Osteoclast inhibition impairs chondrosarcoma ... - Wiley Online Library

Osteoclast Inhibition Impairs Chondrosarcoma Growth and Bone Destruction Jesse E. Otero,1 Jeff W. Stevens,2 Allison E. Malandra,3 Douglas C. Fredericks,3 Paul R. Odgren,4 Joseph A. Buckwalter,5,6 Jose Morcuende7 1

Department of Orthopaedic Surgery, University of Iowa, 200 Hawkins Drive, 01051 JPP, Iowa City, Iowa 52242, 2Department of Internal Medicine, University of Iowa School of Medicine, 500 Newton Road, Bldg. Medical Laboratories 3160, Iowa City, IA 52242, 3Department of Orthopaedic Surgery Bone Healing Research Laboratory, University of Iowa, 2662 Crosspark Road, MTP 4, Coralville, Iowa 52241, 4Department of Cell and Developmental Biology, S7-242, University of Massachusetts Medical School, 55 Lake Ave North, North Worcester, MA 01655, 5 Department of Orthopaedic Surgery, University of Iowa, 200 Hawkins Drive, 01023 JPP, Iowa City, Iowa 52242, 6Iowa City Veterans Administration Medical Center, 601 Highway 6 West, Iowa City, Iowa 52242, 7University of Iowa Department of Orthopaedic Surgery, 200 Hawkins Drive, 01023 JPP, Iowa City, Iowa 52242 Received 20 March 2014; accepted 14 July 2014 Published online 13 August 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jor.22714

ABSTRACT: Because Chondrosarcoma is resistant to available chemotherapy and radiation regimens, wide resection is the mainstay in treatment, which frequently results in high morbidity and which may not prevent local recurrence. There is a clear need for improved adjuvant treatment of this malignancy. We have observed the presence of osteoclasts in the microenvironment of chondrosarcoma in human pathological specimens. We utilized the Swarm rat chondrosarcoma (SRC) model to test the hypothesis that osteoclasts affect chondrosarcoma pathogenesis. We implanted SRC tumors in tibia of Sprague-Dawley rats and analyzed bone histologically and radiographically for bone destruction and tumor growth. At three weeks, tumors invaded local bone causing cortical disruption and trabecular resorption. Bone destruction was accompanied by increased osteoclast number and resorbed bone surface. Treatment of rats with the zoledronic acid prevented cortical destruction, inhibited trabecular resorption, and resulted in decreased tumor volume in bone. To confirm that inhibition of osteoclasts per se, and not off-target effects of drug, was responsible for the prevention of tumor growth and bone destruction, we implanted SRC into osteopetrotic rat tibia. SRC-induced bone destruction and tumor growth were impaired in osteopetrotic bone compared with control bone. The results from our animal model demonstrate that osteoclasts contribute to chondrosarcoma-mediated bone destruction and tumor growth and may represent a therapeutic target in particular chondrosarcoma patients. ß 2014 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 32:1562–1571, 2014. Keywords: chondrosarcoma; osteoclast; bisphosphonate; osteolysis

Chondrosarcoma is highly resistant to currently available chemotherapy and radiation treatment regimens.1–3 As a result, surgery remains the cornerstone in CHS treatment. Several studies have demonstrated the benefit of resection of tumor with wide surgical margins.4–6 However, even with adequate margins, there is a high risk of local recurrence with high-grade as well as low-grade tumors.5 Furthermore, up to 20– 50% of patients with high-grade lesions develop distant metastasis.5,7 Overall 10-year survival for patients with chondrosarcoma depends on grade of tumor and adequacy of surgical margins, with only 30–50% of patients surviving who have a high-grade Disclosure: None of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of any aspect of this work. None of the authors, or his or her institution, has had a financial relationship, in the thirty-six months prior to submission of this work, with an entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. No author has had any other relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article. This work was funded through the Orthopaedic Research and Education Foundation Resident Clinician Scientist Training Grant Correspondence to: Jesse E. Otero (T: 319-356-1616; F: 319-3536754; E-mail: [email protected]) # 2014 Orthopaedic Research Society. Published by Wiley Periodicals, Inc.

1562

JOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014

tumor or positive margins after resection.4–6,8 Chondrosarcoma which is adequately resected has a much higher cure rate than disease for which wide surgical margins cannot be obtained.4–6,9 When distant metastasis develops, the disease is almost invariably fatal. This fact highlights the importance of local control of the tumor in the treatment of CHS. Therapies which inhibit local recurrence and distant advancement of this tumor would undoubtedly improve outcomes. We have observed consistently that chondrosarcoma induces a local osteolytic reaction with bone remodeling at the site of tumor on gross and histopathological specimens. The presence of Howship’s lacunae in bone of the tumor microenvironment suggests that this phenomenon is mediated by the action of osteoclasts, monocyte-derived bone resorbing cells.10 It was previously shown that cathepsin K, a specific osteoclast marker, is expressed at disproportionate levels in a mammal model for chondrosarcoma.11 We postulated that osteoclasts in the tumor microenvironment could be exploited to prevent tumor-mediated bone destruction and control tumor growth. The Swarm rat chondrosarcoma (SRC)12 was used in the present study to test the hypothesis that osteoclasts contribute to the growth of-and bone destruction mediated by- chondrosarcoma. Importantly, zoledronic acid, an inhibitor of osteoclasts has been shown previously to directly inhibit SRC cell cycle progression, induce apoptosis, and to reduce peri-skeletal tumor growth in vivo. However, the study did not address effects of osteo-

CHONDROSARCOMA GROWTH AND BONE DESTRUCTION

clasts per se, on chondrosarcoma growth. Nor did the referenced studies use a true skeletally-based tumor model.13 We sought to examine the effect of osteoclast inhibition on chondrosarcoma growth and bone destruction in rat tibia using a SRC implantation protocol we developed. We employed a pharmacologic inhibition strategy using zoledronic acid and confirmed the direct contribution of osteoclasts to bone-destruction by tumor and skeletal growth of tumor using a genetic approach with osteopetrotic rats.

METHODS Animals All animal experimentation was performed in accordance with the Guide for the Care and Use of Laboratory Animals, National Institutes of Health, and was approved by the University of Iowa Institutional Animal Care and Use Committee (IACUC). For experimentation purposes, Sprague-Dawley 4 week-old male wild-type rats were purchased from Harlan Laboratories (Indianapolis, IN, USA). Osteopetrotic rats and control littermates were from the Fischer background tl colony maintained at the University of Massachusetts Medical School as described previously.14 For experiments involving osteopetrotic rats, mutants and littermate controls were fed ground chow. Rat Chondrosarcoma Model The rat tibial chondrosarcoma model used was described previously.11 Briefly, a slurry of 1 106 Swarm rat chondrosarcoma cells was injected subcutaneously into the flank of 4 week-old male Sprague-Dawley rats. After 10–14 days, when tumor reached 1 cm in greatest cross-sectional diameter, tumor-bearing rats were euthanized and tumor was harvested for implantation. Experimental rats, 7 week-old males, were anesthetized and the right leg was prepped sterilely and draped. A midline incision was made distal to the knee joint, and the hamstrings insertion on the medial proximal tibia was sharply released and subperiosteally elevated. A 2 mm3 corticotomy in the medial tibial metadiaphysis was created with a high-speed pencil tip bur. The bur hole was carried through the entire medial cortex, and a 1 mm curette was used to smoothen the cortical bone window to allow easy implantation of tumor. We chose tumor block implantation rather than SRC cell slurry injection into the bone because injection of tumor cell slurry into the medullary cavity results in immediate colonization of the lungs with metastatic tumors, which compromises the health of animals.15 A 2 mm3 block of tumor was implanted into the bone defect, and the cortex was sealed with bone wax (Ethicon, Somerville, NJ, USA). Control rats had no tumor placed into the corticotomy. Skin was closed with nylon suture. Rats were sacrificed at day 21 post-operatively and tibias were harvested for analysis.

1563

tration of 0.1 mg/kg. In the first experiment, 4 weekold rats were injected intravenously weekly for three weeks prior to tumor implantation at age 7 weeks. A total of 15 rats were utilized in this experiment, with three groups containing 5 rats each. Group 1: PBS injection with Sham surgery (corticotomy without tumor implantation); Group 2: PBS injection with SRC implantation; Group 3: ZA injection, SRC implantation In a second experiment, 7 week-old rats underwent tumor implantation, and beginning on post-operative day 4, they received weekly tail vein injections of zoledronic acid until sacrifice at day 21 post-implantation, for a total of three injections. Two groups of rats were used in this experiment with 5 rats in each group. Group 1: SRC implantation followed by PBS injection; Group 2: SRC implantation followed by ZA injection. In this experiment, right tibia underwent tumor implantation, and left tibia underwent sham corticotomy, therefore, serving as internal control. Toothless Rat Experiment Toothless rats and their normal littermate controls were subjected to tibial corticotomy and SRC implantation into the right tibia. Left tibia served as sham control. Five animals were utilized in each group. Serum Collagen Crosslink Measurement Prior to tumor implantation, and at four to seven day intervals, 300 ml whole blood was collected from rats by tail artery puncture and placed in microcentrifuge tubes. Serum was separated by high-speed centrifugation and stored at 80˚C. Carboxy-terminal collagen crosslink concentration was measured with RatLaps enzyme-linked immunosorbent assay (Immuno Diagnostic Systems, Scottsdale, AZ, USA). Tissue Processing and Histology On day 21 after tumor implantation, rats were euthanized, and tibias were dissected free from surrounding soft tissue. Bones were placed in 10% neutral buffered formalin for 24 h. For histological analysis, the bones were then decalcified in 14% EDTA (Sigma, St. Louis, MO, USA) until complete decalcification, determined by x-ray. Bones were then dehydrated with graded ethanol (to 70%), cleared through xylene, and embedded in paraffin for sectioning. Slides were stained with hematoxylin and eosin (H&E), tartrate-resistant acid phosphatase (TRAP), and safranin O. TRAP stains osteoclasts fuscia. Safranin O stains SRC tumor orange. Histological images were obtained with the Olympus BX-61 light microscope (Olympus, Tokyo, Japan).

Experiments Zoledronic Acid Treatment Zoledronic acid (ZA) (Sigma, St. Louis, MO, USA) powder was diluted in sterile phosphate-buffered saline (PBS) at the time of administration. Sterile PBS served as the injection in control groups. Rat tail-veins were injected with zoledronic acid for a total concen-

Histomorphometry Tumor Size Measurement: Mid-sagittal sections of rat tibia possessing implanted tumor were obtained from each rat in the study. Three mid-sagittal/parasagittal sections were analyzed per limb. Tumor perimeter was drawn manually on 10 magnification, and enclosed


1564

OTERO ET AL.

area was calculated using the area tool on cellSens software (Olympus). The tumor area from the three sections was averaged. A mean of averages from the 5 animals was calculated and groups (control versus ZA or tl) were compared statistically as described below. Calculation of proportion resorbed surface: H&E stained sections of all tumor-bearing tibias were analyzed at 20 magnification. Five separate fields were analyzed per bone. Using Olympus cellSens software, the bone surface was marked with a line. Smooth lines were scored as a non-resorbed surface. Serrated or irregular surface was scored as resorbed surface. The length of resorbed surface was divided by the length of the total surface (smooth þ resorbed) to give resorbed surface/total surface presented as a proportion. Osteoclast Number: TRAP-stained sections were analyzed at 20 magnification. All TRAP-positive cells were counted in five separate 20 fields. Averages were taken from the five fields and the mean of the samples was calculated. Radiographic Analysis Rats were radiographed preoperatively and postoperatively at the time of euthanasia with a SimonDR (Quantum, Ronkonkoma, NY) RAD-X High Frequency Radiographic Imaging System, model E7242X. Images were stored using WhiteCap PACs system. Bone mineral density was determined using peripheral quantitative CT analysis with the STRATEC XCT3000 pQCT (Stratec Medizintechnik, Germany). Bones were dissected of soft tissue and stored in 15 ml conical tubes with 10% neutral buffered formalin. Two millimeter slices were scanned 4 mm, 7 mm, and 10 mm distal to the knee joint line. Micro CT images were obtained using a Skyscan 1076 microCT, and images were analyzed with Skyscan control software (Bruker MicroCT, Kontich, Belgium). Statistical Analysis All statistical comparisons performed in the study were accomplished with the student’s t-test assuming a two-tailed null hypothesis for difference between means. A p-value less than 0.05 was considered statistically significant.

RESULTS Swarm Rat Chondrosarcoma Induces Local Osteoclast Activation and Bone Resorption SRC has been shown previously to induce local osseous destruction.16 It has been unclear whether the tumor itself causes bone erosion or whether the tumor induces bone loss indirectly through other factors in the microenvironment. To determine the association between tumor-induced bone loss and local osteoclast activation, we implanted SRC tumor into the proximal tibia of Sprague-Dawley rats. Three weeks after tumor implantation, x-rays demonstrated both metaphyseal osteolysis and cortical destruction around the site of tumor implantation. Tibias with a sham corticotomy JOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014

demonstrated preservation of local cortical and cancellous bone [Fig. 1A]. Histological sections showed that in the presence of tumor, there was an increase in osteoclast number (p ¼ 0.0002) and an accompanying increase in resorbed trabecular surface (p ¼ 0.02) in the tibial metaphysis compared with the metaphysis of tibias with a sham corticotomy [Fig. 1B and C]. Furthermore, at 21 days post-SRC tumor implantation, peripheral quantitative CT (PQCT) demonstrated decreased bone mineral density surrounding tumor compared with sham corticotomy (p ¼ 0.008) [Fig. 1D]. These data altogether highlight the association between osteoclast activation in the microenvironment of SRC and local bone destruction. Pharmacologic Inhibition of Osteoclasts Prevents Local Bone Loss Caused by Chondrosarcoma With establishment that SRC induces both osteoclast activation and bone destruction, we sought to test whether pharmacologic inhibition of osteoclasts prevents bone loss in the presence of tumor. To this end, Sprague-Dawley rats were treated with intravenous ZA, an amino-bisphosphonate with potent anti-osteolytic activity used to treat bone pain and prevent skeletal tumor burden in metastatic breast carcinoma.17,18 The drug effect was apparent physiologically after only two injections with increased bone density most apparent in metaphyseal segments of long bones [Fig. 2A]. Three weeks after tumor implantation, radiographic evidence of osteolysis was apparent in SRC-implanted tibias, while there was no bone destruction surrounding implant sites in ZA-treated bones [Fig. 2B]. Importantly, at two weeks post operatively, there was decreased concentration of serum carboxy-terminal collagen breakdown products in ZA-treated tumor-bearing rats compared with nonZA treated animals [Fig. 2C], offering biochemical evidence of an association between osteoclast inhibition and prevention of tumor-mediated bone loss. Three weeks after tumor implantation, we analyzed tumor-bearing tibias of rats treated with vehicle and ZA. There was a significant decrease in the proportion of resorbed trabecular surface (0.05 0.01 compared with 0.36 0.04, p ¼ 0.005) and number of osteoclasts per high powered field (22.2 2.5 compared with 37 4.8, p ¼ 0.005) surrounding tumor in ZA-treated compared with PBS-treated bones [Fig. 2D]. In untreated rats, SRC tumor infiltrated the medullary cavity of the diaphysis, replacing normal marrow elements, and invaded metaphyseal bone leading to trabecular rarefaction. Additionally, the SRC tumor eroded through cortical bone. In ZA treated rats, metaphyseal trabeculae were preserved, and cortical bone remained intact [Fig. 3]. Furthermore, PQCT showed preservation of bone mineral density proximal to the site of tumor implantation in ZA-treated rats [Fig. 4A]. Micro-CT showed that in the absence of ZA, SRC destroyed surrounding trabeculae and cortical bone. In the presence of ZA, the tumor implantation


1565

Figure 1. Swarm rat chondrosarcoma induces osteoclast activation and bone loss. (A) Sham tibia (left panel) and tibia implanted with Swarm rat chondrosarcoma (SRC) (right panel) demonstrating cortical bone destruction (arrows) proximal to the tumor implantation site three weeks after tumor implantation. (B) Tartrate-resistant acid phosphatase-stained (TRAP) histological sections at 20 magnification of sham (left panel) and SRC-implanted (right panel) tibias demonstrating increased number and size of osteoclasts (arrows) in the presence of SRC. Osteoclasts stain fuscia with TRAP stain. Inset represents same area stained with safranin O, which stains cartilage proteoglycans orange/red, marking the tumor. Scale bars in lower right demonstrate equivalent magnification. (C) Histomorphometric analysis demonstrating increased number of osteoclasts (OC) ( p ¼ 0.0002) and increased proportion of resorbed bone surface (sfc) ( p ¼ 0.02) in SRC-implanted tibias compared with sham tibias. (D) Peripheral quantitative computed tomography PQCT analysis for bone mineral density in metaphyseal bone proximal to corticotomy of sham surgical tibias and SRC-implanted tibias (Right) compared with the non-operative contralateral limbs (Left). Sham surgery had no effect on local bone mineral density while there was a significant reduction in bone mineral density in the presence of SRC ( p ¼ 0.008).

site appeared sealed off by a sclerotic bony reaction [Fig. 4B]. These data establish that SRC relies upon the activity of osteoclasts to destroy local bone. To address the possibility that ZA pre-treatment prevents SRC tumor implantation, indirectly influencing tumor effects on bone, we first implanted tumor into the proximal tibia of rats and then treated with either intravenous vehicle or ZA beginning four days post-operatively. At three weeks, vehicle-treated tibias displayed typical destructive changes on plain radiographs, while those treated with ZA demonstrated preservation of bone around the tumor implantation site [Fig. 5A]. Histologically, tumors in control rats infiltrated proximal to the implantation site destroying metaphyseal trabeculae and eroded through diaphyseal cortical bone. Treatment with ZA after tumor implantation prevented bone destruction by tumor and inhibited local advancement of the tumor throughout the bone [Fig. 5B]. Importantly, tumors were signifi-

cantly smaller with ZA treatment compared with controls (5.3 1.7 mm2 compared with 16.6 3.0 mm2, p ¼ 0.02) [Fig. 5C]. Analysis of BMD in the metaphysis proximal to corticotomy with PQCT showed preservation of bone mass around the tumor implantation site in ZA-treated rats [Fig. 5D]. SRC Growth and Bone Destruction are Impaired in Osteopetrotic Rats Pharmacologic experiments with ZA showed clear evidence that inhibition of osteoclasts hampered tumor growth in bone and prevented tumor-mediated bone destruction. These experiments left the possibility that the drug potentially affects tumor activity directly. We pursued a genetic model of osteoclast inhibition to address this issue. The toothless (tl) rat strain possesses a naturally occurring mutation in the Csf-1 gene that leads to failure of osteoclast differentiation and osteopetrosis.14 We utilized this rat strain as a JOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014

1566

OTERO ET AL.

Figure 2. Zoledronic acid prevents tumor-mediated bone resorption and reduces osteoclast activity in the presence of SRC. (A) Two weeks of intravenous zoledronic acid (ZA) administration results in radiographically apparent increased bone density. The most affected sites are areas of high bone turnover distal to epiphyseal growth plates in the long bones. Phosphate buffered saline (PBS) was the solvent for ZA and served as a control vehicle injection. (B) Plain radiographs of operative tibia from each experimental group. Metaphyseal and diaphyseal bone is eroded around the tumor implantation site (arrows). Sclerotic bone fills the metaphysis in the presence of ZA and is impenetrable by tumor. (C) Serum collagen carboxy-terminal crosslink [CTX] concentration serves as a marker for bone resorption. There was a significant reduction in serum [CTX] in ZA-treated SRC-bearing rats compared with SRC-bearing rats treated with PBS, ( p ¼ 0.009). (D) Histomorphometric analysis of rat tibias shows that Zoledronic acid (ZA) administration results in significant reduction in osteoclast (OC) number and proportion of resorbed bone surface (sfc) in bone implanted with SRC ( p ¼ 0.005).

genetic model to test whether osteoclasts per se, and not off-target effects of bisphosphonate, are responsible for ZA mediated inhibition of bone destruction by SRC and tumor growth in bone. To this end, we implanted SRC tumors into a right tibial corticotomy of tl rats and normal littermate controls. A sham left tibial corticotomy was used in all rats as an internal control. At 21 days, tibias were analyzed by micro CT and with histopathology. On micro CT analysis of normal littermate rat tibias, there was loss of trabecular bone in the metaphysis proximal to the tumor implantation site in all specimens compared with control tibias, corresponding to a significant decrease in total metaphyseal bone content [Fig. 6A and B]. Furthermore, there were numerous sites of micro-erosion into cortical bone surrounding the tumor implantation site [Fig. 6A]. In contrast, tl rat tibias, which possess a genetic defect in osteoclast development, showed no loss of bone surrounding the tumor implantation site compared with sham tl tibias [Fig. 6B]. Correspondingly, tumors JOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014

implanted into osteopetrotic bone were unable to advance beyond the corticotomy site, while in normal rats, the tumor invaded proximally into the metaphysis replacing marrow and destroying trabecular bone [Fig. 6C]. Furthermore, tumors were significantly smaller in toothless rat bones compared with normal littermate controls (3.3 0.53 mm2 compared with 17.8 2.0 mm2, p ¼ 0.0002).

DISCUSSION Current treatment of chondrosarcoma involves wide surgical excision and reconstruction because chondrosarcoma is notoriously resistant to chemotherapy and radiation regimens.2,3 Even with surgical treatment, local recurrence, metastasis and death do occur in a subset of patients with chondrosarcoma.5 For this reason, significant research effort has been dedicated to understand the biological basis for chondrosarcoma pathogenesis and identify pharmacologic targets to aid in adjuvant treatment.18–25 Despite increasing under-


1567

Figure 3. Osteoclast inhibition prevents SRC mediated bone destruction. Safranin O (Saf O) and corresponding TRAP-stained histological sections of SRC-bearing tibias treated with vehicle control (Veh) and ZA. Swarm rat chondrosarcoma tumor (T), which stains orange/red by Safranin O, invades and erodes local bone (b). With ZA treatment, bone (b) is essentially unaffected by tumor. Note the absence of osteoclasts at the tumor-bone interface with ZA treatment. In the presence of ZA, osteoclasts (arrows) were small and rounded (inset), commensurate with amino-bisphosphonate-mediated inhibition of proteins involved in osteoclast attachment.

standing of chondrosarcoma biology, there are no available chemotherapeutic strategies to improve surgical outcomes. In the current study, we utilized SRC, an established in vivo model,11,12,16,19–22 to study mechanisms of chondrosarcoma pathogenesis. With the current difficulty in targeting chondrosarcoma directly, we sought to focus on microenvironmental factors, which could potentially be exploited to modulate tumor behavior.11 The radiographic characteristics of CHS offer clues regarding the tumor’s interaction with the microenvironment. Chondrosarcoma of bone possesses an aggressive radiographic appearance characterized by osteopenia, cortical erosion, and potentially, pathologic fracture. Computed tomography further characterizes the destructive radiographic phenotype, which has been shown to reflect the pathologic appearance of the tumor under the microscope.23 Trabecular erosion, cortical invasion and remodeling, as well as endosteal scalloping are the radiographic footprint of CHS in its bone environment. Based on this observation, we hypothesized that CHS activates osteoclasts in the tumor microenvironment,

which in turn resorb local bone. Inherent in this hypothesis is the assertion that the tumor itself lacks the ability to resorb bone, contrary to previous hypotheses.11 Indeed, the osteoclast has been described as the sole bone-resorbing cell.10 We utilized the SRC model to characterize and quantify osteoclast activation in the microenvironment of tumor. We showed an increase in osteoclast number in the presence of SRC. There was a concomitant increase in resorbed bone surface, implicating osteoclasts in the destructive process. Importantly, the histological data correlated with quantitative radiographic data, with reduced bone mineral density proximal to the implantation site in the presence of tumor compared with sham corticotomy. Qualitatively, bone harboring SRC appeared radiographically similar to the human tumor, with obvious bone loss on plain x-ray, and trabecular destruction, cortical micro-invasion, endosteal erosion and periosteal reaction on micro-CT. In a recent report describing the pathological features of human chondrosarcoma, authors observed a JOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014

1568

OTERO ET AL.

Figure 4. Computed tomography analysis. (A) Zoledronic acid treatment results in increased bone mineral density (BMD) and prevents SRC-mediated bone destruction. Left tibia served as non-operative internal control to the operative right tibia. Sham animals underwent corticotomy of the right tibia without SRC implantation. BMD was measured in the metaphysis proximal to the corticotomy site. (B) Mid-sagittal reconstructions of micro-CT images. Corticotomy sites are denoted with an arrow. Note the expansion of the bone cavity surrounding the corticotomy site in the presence of SRC. The edges of the corticotomy are well visualized owing to a sclerotic reaction. In ZA-treated rats, the tumor is confined within the corticotomy site, which is surrounded by sclerotic reactive bone. Asterisks denote periosteal bone formation induced by SRC.

tight association between the presence of tumor and osteoclast ‘resorption foci’ in bone.24 They concluded that microenvironmental factors including osteoclasts likely contribute to the pathogenesis of CHS. We sought to test the hypothesis that osteoclast inhibition would prevent bone destruction by tumor. We took advantage of two established models for diminished osteoclast function: Inhibition of osteoclasts with aminobisphosphonates25 and genetic deficiency of osteoclasts with an osteopetrotic rat strain.14 In both settings, there was complete abrogation of bone destruction in the presence of SRC. More importantly, tumor size was markedly diminished in both models of osteoclasts inhibition. Data from our animal model provide pharmacologic and genetic evidence that osteoclasts are solely responsible for the bone destructive behavior of CHS and that osteoclast activity may contribute to the growth of chondrosarcoma in bone. Indeed, previous work in the SRC model showed increased histological grade of tumor when implanted in bone compared with implantation in soft tissues.19 It is possible that osteoclasts contribute to the pathoJOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014

genesis of CHS through physical creation of the niche, through cell-cell or paracrine communication, or through a combination. We did not address this mechanism in the current study. Several studies have defined the expression of cytokines and hormones in chondrosarcoma that orchestrate maturation and activation of osteoclasts, including interleukins (ILs), transforming growth factor-beta (TGF-b2), parthyroid hormone related protein (PTH-rP), tumor necrosis factor (TNF), and heparanase.26 In a previous study, we performed Serial Analysis of Gene Expression (SAGE) to examine the impact the tumor microenvironment has on global gene expression of the SRC tumor grown subcutaneously and within the tibial medullary cavity.11 Reexamining this data, TGFs, TNF, ILs, and heparanase-2 are clearly identified in the SRC tumors. Interestingly 18 SAGE tags (normalization to 100,000 tags per library) of TGF-b-2 were identified in tibia medullary cavity tumors, while no SAGE tags were identified in subcutaneous tumors, emphasizing the importance the microenvironment plays on gene ex-


1569

Figure 5. Delayed treatment with zoledronic acid prevents bone destruction by SRC and inhibits tumor growth in bone. Rats underwent surgical implantation of SRC into the proximal right tibia. Starting four days after implantation, control rats were administered IV PBS vehicle injections weekly, while experimental rats underwent ZA injection. (A) Plain radiographs of SRC-bearing tibias treated with vehicle (left panel) and ZA (right panel). The tumor invades bone surrounding the corticotomy site leading to metaphyseal cortical remodeling and destruction of cancellous bone. Invasion and destruction of bone is prevented by ZA treatment. (B) H&E stained sections at 4 magnification of SRC in the absence (left panel) and presence of ZA (right panel). The tumor (T) invades beyond bony confines, destroying cortical bone (b) and expanding beyond the bone compartment. With ZA administration, the SRC tumor (T) is confined to the corticotomy site. Middle two panels are images of the same sections taken at 20 magnification. Scale bars represent relative magnification. Lower two panels are corresponding Safranin O-stained sections to highlight tumor. Note the inability of tumor to penetrate bone in the presence of ZA. (C) Tumor size was ascertained by measuring cross-sectional area histomorphometrically in mid-sagittal histological sections. Treatment of rats with ZA 4 days after implantation of tumor resulted in significantly reduced tumor size in bone (sq mm, square millimeters, p ¼ 0.02). (D) PQ CT analysis was used to compare bone mineral density proximal to the corticotomy site of the operative (R) tibia and the corresponding site in the non-operative control (L) tibia in ZA and PBS treated rats. The average bone mineral density of (R) and (L) tibias was divided by that of (L) tibias to give percent bone remaining. There was a 29% reduction in bone mineral density with PBS treatment that did not reach statistical significance (n.s.). ZA prevented bone loss.

pression and its potential ability to activate the osteoclasts in the medullary cavity. PTH-rP is a potent agent in activation of osteoclasts; however we did not observe SAGE tags in our gene profile analysis, which appears to have resulted from the lack of the rat PTHrP sequence in the database we used for analysis. However, PTH-rP is clearly present in the SRC tumor as seen in a separate study whereby we showed strong immunohistochemical staining for the PTH-rP in SRC tumor when grown in the rat tibial medullary cavity.27 Additionally, rtPCR identified mRNA transcripts of PTH-rP in the SRC cells.27 In total we propose that SRC tumor mechanistically orchestrates activation of osteoclast through the presence of previously defined

cytokines and hormones shown to influence osteoclast activation. There are several limitations of our study. First, with use of an animal model, there is potential lack of applicability to human disease. We utilized the current model because of the similarities between SRC and human chondrosarcoma, including the ability of the tumor to grow within and destroy bone, spread locally, and colonize the lungs. Also, the high rate of growth of tumor allows for rapid assessment of treatment effects. Another limitation in the study is that we did not assess the affect of ZA on the tumor directly. In this regard, a recent study demonstrated that ZA treatment inhibited the growth of subcutaneous SRC, a JOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014

1570

OTERO ET AL.

Figure 6. Osteopetrotic rats are resistant to SRC-mediated bone destruction. Toothless rats (tl) are osteopetrotic, owing to a genetic defect in Csf1 resulting in absence of osteoclasts. Breeding of heterozygous mates results in three normal littermates for every tl osteopetrotic mutant. Phenotypically normal littermates possess at least one wild-type allele (þ/?). (A) Mid-sagittal CT reconstructions of sham and SRC-implanted tibias from normal littermates (þ/?) and toothless (tl) rats. There is loss of trabecular bone in the metaphysis (denoted with golden box) around SRC implantation site compared with sham surgical tibia. Note the micro-invasions into surrounding cortex (arrows) around the tumor implantation site. The tl tibias have increased bone density compared with normal littermate limbs, and there is no apparent reduction in bone mass in the presence of SRC. Note the thick sclerotic rim surrounding the tumor implantation site in SRC-implanted tl tibias (dashed enclosure). (B) Metaphyseal bone content was ascertained by measuring bone area/total area (B. Ar./T. Ar.) in 5 mm2 sections of metaphyseal bone on mid-sagittal micro CT images of tl and normal (þ/?) tibias implanted with SRC or which underwent sham surgery. Mean value þ standard deviation is shown. In the presence of SRC, there was a significant reduction in metaphyseal bone content in normal rats (þ/?), p ¼ 0.015. Osteopetrotic rats (tl) were resistant to tumorinduced bone loss. (C) Upper panels: H&E stained histological section of tibias at 4 magnification from control (þ/?) and toothless (tl) rats with implanted SRC. Left panel shows tumor (t), which has advanced well proximal to the implantation site replacing normal marrow elements and destroying metaphyseal trabeculae. Right panel shows tumor (t) unable to advance beyond corticotomy borders in osteopetrotic bone. Scale bars indicate identical magnification. Lower panels show 20 magnification of same specimens stained with safranin O. Note the inability of tumor to penetrate bone (b) edges in osteopetrotic rats. The bright pink bars (Arrow) in the toothless rat bone are retained primary spongiosa, characteristic of osteopetrotic bone. Scale bars show relative magnification.

scenario in which osteoclasts are not present to affect growth.28 Additionally, in tissue culture, ZA inhibited cell cycle progression and induced apoptosis in SRC.13 These results suggest that impairment of tumor growth and bone destruction by ZA treatment in our tibial SRC model may not be entirely explained by osteoclast-dependent mechanisms. Nevertheless, our results in the osteopetrotic rat model show that SRC tumor growth is affected by osteoclast-derived factors in vivo. Therefore, ZA may be beneficial clinically through both direct and indirect mechanisms of tumor inhibition. ZA has an established role as an adjunct therapy in the oncologic setting. It is administered in combination JOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014

with standard chemotherapy to patients with metastatic breast carcinoma, and has been shown to reduce bone-related events and to improve overall quality of life in this patient population.17,18 We used ZA to study the biology of CHS, a primary bone sarcoma, but osteoclast-targeted drugs may have clinical utility in CHS patients as well. Bisphosphonates or other osteoclast-targeted adjunctive therapy could potentially reduce the incidence of clinically evident local recurrence after surgery or decrease the aggressiveness of residual CHS after resection with positive, contaminated, or planned-positive margins. Further research is necessary to delineate the molecular and cellular details of CHS-osteoclast interaction in the tumor


microenvironment, and the current study forms a rational basis for the development of osteoclast-targeted therapies as adjunctive treatment for chondrosarcoma.

REFERENCES 1. Hameed M, Dorfman H. 2011. Primary malignant bone tumors–recent developments. Semin Diagn Pathol 28:86– 101. 2. Kim MJ, Cho KJ, Ayala AG, et al. 2011. Chondrosarcoma: with updates on molecular genetics. Sarcoma 2011:405437. 3. Moussavi-Harami F, Mollano A, Martin JA, et al. 2006. Intrinsic radiation resistance in human chondrosarcoma cells. Biochem Biophys Res Commun 346:379–385. 4. Donati D, El Ghoneimy A, Bertoni F, et al. 2005. Surgical treatment and outcome of conventional pelvic chondrosarcoma. J Bone Joint Surg Br 87:1527–1530. 5. Lee FY, Mankin HJ, Fondren G, et al. 1999. Chondrosarcoma of bone: an assessment of outcome. J Bone Joint Surg Am 81:326–338. 6. Pring ME, Weber KL, Unni KK, et al. 2001. Chondrosarcoma of the pelvis. A review of sixty-four cases. J Bone Joint Surg Am 83-A:1630–1642. 7. Sheth DS, Yasko AW, Johnson ME, et al. 1996. Chondrosarcoma of the pelvis. Prognostic factors for 67 patients treated with definitive surgery. Cancer 78:745–750. 8. Pritchard DJ, Lunke RJ, Taylor WF, et al. 1980. Chondrosarcoma: a clinicopathologic and statistical analysis. Cancer 45:149–157. 9. Wirbel RJ, Schulte M, Maier B, et al. 2000. Chondrosarcoma of the pelvis: oncologic and functional outcome. Sarcoma 4:161–168. 10. Teitelbaum SL. 2007. Osteoclasts: what do they do and how do they do it? Am J Pathol 170:427–435. 11. Hamm CA, Stevens JW, Xie H, et al. 2010. Microenvironment alters epigenetic and gene expression profiles in Swarm rat chondrosarcoma tumors. BMC Cancer 10:471. 12. Breitkreutz D, Diaz de Leon L, Paglia L, et al. 1979. Histological and biochemical studies of a transplantable rat chondrosarcoma. Cancer Res 39:5093–5100. 13. Gouin F, Ory B, Redini F, et al. 2006. Zoledronic acid slows down rat primary chondrosarcoma development, recurrent tumor progression after intralesional curretage and increases overall survival. Int J Cancer 119:980–984. 14. Van Wesenbeeck L, Odgren PR, MacKay CA, et al. 2002. The osteopetrotic mutation toothless (tl) is a loss-of-function frameshift mutation in the rat Csf1 gene: evidence of a crucial role for CSF-1 in osteoclastogenesis and endochondral ossification. Proc Natl Acad Sci U S A 99:14303–14308.

1571

15. Morcuende JA, Stevens JW, Scheetz TE, et al. 2012. Swarm rat chondrosarcoma cells as an in vivo model: lung colonization and effects of tissue environment on tumor growth. Iowa Orthop J 32:46–53. 16. Kenan S, Steiner GC. 1991. Experimental transplantation of the Swarm rat chondrosarcoma into bone: radiological and pathological studies. J Orthop Res 9:445–451. 17. Gnant M, Mlineritsch B, Schippinger W, et al. 2009. Endocrine therapy plus zoledronic acid in premenopausal breast cancer. N Engl J Med 360:679–691. 18. Wardley A, Davidson N, Barrett-Lee P, et al. 2005. Zoledronic acid significantly improves pain scores and quality of life in breast cancer patients with bone metastases: a randomised, crossover study of community vs hospital bisphosphonate administration. Br J Cancer 92:1869–1876. 19. Grimaud E, Damiens C, Rousselle AV, et al. 2002. Bone remodelling and tumour grade modifications induced by interactions between bone and swarm rat chondrosarcoma. Histol Histopathol 17:1103–1111. 20. Hascall GK, Kimura JH. 1981. The ultrastructure of cultures from the Swarm rat chondrosarcoma. Anat Rec 200:287–292. 21. Kimura JH, Hardingham TE, Hascall VC, et al. 1979. Biosynthesis of proteoglycans and their assembly into aggregates in cultures of chondrocytes from the Swarm rat chondrosarcoma. J Biol Chem 254:2600–2609. 22. Maibenco HC, Krehbiel RH, Nelson D. 1967. Transplantable osteogenic tumor in the rat. Cancer Res 27:362–366. 23. Murphey MD, Walker EA, Wilson AJ, et al. 2003. From the archives of the AFIP: imaging of primary chondrosarcoma: radiologic-pathologic correlation. Radiographics 23:1245– 1278. 24. David E, Blanchard F, Heymann MF, et al. 2011. The bone niche of chondrosarcoma: a sanctuary for drug resistance, tumour growth and also a source of new therapeutic targets. Sarcoma 2011: 932451. 1–8. 25. Ebetino FH, Hogan AM, Sun S, et al. 2011. The relationship between the chemistry and biological activity of the bisphosphonates. Bone 49: 20–33. 26. Suva LJ, Washam C, Nicholas RW, et al. 2011. Bone metastasis: mechanisms and therapeutic opportunities. Nat Rev Endocrinol 7:208–218. 27. Oji GS, Gomez P, Kurriger G, et al. 2007. Indian hedgehog signaling pathway differences between swarm rat chondrosarcoma and native rat chondrocytes. Iowa Orthop J 27:9– 16. 28. Miot-Noirault E, David E, Vidal A, et al. 2013. 99mTc-NT P 15-5 assessment of the early therapeutic response of chondrosarcoma to zoledronic acid in the Swarm rat orthotopic model. EJNMMI Res 3:40.
