Vol. 119 No. 1 January 2015

RANK, RANKL, and OPG in recurrent solid/multicystic ameloblastoma: their distribution patterns and biologic significance Chong Huat Siar, BDS, MSc, FDSRCPS, FRCPath,a Hidetsugu Tsujigiwa, DDS, PhD,b Ismadi Ishak, BDS, FDSRCS, MBBS,c Nurmawarnis Mat Hussin, BDS,d Hitoshi Nagatsuka, DDS, PhD,e and Kok Han Ng, BDS, MSc, FDSRCPS, FRCPathf Objectives. To determine the distribution patterns of bone resorption regulators, receptor activator of nuclear factor k-B (RANK), RANK ligand (RANKL), and osteoprotegerin (OPG) in recurrent ameloblastoma (RAs) and to clarify their impact on the biologic behavior of these neoplasms. Materials and Methods. Fifteen paraffin-embedded RA cases were subjected to immunohistochemistry for expression of RANK, RANKL, and OPG. Results. The RANK-RANKL-OPG triad was heterogeneously detected in RA samples. RANK, essential for osteoclast differentiation, was strongly expressed in tumoral epithelium. Conversely, RANKL, an osteoclast activator, was markedly underexpressed, and protein localization was predominantly stromal. OPG, an osteoclastogenesis inhibitory factor, was detected in neoplastic epithelium more than in stroma, suggesting functional inactivation of RANKL. Most RA (n ¼ 12/15; 80%) exhibited a bimolecular spatial expression pattern, the most common being RANK-positive/OPG-positive (n ¼ 8/15; 53.3%). All three proteins showed no significant correlation with the clinical/histopathologic parameters in RA patients (P > .05). Conclusions. The RANKþ/RANKLlow//OPGþ phenotype observed in RA suggests an altered local bone metabolism characterized by low bone resorptive activity in these recurrent tumors. (Oral Surg Oral Med Oral Pathol Oral Radiol 2015;119:83-91)

Recent advances in bone biology saw the identification of a novel set of cytokines/cytokine receptors that are key regulators for osteoclast formation, differentiation, and activity.1,2 These regulatory molecules, which belong to the tumor necrosis factor ligand and receptor superfamilies, are widely expressed in human tissues and by their distributions would account, in part, for their multifunctional roles, including involvement in bone metabolism and in immune and vascular a Professor of Oral Pathology and Oral Medicine, Department of OroMaxillofacial Surgical and Medical Sciences, Faculty of Dentistry, University of Malaya, Kuala Lumpur, Malaysia. b Professor of Histopathology, Laboratory of Histopathology, Department of Life Science, Faculty of Science, Okayama University of Science, Okayama, Japan. c Senior Lecturer and Consultant Oral and Maxillofacial Surgeon, Department of Oro-Maxillofacial Surgical and Medical Sciences, Faculty of Dentistry, University of Malaya, Kuala Lumpur, Malaysia. d Master of Clinical Dentistry (Oral Medicine and Oral Pathology) Resident, Department of Oro-Maxillofacial Surgical and Medical Sciences, Faculty of Dentistry, University of Malaya, Kuala Lumpur, Malaysia. e Professor of Oral Pathology and Medicine, Department of Oral Pathology and Medicine, Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama University, Okayama, Japan. f Former Director and Senior Consultant Oral Pathologist, Unit of Stomatology, Institute for Medical Research, Jalan Pahang, Kuala Lumpur, Malaysia. Received for publication Jul 1, 2014; returned for revision Aug 19, 2014; accepted for publication Sep 15, 2014. Ó 2015 Elsevier Inc. All rights reserved. 2212-4403/$ - see front matter http://dx.doi.org/10.1016/j.oooo.2014.09.017

systems.3-6 RANKL (receptor activator of nuclear factor k-B ligand) is present as membrane-bound or soluble forms in a variety of cell types, including osteoblasts, dental cells, and other cells of the microenvironment. RANK (receptor activator of nuclear factor k-B) is located on the surface of cells of myeloid lineage (osteoclasts, monocytic, and dendritic cells), whereas OPG (osteoprotegerin) is a soluble glycoprotein expressed by bone (osteoblasts, mesenchymal stem cells), immune cells (T and B cells) and vessels (endothelial and vascular smooth muscle cells). Differentiation and activation of pro-osteoclasts (osteoclast precursor cells) to mature (active) osteoclasts requires binding of RANKL to RANK. The RANKLeRANK interaction is inhibited by OPG, which acts as a soluble decoy receptor for RANKL and thereby hinders osteoclast activation.7 The RANK/RANKL/OPG system, with its functional equilibrium, is crucial for the homeostasis of bone remodeling, including osteoimmunology, and its dysregulation has been implicated in a range of neoplastic and nonneoplastic diseases.3,8-12

Statement of Clinical Relevance Clarification on the distribution patterns of RANK, RANKL, and OPG would contribute to a better understanding regarding the growth characteristics of recurrent ameloblastoma. 83

ORAL AND MAXILLOFACIAL PATHOLOGY 84 Siar et al.

The potential role of the RANK/RANKL/OPG axis in inducing osteolysis in odontogenic cysts and tumors has gained considerable attention recently.13-25 This molecular triad has been investigated in different subtypes of ameloblastoma and implicated in its local invasiveness.15-23 From these studies, it has been reported that (1) the odontogenic epithelium in ameloblastoma has a similar expression for RANK, RANKL, and OPG17; (2) the molecular system of RANK/RANKL/OPG is variably expressed in ameloblastoma19; (3) no differences were observed in cell reactivity for RANK, RANKL, and OPG in central and peripheral cells in ameloblastoma17; (4) RANKL and OPG reactivity in tumor cells was weaker than in stromal cells13; (5) the majority of solid/ multicystic and unicystic ameloblastomas has higher numbers of RANKL-positive than OPG-positive cells in the stroma17,20; and (6) RANKL immunoreactivity is localized to cells scattered within the connective tissue stroma of ameloblastoma.14 The main conclusions drawn were that (1) imbalance in RANK, RANKL, and OPG expression might locally regulate bone metabolism and contribute to the differential bone resorption activity and consequently mediate progression of ameloblastoma13-20; and (2) these factors might also be involved in tumor cell differentiation and/or tumor tissue restructuring of ameloblastoma.13 To date, it is unclear whether RANK, RANKL, and OPG differential expression patterns encountered in primary ameloblastomas are reflected in the recurrent tumor as well. It is known that morphologic, signaling, and genetic differences may exist between primary and recurrent tumors.26 Therefore in this study, the two objectives were (1) to delineate and compare the distribution patterns of RANK, RANKL, and OPG in recurrent ameloblastoma (RA) and (2) to correlate these results with the clinical and histopathologic parameters in patients presenting with these tumors. Our working hypothesis was that the RANK/RANKL/OPG signaling pathway might be dysregulated in RA and that this aberrant event would be paralleled by an imbalance in the immunophenotypic expression and distribution of these protein factors in the tumor epithelium and stroma of RA. To test this hypothesis, a study based on an immunohistochemical analysis for RANK, RANKL, and OPG in a sample of 15 solid/ multicystic RA was carried out. It is hoped that the results of this study would contribute to a better understanding of the role of these factors in the biologic behavior of these recurrent neoplasms.

MATERIALS AND METHODS Tissue sample This study was exempted from institutional review board approval. Archival paraffin-embedded tissue blocks of 15 cases of RA were selected from the Oral

OOOO January 2015

Table I. Patients’ characteristics Patients’ details No.

Age/Gender/Ethnic

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

10/M/Malay 16/F/Chinese 21/F/Chinese 23/F/Chinese 25/F/Indonesian 27/M/Malay 27/M/Malay 29/M/Malay 31/M/Malay 34/M/Malay 35/F/Chinese 37/M/Malay 60/F/Chinese 61/M/Indian 70/M/Malay

Tumor details Site

#27 e R condyle L mand #21 e 32 R angle mand #20 e L angle mand #24 e 25 L mand Submental R mand R body mand R submand Mand NOS Mand NOS R max L mand

RA subtypes SMA SMA SMA SMA SMA SMA SMA SMA SMA SMA SMA SMA SMA SMA SMA

M, male; F, female; R, right; L, left; Mand, mandible; max, maxilla; submand, submandibular; NOS, not otherwise specified; #20, left mandibular second premolar; #21, left mandibular first premolar; #24, left mandibular central incisor; #25, right mandibular central incisor; #27, right mandibular canine; #32, right mandibular third molar (ADA Universal Tooth Numbering System); RA, recurrent ameloblastoma; SMA, solid/multicystic ameloblastoma.

Pathology Diagnostic and Research Laboratory, Faculty of Dentistry, University of Malaya, Kuala Lumpur, Malaysia. These cases were reviewed and selected according to the World Health Organization Classification of Tumors.27 Case selection inclusion criteria included surgical specimens with adequate intratumoral tissues and invasive front of tumor tissues. Patient details, including onset age, gender, anatomic site, and case summaries, were recorded. Immunohistochemistry New 5-mm thick sections mounted on salinized slides were used. Immunohistochemical staining was done as previously described.28 The primary antibodies used were mouse monoclonal RANK (AB13918) (Abcam Inc. Cambridge, MA; diluted 1:200), mouse monoclonal RANKL (AB45039) (Abcam, 1:400), and rabbit polyclonal osteoprotegerin (OPG) (AB9986) (Abcam, 1:100). A case of central giant cell granuloma was used as positive control. In addition, positive labeling in the stromal cells (macrophages, endothelium, and lymphocytes) and bone tissue cells (osteoblasts and osteoclasts) located at the tumor periphery served as internal positive controls. For negative control, sections were treated as above but without the primary antibody. All the control sections were negative. Immunohistochemical analysis The spatial expression patterns and immunoreactivity for all markers were evaluated by descriptive and

OOOO Volume 119, Number 1

ORIGINAL ARTICLE Siar et al. 85

Fig. 1. Representative sections of a RANKþ/RANKLþ/OPG recurrent ameloblastoma showing co-localization of RANK and RANKL in the cytoplasm and membrane of peripheral columnar cells (B and D). Note stromal and cystic accumulation of RANKL (C). Original magnification: A, C, and E, 40; B, D, and F, 400.

semiquantitative methods. Digitized images (Olyvia DotSlide Virtual Slide System, Olympus Imaging Inc., Tokyo, Japan) of all slides in each case were assessed by two investigators (CHS and KHN). Protein staining was assessed according to the methods described by Mello et al.29 and Bello et al.30 with minor modifications. Histomorphometry was performed at 200 magnification using the Image-Pro Express software (Media Cybernetics Inc., Bethesda, MD) for Windows. Each case was rated according to a score based on the product of percentage of immunoreactive tumor epithelial cells present: () negative when none of these cells were positively stained; (þ) mild when staining was present in focal areas (50%). Stromal cellular components were similarly quantified.

Statistical analysis Spearman correlation coefficient was carried out to assess correlations among RANK, RANKL, or OPG and the patients’ clinical and histopathologic parameters. Comparative analysis of expression levels for each marker in tumor epithelium versus stroma was performed by the Wilcoxon Signed Ranks test. P value of less than .05 was considered significant. All statistical analyses were performed using SPSS software version 12.0 (SPSS Inc., Chicago, IL).

RESULTS Patients’ characteristics The median age of 15 patients with RA was 29 years (mean, 33.7  17.1 years; range, 10-70 years) (Table I). There were 9 male and 6 female patients, with a male-tofemale ratio of 1.5:1. All patients presented with a known history of primary ameloblastoma removal. The duration between primary treatment and recurrence

ORAL AND MAXILLOFACIAL PATHOLOGY 86 Siar et al.

OOOO January 2015

Fig. 2. Representative sections of a RANKþ/RANKLþ/OPG recurrent ameloblastoma showing lack of co-localization for RANK and RANKL. Instead, RANK is localized in the cytoplasm and membrane of peripheral columnar epithelial cells (B), and RANKL is in central polyhedral cells (D). Note stromal positivity for RANKL (C). Original magnification: A, C, and E, 40; B, D, and F, 200.

ranged from 1 to 29 years (mean, 8.3 years). Clinically, the most common presenting complaint was swelling (n ¼ 10/15; 66.7%), followed in each case by mandibular pathologic fracture (n ¼ 1/15; 6.7%) and pus discharge (n ¼ 1/15; 6.7%). Nearly all cases (n ¼ 14/15; 93.3%) were preoperatively diagnosed as RA. Six (40%) cases were treated with excision, 3 cases each with enucleation (20%) and curettage (20%), 2 (13.3%) with resections, and 1 (6.7%) with hemimandibulectomy. In all cases, no follow-up records were available to determine the outcome after treatment. Immunohistochemical findings The distribution patterns of RANK, RANKL, and OPG in RA tumors are illustrated in Figures 1 to 4, and the results of their staining levels according to the various clinical and histopathologic parameters are detailed in Table II. RANK. RANK was clearly identified in almost all RA tumors (n ¼ 14/15; 93.3%). In RANK-positive tumors, a heterogeneous distribution pattern was observed (Figure 1, A and B and Figure 2, A and B).

Their protein localization was predominantly cytoplasmic and sometimes membranous (Figure 2, A and B). RANK was more frequently detected in peripheral cells with a columnar morphology (Figure 2, A and B), whereas it was absent or mild-to-moderate in central polyhedral cells (Figure 1, B; Figure 2, B; and Figure 3, C). Stromal and bone cells were variably positive for RANK (not shown). Expression levels for RANK between epithelial and stromal cells were significantly different (Table III) (P < .05). RANKL. RANKL was markedly underexpressed in RA tumors (n ¼ 3/15; 20%). In RANKL-expressing tumors, the protein was more frequently detected in the stroma than within the tumor epithelium (Figure 1, C and D; and Figure 2, C and D). In the latter, protein localization was mainly cytoplasmic and sometimes membranous. Stromal and bone cells variably expressed RANKL (Figure 4, C). Expression levels for RANKL between epithelial and stromal cells were not significantly different (Table III) (P > .05). OPG. OPG was widely expressed in RA (n ¼ 11/15; 73.3%). In OPG-positive tumors, a heterogeneous

OOOO Volume 119, Number 1

ORIGINAL ARTICLE Siar et al. 87

Fig. 3. Representative sections of a RANKþ/RANKL/OPGþ recurrent ameloblastoma showing mild expression for RANK (B, C) and strong expression for OPG in both the peripheral and central epithelial cells (H, I). Note stromal positivity for OPG. (Original magnification: A, D, and G, 40; B, E, and H, 100; C, F, and I; magnification 200.)

distribution pattern was observed. OPG protein localization was more frequently cytoplasmic than membranous. It was variably distributed in both the peripheral and central polyhedral epithelial cells (Figure 3, G and H; and Figure 4, E and F). Expression levels for OPG between epithelial and stromal cells were significantly different (Table III) (P < .05). RANK/RANKL/OPG. Most RA tumors exhibited a bimolecular spatial expression pattern where RANKþ/ RANKLe/OPGþ profile (n ¼ 8/15, 53.3%) was most prevalent, followed by RANKþ/RANKLþ/OPGe (n ¼ 2/15, 13.3%). Only a small number of tumors demonstrated positivity for all three proteins (n ¼ 2/15, 13.3%). The remaining RA tumors expressed either RANK (n ¼ 2/15, 13.3%) or OPG (n ¼ 1/15, 6.7%). Statistical analysis Spearman correlation coefficient did not identify any significant correlation among RANK, RANKL, and OPG and the patients’ clinical and histopathologic

parameters (P > .05) (Table II). Wilcoxon Signed Ranks test detected significant differences in expression levels between tumor epithelium and stroma for RANK and OPG (P > .05) (Table III).

DISCUSSION The most significant observation in this study was the distinct RANKL-low/negative expression pattern demonstrated by most tumors in our RA sample (n ¼ 12/15; 80%). In contrast, earlier reports on ameloblastoma described wide expression of RANKL in the tumor parenchyma and/or stromal cells.13,14,16-20 The reasons for this discrepancy in the findings are unknown. The ameloblastoma is well recognized as a locally aggressive odontogenic epithelial neoplasm that can cause extensive bone destruction. According to the current and aforementioned studies,13-20 local bone remodeling activities seem to take their molecular cues from the RANK/RANKL/OPG pathway usurped by ameloblastoma (Figure 5, A). Therefore, an elevated

ORAL AND MAXILLOFACIAL PATHOLOGY 88 Siar et al.

OOOO January 2015

Fig. 4. Representative sections of a RANKe/RANKLþ/OPGþ recurrent ameloblastoma showing differential expression of RANKL and OPG. Both RANKL and OPG are mildly expressed in central polyhedral cells (D, F). Note stromal positivity for RANKL (C). (Original magnification: A, C, and E, 40; B, D, and F, 200.)

RANKL reactivity, as noted earlier,13,14,16-20 is not an unexpected finding in an osteolytic lesion exemplified by this neoplasm. Our current data, on the other hand, registered marked loss of expression for RANKL in most RA tumors, and in those few RANKL-positive cases, stromal positivity clearly dominated over tumor epithelial reactivity.14,17 It is unclear whether this RANKL-low/negative phenotype is unique to recurrent tumors of ameloblastoma. There is no known previous study examining this cytokine in RA for our baseline reference or comparison. Moreover, we were unable to substantiate or verify this view as well because the primary and recurrent tumors from the same patient were not available for evaluation. Nevertheless, a plausible explanation for the disparity in RANKL findings might be that the RA group of tumors investigated here most likely represents an indolent/dormant subset of tumors distinct from their primary counterparts. The low/non-expression of RANKL in these tumors might imply an altered local bone metabolism with low resorptive activities

(Figure 5, B). The other alternative explanation is that RANKL upregulation or downregulation in ameloblastoma might be reflective of the dynamic ongoing process associated with tumor-induced bone remodeling. Differing results with regard to RANKL expression were likewise reported in breast carcinoma.8,31,32 In those rare cases of breast carcinoma that were RANKL-positive, the protein also resides mainly in the stromal compartment.8 Other minor concerns that might contribute to these disparities in findings include differences in the sources of antibodies used, methodologies for tissue processing, and histomorphometry. In the present study, analysis of RANK/RANKL/OPG spatial expression patterns revealed that most RA tumors expressed a maximum of two protein molecules. The most prevalent bimolecular protein profile observed here was RANK-positive or OPG-positive. This wide expression for RANK16,17 and OPG17 correlated well with previous reports. Of significance here is the fact that since most tumors were RANKL-low/negative, RANK

OOOO Volume 119, Number 1

ORIGINAL ARTICLE Siar et al. 89

Table II. Correlation among RANK, RANKL, and OPG and the clinical and histopathologic parameters in 15 patients with recurrent ameloblastoma RANK Factors

RANKL

N ¼ 15



þ

þþ

þþþ

8 7

1 0

1 3

6 3

0 1

9 6

1 0

2 2

5 4

1 0

8 5 1 1

1 0 0 0

2 1 0 1

4 4 1 0

1 0 0 0

1 12 2

0 1 0

0 3 1

1 7 1

0 1 0

11 3 1

1 0 0

3 0 1

6 3 0

1 0 0

P value

*

Age (years) 29 >29 Gender Male Female Ethnicity Malay Chinese Indian Indonesian Location Maxilla Mandible Others Histologic subtypes Plexiform (P) Follicular (F) PþF

OPG



þ

þþ

þþþ

7 5

0 1

1 0

0 1

8 4

0 1

0 1

1 0

7 4 1 0

0 1 0 0

0 0 0 1

1 0 0 0

1 10 1

0 1 0

0 1 0

0 0 1

8 3 1

1 0 0

1 0 0

1 0 0

.950

P value



þ

þþ

þþþ

3 1

5 3

0 1

0 2

1 3

6 2

0 1

2 0

1 3 0 0

5 1 1 1

0 1 0 0

2 0 0 0

0 4 0

1 5 2

0 1 0

0 2 0

4 0 0

5 3 0

1 0 0

1 0 1

P value

.477

.899

.073

.420

.930

.124

.391

.885

.336

.368

.978

.823

.283

.202

* Median age 29 years in RA sample was used as cutoff point for binary measurements of patients’ ages; P value was determined by Spearman correlation coefficient test. The level of significance was set at 0.05.

Table III. Staining intensity levels for RANK, RANKL, and OPG in tumor epithelial and stromal cells of recurrent ameloblastoma (n ¼ 15) None

Mild

Moderate Strong

None

Mild

Moderate Strong

P value*

RANK

1

5

8

1

RANK

2

12

1

0

.008

RANKL

12

1

1

1

RANKL

12

1

1

1

.317

OPG

4

9

1

1

OPG

13

2

0

0

.021

Number of cases distributed by staining intensity

Number of cases distributed by staining intensity

(Epithelial cells)

(Stromal cells)

*Differences were determined by Wilcoxon signed ranks test. Boxed areas (red) indicate statistical significances (P < .05).

and OPG overexpression appear to be important factors determining osteoclastogenesis in RA. The heterogeneous distribution pattern of RANK, RANKL, and OPG in the tumors expressing these proteins could be extrapolated to the slow-growing nature of ameloblastoma. Tumor locations demonstrating positive identification of these molecules most likely represented active bone remodeling sites, whereas the downregulation of these factors would

represent quiescent areas. Furthermore, there is a likelihood that other mechanism(s) besides the RANK/ RANKL/OPG signaling pathway may be involved in the bone remodeling process. The main limitation of this study was the nonavailability of the primary and recurrent tumors from the same patient to compare their expression levels and distribution patterns for RANK, RANKL, and OPG. The other setback was the lack of information on the

ORAL AND MAXILLOFACIAL PATHOLOGY 90 Siar et al.

OOOO January 2015

Fig. 5. Schematic diagram illustrating RANK/RANKL/OPG pathway interactions and local bone remodeling activities in the presence of ameloblastoma. A, Ligation of RANKL to RANK, its cognate receptor, results in preosteoclast recruitment, fusion, differentiation, activation, and survival of osteoclasts. B, OPG, a decoy receptor for RANKL, inhibits interaction between RANKL and RANK and results in functional inactivation of osteoclasts.

clinical outcome of RA cases for correlative analysis against the expressions of these molecules. Furthermore, all RA tumors evaluated here were solid or multicystic ameloblastomas. This means that recurrent unicystic and other recurrent subsets of this neoplasm have yet to be investigated. Given these shortcomings, future studies aimed at clarifying these issues are needed. In conclusion, ameloblastoma still poses questions regarding the mechanism(s) underlying its ability to induce bone resorption and grow aggressively within the jaws. In this study, we presented immunohistochemical evidence that the RANK/RANKL/OPG pathway might be dysregulated in RA. Our results of a RANKL-low/negative, RANK-positive and OPG-positive distribution pattern in RA tumors suggest an imbalance in local bone metabolism and a shift away from osteoclastogenesis. These findings may thus shed new light on the differential roles of these molecules in tumor-induced bone remodeling activities in RA, and may also provide directions for clinical treatment of these recurrent tumors. This study was jointly supported by the Ministry of Higher Education Malaysia Fundamental Research Grant Scheme (FP038-2013 A), Japan Society for Promotion of Science

(JSPS) KAKENHI Grant-in-Aid for Scientific Research (C26462783) and Grant-in-Aid for Research Activity Startup (25893141). REFERENCES 1. Abu-Amer Y. Advances in osteoclast differentiation and function. Curr Drug Targets Immune Endocr Metab Disord. 2005;5:347-355. 2. Roodman GD. Regulation of osteoclast differentiation. Ann NY Acad Sci. 2006;1068:100-109. 3. Khosla S. Mini-review: the OPG/RANKL/RANK system. Endocrinology. 2001;142:5050-5055. 4. Proff P, Römer P. The molecular mechanism behind bone remodelling: a review. Clin Oral Investig. 2009;13:355-362. 5. Rani CS, MacDougall M. Dental cells express factors that regulate bone resorption. Mol Cell Biol Res Commun. 2000;3:145-152. 6. Ohazama A, Courtney JM, Sharpe PT. OPG, RANK, and RANKL in tooth development: co-ordination of odontogenesis and osteogenesis. J Dent Res. 2004;83:241-244. 7. Shipman CM, Croucher PI. Osteoprotegerin is a soluble decoy receptor for tumor necrosis factor-related apoptosis-inducing ligand/Apo2 ligand and can function as a paracrine survival factor for human myeloma cells. Cancer Res. 2003;63:912-916. 8. Ibrahim T, Sacanna E, Gaudio M, et al. Role of RANK, RANKL OPG and CXCR4 tissue markers in predicting bone metastases in breast cancer patients. Clin Breast Cancer. 2011;11:369-375. 9. Ito R, Nakayama H, Yoshida K, et al. Expression of osteoprotegerin correlates with aggressiveness and poor prognosis of gastric carcinoma. Virchows Arch. 2003;443:146-151.

OOOO Volume 119, Number 1 10. Shang M, Lin L, Cui H. Association of genetic polymorphisms of RANK, RANKL and OPG with bone density in Chinese peri- and postmenopausal women. Clin Biochem. 2013;46:1493-1501. 11. Samara S, Dailiana Z, Chassanidis C, et al. Expression profile of osteoprotegerin, RANK and RANKL genes in the femoral head of patients with avascular necrosis. Exp Mol Pathol. 2014;96:9-14. 12. Cao H, Li Q, Li M, et al. Osteoprotegerin/RANK/RANKL axis and atrial remodelling in mitral valvular patients with atrial fibrillation. Int J Cardiol. 2013;166:702-708. 13. Kumamoto H, Ooya K. Expression of parathyroid hormone related protein (PTHrP), osteoclast differentiation factor (ODF)/receptor activator of nuclear factor-kappa B ligand (RANKL) and osteoclastogenesis inhibitory factor (OCIF)/ osteoprotegerin (OPG) in ameloblastomas. J Oral Pathol Med. 2004;33:46-52. 14. Tay JYY, Bay BH, Yeo JF, Harris M, Meghi S, Dheen ST. Identification of RANKL in osteolytic lesions of the facial skeleton. J Dent Res. 2004;83:349-353. 15. Sandra F, Hendarmin L, Kukita T, Nakao Y, Norifumi N, Nakamura S. Ameloblastoma induces osteoclastogenesis: a possible role of ameloblastoma in expanding in the bone. Oral Oncol. 2005;41:637-644. 16. Sandra F, Hendarmin L, Nakamura S. Osteoprotegerin (OPG) binds with tumor necrosis factor-related apoptosis-inducing ligand (TRAIL): suppression of TRAIL-induced apoptosis in ameloblastomas. Oral Oncol. 2006;42:415-420. 17. da Silva TA, Batista AC, Mendonca EF, Leles CR, Fukada S, Cunha FQ. Comparative expression of RANK, RANKL, and OPG in keratocystic odontogenic tumors, ameloblastomas, and dentigerous cysts. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2008;105:333-341. 18. Qian Y, Huang HZ. The role of RANKL and MMP-9 in the bone resorption caused by ameloblastoma. J Oral Pathol Med. 2010;39:592-598. 19. Tekkesin MS, Mutlu S, Olgac V. The role of RANK/RANKL/ OPG signalling pathways in osteoclastogenesis in odontogenic keratocysts, radicular cysts, and ameloblastomas. Head Neck Pathol. 2011;5:248-253. 20. de Matos FR, de Moraes M, das Neves Silva EB, Galvao HC, de Almeida Freitas R. Immunohistochemical detection of receptor activator nuclear kB ligand and osteoprotegerin in odontogenic cysts and tumors. J Oral maxillofac Surg. 2013;71: 1886-1892. 21. Andrade FR, Sousa DP, Mendonca EF, Silva TA, Lara VS, Batista AC. Expression of bone resorption regulators (RANK, RANKL, and OPG) in odontogenic tumors. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2008;106:548-552.

ORIGINAL ARTICLE Siar et al. 91 22. Gong Y, Wang L, Wang H, Li T, Chen X. The expression of NFkB, Ki-67 and MMP-9 in CCOT, DGCT and GCOC. Oral Oncol. 2009;45:515-520. 23. de Moraes M, de Lucena HF, de Azevedo PR, Queiroz LM, Costa Ade L. Comparative immunohistochemical expression of RANK, RANKL and OPG in radicular and dentigerous cysts. Arch Oral Biol. 2011;56:1256-1263. 24. Menezes R, Bramante CM, Silva-Paiva KB, et al. Differential patterns of receptor activator of nuclear factor kappa B ligand/ osteoprotegerin expression in human periapical granulomas: possible association with progressive or stable nature of the lesions. J Endod. 2008;34:932-938. 25. Zakopoulos A, Tosios KI, Vlachodimitropoulos D. Expression of OPG, RANKL, and TRAIL in odontogenic keratocysts (OKC): an immunohistochemical study. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2007;103:e24. 26. Friedl P, Alexander S. Cancer invasion and the microenvironment: plasticity. Cell. 2011;147:992-998. 27. Barnes L, Eveson JW, Reichart P, Sidransky D, eds. World Health Organization Classification of Tumours. Pathology and Genetics of Head and Neck Tumours. Lyon, France: IARC Press; 2005. 28. Siar CH, Ng KH. Differential expression of transcription factors Snail, Slug, SIP1 and Twist in ameloblastoma. J Oral Pathol Med. 2014;43:45-52. 29. Mello LA, Figueiredo AL, Ramos EA, et al. CD1 a-positive Langerhans cells and their relationship with E-cadherin in ameloblastomas and keratocystic odontogenic tumors. J Oral Pathol Med. 2013;42:454-461. 30. Bello IO, Soini Y, Slootweg PJ, Salo T. Claudins 1, 4, 5, 7 and occludin in ameloblastomas and developing human teeth. J Oral Pathol Med. 2007;36:48-54. 31. Trinkaus M, Ooi WS, Amir E, et al. Examination of the mechanisms of osteolysis in patients with metastatic breast cancer. Oncol Rep. 2009;21:1153-1159. 32. Chen YC, Sosnoski DM, Mastro AM. Breast cancer metastasis to the bone: mechanisms of bone loss. Breast Cancer Res. 2010;12:215. Reprint requests: Siar CH Department of Oro-Maxillofacial Surgical and Medical Sciences Faculty of Dentistry University of Malaya 50603 Kuala Lumpur Malaysia [email protected]