region was 2-fold greater (p ⬍ 0.05) in ZA-treated than saline-treated animals at 2 weeks. In the saline-treated group, CgV/TV increased between 2 and 4 weeks, followed by a substantial decrease at 6 and 18 weeks, to reach a final level of 0.1% that was maintained to 44 weeks (Table 1). There was a similar pattern in the ZA-treated group, but by 18 and 44 weeks, CgV/TV was still 2–3%. Osteoclast-covered bone surface (Oc.S) in the saline group decreased from 2 to 4 weeks, returned to initial (2 week) levels between 4 and 6 weeks, decreased again at 18 weeks, and ultimately increased to an intermediate level at 44 weeks (Table 1). The same trends were seen in the ZA-treated group in the 2- to 6-week period, with no differences from the Oc.S values in the saline groups at these time-points. In association with the marked difference in cancellous bone histology and the relative immaturity of the ZA-treated callus at 18 weeks (Figs. 3G–3H), osteoclast surface was significantly greater in the ZA-treated regenerates at this time-point (p ⬍ 0.05). Osteoclast number (Oc.N) changes were similar to Oc.S. Osteoblast-covered bone surface (Ob.S) was decreased at both 4 and 6 weeks compared with the 2-week values in both the saline and ZA-treated groups (Table 1). This parameter was not affected by ZA treatment until 6 weeks after surgery, when there was a reduction in the treated group. Osteoblast number (Ob.N) was lower in the ZAtreated group compared with the saline groups at 2 and 6 weeks. There was no measurable Ob.S or Ob.N at 18 and 44 weeks in either group. Mineral apposition rate (MAR) was similar over time in both groups at a rate of ⬃2 ␮m/day, other than a lower value in the ZA-treated group at 6 weeks (1.3 ␮m/day, p ⬍ 0.05; Table 1).

DISCUSSION Animal studies have indicated that bisphosphonates may be of benefit in the clinical situation where increased strength of fractured or distracted bone is required.(6 –13) In our previous rabbit study of distraction osteogenesis, ZA increased regenerate bone volume, BMC, and strength, while also preventing osteopenia at the 6-week time-point evaluated.(16) However, it was not clear in that study whether the increased mineralized callus volume was caused by an increase in production of bone or a lack of remodeling, and the effects were not evaluated in the long term. This study examined changes in regenerating

MODULATION OF ENDOCHONDRAL BONE REMODELING

bone by radiography, DXA, histology, and histomorphometry to investigate regenerate formation and remodeling over time to better understand the mechanism by which ZA alters short-term callus mass and strength during bone regeneration and to evaluate the consequences, if any, of these effects in the long term. The process of bone repair during distraction osteogenesis progressed in both saline-treated and ZA-treated animals; however, zoledronic acid increased the size and BMC of regenerate bone by delaying the temporal progression of bone healing at the endochondral cartilaginous matrix remodeling stage and thereafter. In both groups of animals, the mineralized cartilaginous matrix was removed, replaced by woven bone, and subsequently remodeled into lamellar bone of the cortices or removed to form the medulla, but these changes were delayed in ZA-treated animals. Thus, the ZA-induced increase in early callus mass and early strength of regenerated bone reported previously(16) was associated with a temporary retention of the mineralized endochondral cartilaginous matrix with subsequent deposition of bone on it, a consequence of inhibited osteoclast function. These results also confirm that normal osteoclast function is required for the removal or resorption of the mineralized cartilaginous matrix during endochondral bone formation in bone regeneration and repair. Importantly, however, despite the increased retention of mineralized cartilaginous matrix and slow removal of trabeculae from the medulla of treated animals, the regenerating bone still progressed into mature lamellar bone, allowing radiographic completion of the bone repair process. Delays in resorption and associated remodeling were evident in the slower resolution of the callus, fixator pin sites, radiodense lines in the metaphyses, and the original cortices in radiographs, as well as in the increased trabecular number of the ZA-treated groups. The delay in remodeling in treated animals was not associated with a reduction in osteoclast number or coverage of bone surfaces and thus is likely to be caused by osteoclast or TRACP⫹ cell dysfunction as reported previously for nitrogen-containing bisphosphonates.(25) The effects of ZA were not solely caused by transient retention of mineralized bone. The distraction gap fills by a combination of endochondral and intramembranous ossification depending on the region and degree of mechanical stability.(26 –29) In this study, the ZA-inhibited removal of endochondral cartilaginous matrix provided a larger scaffold on which osteoprogenitors deposited new bone. The mechanism for this reduction in matrix resorption may not be solely from inhibition of TRACP⫹ cell function. It may also relate to a “chondroprotective” effect of ZA, because nitrogen-containing bisphosphonates have been shown to prevent chondrocyte apoptosis,(30) which seems to be involved in degradation of the cartilaginous matrix.(31–33) Interestingly, there may have been active degradation of nonmineralized cartilaginous matrix by vascular endothelial cells, because these cells are capable of degrading nonmineralized cartilage in the presence of bisphosphonates. (34) However, this was

1703

clearly insufficient to remove the mineralized cartilaginous matrix to control levels. In contrast to observations in cell culture,(21–23) histomorphometry in this in vivo study did not provide any evidence for a direct anabolic effect of ZA. In fact, at the earlier time-points, osteoblast density was lower in ZA-treated animals, MAR was reduced, and there was no change in trabecular thickness. MAR and bone formation rate were similarly reduced in a ZA-treated model of traumatic avascular necrosis.(35) The osteoclastic disruption in the ZA bones may relate to reduced osteoblast recruitment and function, because bone repair relies on coordination between osteoblasts and osteoclasts.(36,37) The mechanical stimulation in distraction osteogenesis is a potent stimulant of an exuberant osteoblast response, and thus, it is difficult to observe any added stimulatory effects of an exogenous or systemic agent unless those added effects were very marked. The effect of ZA in our model was most apparent by 6 weeks and beyond, presumably as a result of both transient retention of the endochondral cartilaginous matrix and slower remodeling of the woven bone that formed on it. Importantly, despite the observed delay in the temporal progression of endochondral bone remodeling, regenerating bone of ZA-treated animals was radiographically healed and similar to controls by 44 weeks after surgery. Although endochondral cartilaginous matrix in trabeculae was easily removed as a result of medulla formation, a few remnants were still present in the cortical bone at 44 weeks after surgery, perhaps as a consequence of slower remodeling of cortical bone. Based on previous modeling studies, the increases in trabecular number at early time-points in the ZA-treated groups are likely to be responsible for the increase in strength at the 6-week time-point recorded previously.(16,38) In the clinical situation, a ZA-induced increase in callus strength early in the healing process may allow earlier frame removal, encouraging patient return to full activity, allowing callus remodeling under more normal loading conditions and potentially reducing disuse osteopenia. However, the long-term impact of cartilaginous remnants in cortical bone still requires consideration, because changes in bone homogeneity may alter bone strength. In summary, the mechanism by which noncontinuous treatment with a nitrogen-containing bisphosphonate transiently increases the amount of bone present during repair is through retention of a larger, possibly osteoinductive, endochondral cartilaginous substrate for bone deposition in the early stages and more gradual remodeling of trabeculae thereafter. Delay in the temporal progression of endochondral bone remodeling increased the amount of transiently retained bone, which in the early stages is stronger,(16) a time when, in the clinical situation, newly formed bone is often fragile and prone to re-fracture. Importantly, short-term treatment with the potent nitrogen-containing bisphosphonate ZA did not prevent completion of the bone repair process. The early increased callus strength and structure with normal, albeit delayed, remodeling and eventual progression to

1704

SMITH ET AL.

lamellar bone structure has clear clinical implications, not only in distraction osteogenesis, but also in healing of traumatic fractures and osteoporotic fractures in individuals on treatment with potent bisphosphonates.

ACKNOWLEDGMENTS We thank Rachael Bugler and Rachel Peat for technical assistance, Kathy Mikulec and Hao-Xu Lu for ethical care of the rabbits, Dr Elisabeth Gorgeons for performing surgical procedures, Madeleine Thompson and Julie Briody for performing the DXA scans, Drs Christopher Cowell and Nicholas Smith for stimulating discussions, and Associate Professor Daniel Cass for comments on the manuscript. This study was funded in part by grants from the Children’s Hospital at Westmead Orthopaedic Research Fund and Novartis Pharmaceuticals Pty Ltd. AM was funded by contributions from Smith & Nephew Surgical and Ingham Enterprises. ES was funded by the NH&MRC Dora Lush BioMedical Scholarship and the J Schougall Orthopaedic Scholarship.

REFERENCES 1. Russel RG, Rogers MJ 1999 Bisphosphonates: From the laboratory to the clinic and back again. Bone 25:97–106. 2. Fleisch H 1991 Bisphosphonates. Pharmacology and use in the treatment of tumour-induced hypercalcaemic and metastatic bone disease. Drugs 42:919 –944. 3. Seeman E 1997 Osteoporosis: Trials and tribulations. Am J Med 103:S74 –S89. 4. Storm T, Thamsborg G, Steiniche T, Genant HK, Sorensen OH 1990 Effect of intermittent cyclical etidronate therapy on bone mass and fracture rate in women with postmenopausal osteoporosis. N Engl J Med 322:1265–1271. 5. Hosking DJ, Eusebio RA, Chines AA 1998 Paget’s disease of bone: Reduction of disease activity with oral risedronate. Bone 22:51–55. 6. Nyman MT, Paavolainen P, Lindholm TS 1993 Clodronate increases the calcium content in fracture callus. An experimental study in rats. Arch Orthop Trauma Surg 112:228 –231. 7. Fleisch H 2001 Can bisphosphonates be given to patients with fractures? J Bone Miner Res 16:437– 440. 8. Goodship AE, Walker PC, McNally D, Chambers T, Green JR 1994 Use of a bisphosphonate (pamidronate) to modulate fracture repair in ovine bone. Ann Oncol 5(Suppl):S53–S55. 9. Nyman MT, Gao T, Lindholm TC, Lindholm TS 1996 Healing of a tibial double osteotomy is modified by clodronate administration. Arch Orthop Trauma Surg 115:111–114. 10. Peter CP, Cook WO, Nunamaker DM, Provost MT, Seedor JG, Rodan GA 1996 Effect of alendronate on fracture healing and bone remodelling in dogs. J Orthop Res 14:74 –79. 11. Madsen JE, Berg-Larsen T, Kirkeby OJ, Falch JA, Nordsletten L 1998 No adverse effects of clodronate on fracture healing in rats. Acta Orthop Scand 69:532–536. 12. Li J, Mori S, Kaji Y, Kawanishi J, Akiyama T, Norimatsu H 2000 Concentration of bisphosphonate (incadronate) in callus area and its effects on fracture healing in rats. J Bone Miner Res 15:2042– 2051. 13. Li J, Mori S, Li J, Kaji Y, Akiyama T, Kawanishi J, Norimatsu H 2002 Long term effect of icandronate disodium (YM-175) on fracture healing of femoral shaft in growing rats. J Bone Miner Res 16:429 – 436. 14. Little DG, Cornell MS, Briody J, Cowell CT, Arbuckle S, CookeYarborough CM 2001 Intravenous pamidronate reduces osteoporosis and improves formation of the regenerate during distraction osteogenesis. A study in immature rabbits. J Bone Joint Surg Br 83:1069 –1074. 15. Little DG, Cornell MS, Hile MS, Briody J, Cowell CT, Bilston L 2001 Effect of pamidronate on distraction osteogenesis and fixatorrelated osteoporosis. Injury 32(Suppl):SD14 –SD20.

16. Little DG, Smith NC, Williams PR, Broidy JN, Bilston LE, Smith EJ, Gardiner EM, Cowell CT 2003 Zoledronic acid prevents osteopenia and increases bone strength in a rabbit model of distraction osteogenesis. J Bone Miner Res 18:1300 –1307. 17. Simpson AH, Kenwright J 2000 Fracture after distraction osteogenesis. J Bone Joint Surg Br 82:659 – 665. 18. Danziger MB, Kumar A, DeWeese J 1995 Fractures after femoral lengthening using the Ilizarov method. J Pediatr Orthop 15:220 – 223. 19. Schenk R, Merz WA, Muhlbauer R, Russell RG, Fleisch H 1973 Effect of ethane-1-hydroxy-1,1-diphosphonate (EHDP) and dichloromethylene diphosphonate (Cl 2 MDP) on the calcification and resorption of cartilage and bone in the tibial epiphysis and metaphysis of rats. Calcif Tissue Res 11:196 –214. 20. Larsson A, Larsson SE 1978 The effects of ethylene-1-hydroxy1,1-diphosphonate on cellular transformation and organic matrix of the epiphyseal growth plate of the rat—a light microscopic and ultrastructural study. Acta Pathol Microbiol Scand 86:211–223. 21. Reinholz GG, Getz B, Pederson L, Sanders ES, Subramaniam M, Ingle JN, Spelsberg TC 2000 Bisphosphonates directly regulate cell proliferation, differentiation, and gene expression in human osteoblasts. Cancer Res 60:6001– 6007. 22. Giuliani N, Pedrazzoni M, Negri G, Passeri G, Impicciatore M, Girasole G 1998 Bisphosphonates stimulate formation of osteoblast precursors and mineralized nodules in murine and human bone marrow cultures in vitro and promote early osteoblastogenesis in young and aged mice in vivo. Bone 22:455– 461. 23. Giuliani N, Pedrazzoni M, Passeri G, Negri G, Impicciatore M, Girasole G 1998 Bisphosphonates stimulate the production of basic fibroblast growth factor and the formation of bone marrow precursors of osteoblasts. New findings about their mechanism of action. Minerva Med 89:249 –258. 24. Parfitt AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H, Meunier PJ, Ott SM, Recker RR 1987 Bone histomorphometry: Standardization of nomenclature, symbols, and units. Report of the ASBMR Histomorphometry Nomenclature Committee. J Bone Miner Res 2:595– 610. 25. Coxon FP, Helfrich MH, van’t Hof R, Sebti S, Ralston SH, Hamilton A, Rogers MJ 2000 Protein geranylgeranylation is required for osteoclast formation, function, and survival: Inhibition by bisphosphonates and GGTI-298. J Bone Miner Res 15:1467– 1476. 26. Li G, Simpson AH, Triffitt JT 1999 The role of chondrocytes in intramembranous and endochondral ossification during distraction osteogenesis in the rabbit. Calcif Tissue Int 64:310 –317. 27. Li G, Virdi AS, Ashhurst DE, Simpson AH, Triffitt JT 2000 Tissues formed during distraction osteogenesis in the rabbit are determined by the distraction rate: Localization of the cells that express the mRNAs and the distribution of types I and II collagens. Cell Biol Int 24:25–33. 28. Aronson J, Harrison BH, Stewart CL, Harp JH Jr 1989 The histology of distraction osteogenesis using different external fixators. Clin Orthop 241:106 –116. 29. Ilizarov GA 1989 The tension-stress effect on the genesis and growth of tissues. Part I. The influence of stability of fixation and soft-tissue preservation. Clin Orthop 239:263–285. 30. Van Offel JF, Schuerwegh AJ, Bridts CH, Stevens WJ, De Clerck LS 2002 Effect of bisphosphonates on viability, proliferation, and dexamethasone-induced apoptosis of articular chondrocytes. Ann Rheum Dis 61:925–928. 31. Einhorn TA, Lee CA 2001 Bone regeneration: New findings and potential clinical applications. J Am Acad Orthop Surg 9:157– 165. 32. Lee FY, Choi YW, Behrens FF, DeFouw DO, Einhorn TA 1998 Programmed removal of chondrocytes during endochondral fracture healing. J Orthop Res 16:144 –150. 33. Hiraoka N, Apel IJ, Gyetko MR, Weiss SJ, Allen E 1998 Matrix metalloproteinases regulate neovascularization by acting as pericellular fibronolysins. Cell 95:365–377. 34. Deckers MML, Van Beek ER, Van der Pluijm G, Wetterwald A, van der Wee-pals L, Cecchini MG, Papapoulos SE, Lowik CWGM 2002 Dissociation of angiogenesis and osteoclastogenesis during endochondral bone formation in neonatal mice. J Bone Miner Res 17:998 –1007.

MODULATION OF ENDOCHONDRAL BONE REMODELING 35. Little DG, Peat RA, McEvoy A, Williams PR, Smith EJ, Baldock PA 2003 Zoledronic acid treatment results in retention of femoral head structure after traumatic osteonecrosis in young Wistar rats. J Bone Miner Res 18:2016 –2022. 36. Jimi E, Nakamura I, Amano H, Taguchi Y, Tsurukai T, Tamura M, Takahashi N, Suda T 1996 Osteoclast function is activated by osteoblastic cells through a mechanism involving cell-to-cell contact. Endocrinology 137:2187–2190. 37. Vaes G 1988 Cellular biology and biochemical mechanism of bone resorption. A review of recent developments on the formation, activation, and mode of action of osteoclasts. Clin Orthop 231: 239 –271. 38. Silva MJ, Gibson LJ 1997 Modeling the mechanical behavior of vertebral trabecular bone: Effects of age-related changes in microstructure. Bone 21:191–199.

1705

Address reprint requests to: Elisabeth J Smith, MSc Bone and Mineral Research Program Garvan Institute of Medical Research 384 Victoria Street Darlinghurst Sydney, NSW 2010, Australia E-mail: [email protected]

Received in original form December 21, 2003; in revised form May 3, 2004; accepted May 21, 2004.