Glucocorticoids, calcitonin, and osteocalcin cannot differentiate between aggressive and nonaggressive central giant cell lesions of the jaws

Vol. 120 No. 3 September 2015

Glucocorticoids, calcitonin, and osteocalcin cannot differentiate between aggressive and nonaggressive central giant cell lesions of the jaws Allisson Filipe Lopes Martins, DDS, MSc,a Paulo Otávio Carmo Souza, DDS,a Inara Carneiro Costa Rege, DDS, MSc,b Marília Oliveira Morais, DDS, MSc,a and Elismauro Francisco Mendonça, DDS, MSc, PhDa Objective. To evaluate the expression of glucocorticoid receptor (GR), calcitonin receptor (CTR), and osteocalcin (OC) in aggressive and nonaggressive central giant cell lesions (CGCLs). The numbers of mitotic and multinucleated giant cells were also evaluated. Study Design. Thirty-one cases of CGCL were submitted for immunohistochemistry. Mitotic figures and multinucleated giant cells were assessed through histochemical analyses. Results. Positive staining for GR, CTR, and OC was observed in all cases studied. There were no differences between CGCL variants with regard to the expression of GR, CTR, or OC. The aggressive group showed a higher number of multinucleated giant cells compared with the nonaggressive group (P < .05). Conclusions. Nonaggressive and aggressive CGCLs cannot be distinguished by OC, CTR, or GR expression, although the number of multinucleated giant cells may help differentiate between CGCL types. (Oral Surg Oral Med Oral Pathol Oral Radiol 2015;120:386-395)

The central giant cell lesion (CGCL) of the jaws is a benign intraosseous lesion. Whether CGCLs are consistent with neoplastic or reactive lesions is not well understood and needs to be clarified.1-4 On the basis of clinical and radiographic findings, CGCLs may be classified as either nonaggressive or aggressive. The aggressive lesions are associated with pain, rapid growth, cortical perforation, root resorption, and tendency to recur after surgical treatment, whereas the nonaggressive lesions cause little or no pain, grow slowly, and do not promote cortical perforation or root resorption.2,5-7 Aggressive lesions may also present with a higher number of multinucleated giant cells.7-9 Analysis of the number of mitotic bodies within mononuclear cells helps distinguish between aggressive and nonaggressive CGCLs;7,9 however, some studies have shown that the lesions cannot be distinguished by the number of proliferating cells.5,10 The classic treatment for CGCL may result in esthetic and functional defects, and thus, surgical procedures are undesirable in some cases, including in children and young adults.11,12 Alternative treatments have been suggested, such as the use of intralesional glucocorticoids, calcitonin, and interferon-alpha.6,13-17 The use of these agents was based on clinical observations and similarities between CGCLs and others lesions.2,3,16 It is believed that glucocorticoids inhibit bone resorption by inducing a

Oral Pathology, Universidade Federal de Goiás, Goiânia, Goiâs, Brazil. b Professor of Oral Radiology, Universidade Paulista (UNIP), Goiânia, Goiâs, Brazil. Received for publication Feb 27, 2015; returned for revision May 15, 2015; accepted for publication May 21, 2015. Ó 2015 Elsevier Inc. All rights reserved. 2212-4403/$ - see front matter http://dx.doi.org/10.1016/j.oooo.2015.05.016

386

osteoclast apoptosis, thus reducing protease production and suppressing inflammatory and angiogenic processes within CGCLs.6,12,17,18 Calcitonin may inhibit multinucleated giant cell motility and DNA synthesis.19,20 According to the literature, the results of the therapeutic treatment for CGCLs are variable and hard to predict.13-17,21-28 To justify the treatment of CGCLs with glucocorticoids and calcitonin, some studies have investigated glucocorticoid receptor (GR) and calcitonin receptor (CTR) expression within lesion cells. Tobón-Arroyave et al.29 were the first to investigate these receptors in aggressive and nonaggressive CGCLs and concluded that aggressive lesions have higher expression levels of CTR. Additionally, Vered et al.4 found that all CGCL cases expressed GR, but CTR was not expressed in all lesions. Finally, Nogueira et al.2 showed that there is no difference between the clinical variants of the lesion with regard to the expression of CTR and GR. Although GR and CTR expression has been studied in CGCL, there is no agreement as to whether the expression of these receptors differs among the CGCL variants.

Statement of Clinical Relevance There is no difference with regard to the expression of glucocorticoids and calcitonin receptors between nonaggressive and aggressive central giant cell lesions (CGCLs), suggesting that treatment with glucocorticoids and calcitonin may be indicated in both types of CGCL. Osteocalcin expression in mononuclear cells of the nonaggressive group may reflect a reduced osteolytic potential.

OOOO Volume 120, Number 3

ORIGINAL ARTICLE Martins et al. 387

Fig. 1. A, Panoramic radiographs showing a multilocular osteolytic lesion with well-defined margins in the left side of the mandible. B, Cone beam computed tomography showing a hypodense lesion with well-defined margins without resorption of the cortical bone or tooth.

Besides the use of glucocorticoids and calcitonin in CGCL treatment, it has been described the use of proteins that act in bone remodeling, such as denosumab.6,30 Previous studies have demonstrated that the bone resorption process in CGCL is similar to that which occurs in normal bone tissue.31-33 The bone resorption process in CGCL is well documented in the literature31-33; however, only a few studies have evaluated the osteogenic profile and bone formation capacity of this lesion.33,34 Osteocalcin (OC) is a protein secreted by osteoblasts and can be used as a marker of bone formation and osteoblastic activity.35 CGCLs express little to no OC,33,34 but a comparative evaluation between the CGCL variants has not been made. The investigation of OC expression in nonaggressive and aggressive lesions may provide new information regarding the bone repair process, differential diagnosis between aggressive and nonaggressive lesions, and new treatment approaches. The present study evaluated and compared the expression of GR, CTR, and OC in aggressive and nonaggressive CGCLs. Furthermore, the number of mitotic mononuclear cells and the number of multinucleated giant cells were evaluated in both groups of lesions, and we verified the association between cell number and lesion aggressiveness. This investigation may help to elucidate any differences between aggressive and nonaggressive CGCLs with regard to the expression of GR, CTR, and OC, thus providing evidence in favor of alternative treatments, such as the prescription of intralesional glucocorticoid and calcitonin, instead of surgical procedures.

MATERIAL AND METHODS Sample For this study, formalin-fixed paraffin-embedded samples from 31 cases of CGCL were selected from the archives of the Oral Pathology Laboratory of the Oral

Pathology Department, Dental School of Federal University of Goiás, Brazil. The inclusion criteria were as follows: well-preserved paraffin-embedded specimens from primary CGCL patients who did not receive prior treatment, as confirmed by assessment of the microscopic and clinical data; complete clinical and demographic records; specimens with sufficient tissue for analysis; and interpretable radiographic images. The cases that did not meet these criteria were excluded from the sample; cases of brown tumor of hyperparathyroidism and cherubism were also excluded. The clinical and radiologic data were retrieved from patient records. Review of the specimens and clinical data was conducted to confirm the CGCL diagnosis. These data were used to classify the lesions, based on criteria established by Chuong et al.,5 into nonaggressive lesions (n ¼ 20), which had little or no symptoms, no evidence of cortical bone or root resorption or teeth displacement, slow growth and low recurrence rates (Figures 1A and 1B), or aggressive lesions (n ¼ 11), which were characterized by pain, root resorption, tooth displacement, cortical bone perforation, rapid growth, and high rates of recurrence (Figures 2A and 2B). A rapid growth rate was considered when the patient’s records indicated that the lesion had evolved over a 6month period. A high rate of recurrence was defined as at least one recurrence within 24 months of the previous treatment (this period was established on the basis of reports by Chuong et al.5). In this specific study the patients were followed up for 12 to 84 months. This research was approved by the Institutional Review Board of the Federal University of Goiás (UFG: No. 556.175/2014). Immunohistochemical staining GR, CTR, and OC expression levels were assessed by immunohistochemical staining. For the staining

ORAL AND MAXILLOFACIAL PATHOLOGY 388 Martins et al.

OOOO September 2015

Fig. 2. A, Periapical radiographs showing an irregular osteolytic lesion with lamina dura resorption in the alveolar process of the maxilla, associated with endodontically treated teeth. B, Cone beam computed tomography showing a hypodense lesion with well-defined margins, corresponding to an aggressive central giant cell lesion (CGCL) with expansion and resorption of cortical platedcoronal view.

procedure, 3-mm sections were coated with 2% 3aminopropyltriethylsilane (Sigma Chemicals, St. Louis, MO), deparaffinized, and hydrated. The sections were incubated in a citrate buffer (pH ¼ 6.0) for 25 minutes (this step was excluded for the antibody anti-osteocalcin) and then incubated with peroxidase blocker (EasyLink One, EasyPath, Sao Paulo, Brazil). The sections were blocked by incubation with goat serum (Easy LinkOne) and then incubated at 4 C overnight with the following primary mouse and rabbit antibodies: anti-glucocorticoid receptor (BuGR2, ab2768, Abcam, Cambridge, MA), anti-calcitonin receptor (ab140778, Abcam) both diluted 1:100, and anti-osteocalcin (FL-100, sc30044; Santa Cruz Biotechnology, Dallas, TX) diluted 1:50. The sections were then treated using the HRP-polymer method (EasyLink One) and incubated in 3.3’-diaminobenzedine (Easy Link One) for 3 minutes for the GR and CTR antibodies and for 1 minute for the OC antibody. Finally, the sections were counterstained with Mayer’s hematoxylin. Human kidney tissue was used as a positive control for OC antibody staining, and human thyroid tissue and human tonsil tissue were used as positive controls for CTR and GR staining, respectively. Negative controls were obtained by the omission of the primary antibody and the use of rabbit or goat serum instead. GR, CTR, and OC analysis The staining analysis was assessed by counting the number of positive and negative cells in the samples based on the methods of Tobón-Arroyave et al.29 The quantitative analysis was evaluated in eight high-power fields of each case under 400 magnification using an integration graticule (4740680000000-Netzmikrometer 12.5; Carl Zeiss, Germany). Mononuclear and multinucleated giant cells were evaluated separately. The number of stained cells was divided by the total number of cells in each field; a percentage of the mean

in the eight high-power fields was used as the estimation for expression in each case. Each sample was then scored, according to the proportion of stained cells, as 0 ¼ no stained cells; 1 ¼ less than 25% stained cells; 2 ¼ 25% to 50% stained cells; or 3 ¼ greater than 50% stained cells. Qualitative analysis of the immunohistochemical staining was conducted to determine if the staining occurred in the cell nucleus, cytoplasm, or cytoplasmic membrane. Additionally, the staining intensity in each field was evaluated and compared with that of the control to classify each sample as 0 ¼ no staining; 1 ¼ weak staining; 2 ¼ moderate staining; and 3 ¼ intense staining. Furthermore, a staining-intensity-distribution (SID) score was calculated by multiplying the stained cell proportion score by the staining intensity score in each field. The mean of the eight fields was used as the SID score for that case. Mitotic figures and multinucleated giant cells The number of mitotic figures in mononuclear cells and the number of multinucleated giant cells (cells with three or more nuclei) were determined by assessment of slides stained with hematoxylin and eosin (H&E). The slides were evaluated in 10 consecutive fields under 400 magnification using an integration graticule (4740680000000-Netzmikrometer 12.5; Carl Zeiss, Germany) on a light microscope. Statistical analysis The statistical analysis was performed by using the IBM SPSS Statistics Software 20.0 for Windows. The nonparametric Mann-Whitney test was used to compare the proportion of stained cells and the SID score between clinical CGCL variants. The Spearman’s correlation coefficient was used to verify correlations.

OOOO Volume 120, Number 3

Statistical significance was set at P < .05. The results are expressed as medians (minimum and maximum values).

RESULTS Sample analysis All cases showed a predominance of female patients (58%), with a mean age of 28 years (standard deviation  22.2). The mandible was the most commonly affected site (61.3%). The clinical and radiographic features of the cases studied classified as nonaggressive and aggressive are summarized in Table I. GR expression analysis Positive immunohistochemical reactivity for GR was detected in the cell nuclei of mononuclear and multinucleated giant cells in all cases analyzed (Figures 3A and 3B). In the nonaggressive group, 50% of cases had high GR expression (percentage of positive cells higher than the median number of positive cells in all cases studied) in mononuclear or multinucleated giant cells. In the aggressive group, 54% of the lesions had high expression of this receptor. The percentage of mononuclear cells positive for GR expression was 33.21% in the nonaggressive group and 44.09% in the aggressive group (P ¼ .312). For multinucleated giant cells, the proportion of positive cells was 61.44% in the nonaggressive group and 67.5% in the aggressive group (P ¼ .901). The SID score analysis in mononuclear cells was 3 in the nonaggressive group and 6 in the aggressive group (P ¼ .411). In multinucleated giant cells, the SID score was 3 in the nonaggressive group and 5.02 in the aggressive group (P ¼ .659). GR expression data are summarized in Tables II and III. CTR analysis The immunoreactivity for CTR was detected in the cytoplasm and cytoplasmic membranes of mononuclear and multinucleated giant cells in 100% of the cases studied (Figures 4A and 4B). Higher CTR expression in mononuclear cells was observed in 50% of the cases in the nonaggressive group and in 55.54% of cases in the aggressive group. For CTR expression in multinucleated giant cells, high expression was observed in 55% of the nonaggressive lesions and in 45.45% of the aggressive lesions. In mononuclear cells, the percentage of cells positive for calcitonin receptor was 34.12% in the nonaggressive group and 32.27% in the aggressive group (P ¼ .836). CTR expression in multinucleated giant cells was 65.74% in the nonaggressive and 62.8% in the aggressive CGCLs (P ¼ .967). The SID scores for the mononuclear component was 4.69 for the nonaggressive group and 4.68 for the

ORIGINAL ARTICLE Martins et al. 389

Table I. Clinical features of the cases studied CGCL (n ¼ 31) Age in years (mean  SD) Gender Female Male Site Mandible Maxilla Symptoms Yes No Evolution in months (mean  SD) Cortical bone Preserved Perforation Teeth displacement or resorption Treatment Surgical Additional treatment Recurrence

28.0 (22.2)

Nonaggressive (n ¼ 20) 31 (24.4)

Aggressive (n ¼ 11) 22.8 (17.4)

58% (18) 42% (13)

60% (12) 40% (8)

54.5% (6) 45.5% (5)

61.3% (19) 38.7% (12)

55% (11) 45% (9)

72.7% (8) 27.3% (3)

29% (9) 0% (0) 71% (22) 100% (20) 12.4 (12.6) 14 (9.7)

81.1% (9) 18.9% (2) 9.6 (17.9)

67.7% (21) 32.2% (10) 35.4% (11)

100% (20) 0% (0) 0% (0)

9.1% (1) 90.9% (10) 100% (11)

100% (31) d

100% (20) d

100% (11) 27.2%* (3)

9.67% (3)

d

27.2% (3)

*The three recurrent cases were treated with a pharmacologic approach. Two were treated with calcitonin nasal spray. The other recurrent case was treated with corticosteroids together with calcitonin nasal spray.

aggressive group (P ¼ .612) (see Table II). In multinucleated giant cells, the score was 2.62 in the nonaggressive group and 3.43 in the aggressive group (P ¼ .265) (see Table III). OC analysis Immunoreactivity for OC was scarce in the cytoplasm of mononuclear and multinucleated giant cells (Figures 5A and 5B). However, a higher staining intensity was observed in mononuclear cells compared with multinucleated giant cells. In 55% of the nonaggressive cases and in 45.45% in the aggresive lesions it was noted that the OC expression in mononuclear cell was higher than the median for all cases analyzed. For the multinucleated giant cell, in 45% of the nonaggressive lesions and 63.63% of the aggressive cases it was noted that the OC expression was higher than the median expression for all cases studied. The OC expression in mononuclear cells was higher in the nonaggressive group (5.35%) compared with the aggressive forms (3.43%), although this difference was not statistically significant (P ¼ .409). In multinucleated giant cells, the percentage of positive cells was 4.76% in the nonaggressive group and 8.39% in the aggressive group (P ¼ .519). The SID score in the mononuclear component was 1.56 in the nonaggressive group and 1.62 in the aggressive group (P ¼ .725). With regard to the SID

ORAL AND MAXILLOFACIAL PATHOLOGY 390 Martins et al.

OOOO September 2015

Fig. 3. Immunohistochemistry staining showing mononuclear (asterisk) and multinucleated giant cells (black arrow) positive for glucocorticoid receptor expression (immunohistochemical staining; original magnification, 400). A, Nonaggressive lesion, staining-intensity distribution (SID) mononuclear cells ¼ 6; SID multinucleated giant cells ¼ 9. B, Aggressive lesion, SID mononuclear cells ¼ 6; SID multinucleated giant cells ¼ 7.5. A high-resolution version of this slide for use with the Virtual Microscope is available as eSlide: VM00684.

Table II. CTR, GR, and OC expression and SID score in mononuclear cells, results are expressed as median (minimum and maximum values) Nonaggressive (n ¼ 20) GR Expression (%) CTR Expression (%) OC Expression (%) GR SID score CTR SID score OC SID score

33.21 34.12 5.84 3 4.69 1.56

(5.53-100) (1.43-61.25) (0-28.8) (1-9) (0.375-8.250) (0-8.25)

Aggressive (n ¼ 11) 45.09 32.27 3.43 6 4.68 1.625

(15.62-78.08) (14.90-43.8) (0.61-23.57) (1-9) (1.875-6.5) (0.625-3.5)

Table III. CTR, GR, and OC expression and SID score in multinucleated giant cells results are expressed as median (minimum and maximum values)

P value* .312 .836 .409 .411 .612 .725

Nonaggressive (n ¼ 20) GR expression (%) CTR expression (%) OC expression (%) GR SID score CTR SID score OC SID score

61.44 65.74 4.76 3 2.62 0.375

(16-100) (8.77-100) (0-30.45) (1-9) (0.375-6.375) (0-6.375)

Aggressive (n ¼ 11) 67.5 62.80 8.39 5.02 3.43 0.875

(15-95.14) (4.6-96.8) (0-47.05) (1-8.25) (0.125-6.375) (0-5.875)

P value* .901 .967 .519 .659 .262 .299

*Mann-Whitney test.

*Mann-Whitney test.

score in multinucleated giant cells, the nonaggressive group presented a score of 0.375 and the aggressive group 0.875; however, this difference was not statistically significant (P ¼ .299). Tables II and III summarize the OC analysis in mononuclear and multinucleated giant cells, respectively.

between CTR expression in mononuclear cells and that in multinucleated giant cells (r ¼ 0.66; P < .01) (Figure 7B). In the aggressive group, a significant correlation was only observed between OC and CTR expression in multinucleated giant cells (r ¼ 0.718; P ¼ .01) (Figure 7C).

Mitotic figures and number of multinucleated giant cells The median number of mitotic figures in 10 high-power fields was 6 (range, 0-18) in the nonaggressive group and 1.0 (range, 4-18) in the aggressive group (P ¼ .262). The aggressive group showed a higher number of multinucleated giant cells (median ¼ 7.4; minimum ¼ 3.6; maximum ¼ 14.5; Figure 6A) compared with that of the nonaggressive group (median ¼ 4.6; minimum ¼ 1.2; maximum ¼ 11; Figure 6B) (P ¼ .01). Correlation tests A significant correlation between CTR expression in mononuclear cells and that in multinucleated giant cells could be observed (r ¼ 0.45; P < .01) (Figure 7A). The nonaggressive group showed a significant correlation

DISCUSSION Our results demonstrated that there is no difference between GR and CTR expression either in aggressive or nonaggressive CGCLs. OC expression was higher in the mononuclear component of the nonaggressive form; however, this difference was not statistically significant. These findings suggest that GR, CTR, and OC expression in mononuclear or multinucleated giant cells cannot distinguish between the aggressive and nonaggressive forms of CGCL. Because both types of CGCL have the same cell lineage, their aggressiveness can be related to other factors, such as the stromal vascular density of lesions36,37 or the production of matrix metalloproteinases.38 The diversity among the results of alternative CGCL treatments has encouraged investigations of GR and CTR expression.2,4,29,39 Two of these studies compared

OOOO Volume 120, Number 3

ORIGINAL ARTICLE Martins et al. 391

Fig. 4. Immunohistochemistry staining showing mononuclear (asterisk) and multinucleated giant cells (black arrow) positive for calcitonin receptor staining (immunohistochemical staining; original magnification, 400). A, Nonaggressive lesion, stainingintensity distribution (SID) mononuclear cells ¼ 7.125; SID multinucleated giant cells ¼ 6.375. B, Aggressive lesion, SID mononuclear cells ¼ 3.4; SID multinucleated giant cells ¼ 3.6. A high-resolution version of these slides for use with the Virtual Microscope are available as eSlide: VM00685 and VM00686.

Fig. 5. Immunohistochemistry staining showing mononuclear cells (asterisk) positive for (asterisk) and multinucleated giant cells (black arrow) negative for osteocalcin expression, the arrow head shows a woven bone showing positive staining for osteocalcin (OC) (immunohistochemical staining; original magnification, 400). A, Nonaggressive lesion, staining-intensity distribution (SID) mononuclear cells ¼ 1; SID multinucleated giant cells ¼ 0.125. B, Aggressive lesion, SID mononuclear cells ¼ 1.25; SID multinucleated giant cells ¼ 0. A high-resolution version of these slides for use with the Virtual Microscope are available as eSlide: VM00687 and VM00688.

the nonaggressive and aggressive variants.2,29 TobónArroyave et al.29 investigated receptor expression in 27 cases of nonaggressive lesions and 10 cases of aggressive lesions, and the authors found a significant difference in CTR expression. However, mononuclear and multinucleated giant cells were not evaluated in a distinct manner as in our study. The literature shows that mononuclear cells proliferate,9,10 whereas multinucleated giant cells are responsible for bone resorption.19,31,32 Thus, glucocorticoids and calcitonin have different mechanisms of action in these cells.12,20 An investigation including these types of cells is important to help understand the appropriate mechanism of action when therapeutic protocols are used. Nogueira et al.2 also studied GR and CTR expression in mononuclear and multinucleated giant cells in 18 cases of nonaggressive and aggressive lesions, and as in our results, they did not find any differences in the receptor expression between the two forms of CGCL in mononuclear or multinucleated giant cells.

Therapies for CGCL using calcitonin and glucocorticoids have been described in the literature, and the results of these therapies vary from complete resolution to lesion enlargement.12,13,15-17,21-28 Nogueira et al.2 showed that lesions with moderate or good response to intralesional glucocorticoid injection showed a higher expression of GR before the treatment. This variety in GR expression may explain why some lesions have a good response and others have a negative response to the use of intralesional glucocorticoids. Although there are reports about the successful use of calcitonin,15,18,39 failure of this treatment may be related to its prolonged use, which can cause an escape phenomenon that downregulates CTR expression.39,40 Due to the absence of a unique protocol establishing the dose to be prescribed for each patient, we believe that more studies describing either success or failure are required. In our study, expression were investigated only in formalin-fixed, paraffin-embedded CGCL material

ORAL AND MAXILLOFACIAL PATHOLOGY 392 Martins et al.

OOOO September 2015

Fig. 6. Microscopic features of central giant cell lesion (CGCL). The presence of multinucleated giant cells in the stroma of mononuclear cells (ovoid and spindle) is observed with the presence of hemorrhagic area. A, Nonaggressive lesion. B, Aggressive lesion; notice the higher amount of multinucleated giant cells (hematoxylin and eosin; original magnification, 200). A highresolution version of these slides for use with the Virtual Microscope are available as eSlide: VM00689 and VM00690.

Fig. 7. Correlations graphs. A, Correlation between calcitonin receptor (CTR) expression in mononuclear cells (MOC) and multinucleated giant cells (MGC) in central giant cell lesion (CGCL) (n ¼ 31; r ¼ 0.45, P < .01, Spearman’s correlation test). B, Graphic of the correlation between the CTR expression in MOC and MGC in the nonaggressive group of CGCL (n ¼ 20; r ¼ 0.66; P < .01, Spearman’s correlation test). C, Correlation between CTR and OC expression in MGC in the aggressive group (n ¼ 11; r ¼ 0.718; P ¼ .01, Spearman’s correlation test).

before the treatment, and our results, together with the results of previous studies,2,4,29,39 suggest that both aggressive and nonaggressive CGCLs can be treated with intralesional glucocorticoids and/or spray or subcutaneous injection of calcitonin. Immunohistochemical study of these receptors can be helpful in making decisions about treatment protocols.

To the best of our knowledge, this is the first study to investigate OC in nonaggressive and aggressive CGCLs. OC expression can be used to determine the level of osteoblast activity and bone formation.35 We believe that OC expression in mononuclear cells may reflect the intraosseous nature of the lesion or an attempt to produce bone against a resorption stimulus.

OOOO Volume 120, Number 3

Additionally, positive OC expression in cells of the CGCL stroma may indicate that at least some cells have bone-producing potential. In this study, we found little OC expression in both nonaggressive and aggressive cases of CGCLs. Park et al.34 and Elias et al.33 described the same expression pattern in previous studies, although these authors did not describe the aggressiveness of the cases studied. The reduced expression of OC in the cells of CGCLs may be related to the high osteolytic potential of the lesions. The excessive production of proteins related to bone resorption in the stroma of CGCLs33 may suppress the maturation of osteoblastic cells, thus diminishing OC production and expression, expression favoring the bone resorption process. Based on our results, we speculate that a lesser expression of OC in mononuclear cells in the aggressive group may reflect larger lesions with a higher potential to resorb bone, including the cortical plate. Associated with this, aggressive lesions may present a higher number of multinucleated giant cells compared with nonaggressive lesions,7-9 as was also observed in our study. It is hypothesized that multinucleated giant cells are osteoclastic cells formed by mononuclear cell fusion.31,32 Consistent with this hypothesis, we observed a positive correlation between CTR expression in mononuclear and multinucleated giant cells, which may indicate that giant cells are formed by the fusion of mononuclear cells expressing CTR because this receptor is a marker that indicates osteoclastic cell lineage.41 Previous studies have suggested that the stromal cells of giant cell tumors of the bone (GCTB) also express little to no OC.42,43 In a study conducted by Steensma et al.43 with cell cultures from fresh GCTB samples, it was observed that these cells exhibit enhanced OC expression when cultured in an osteogenic media, suggesting that the GCTB have an osteoblastic cell lineage that may initiate a bone mineralization process. Due to the similarities described in the literature between GCTB and CGCL44,45 and on the basis of our results, we believe that stromal cells will increase OC expression when stimulated by bone-remodeling agents, and thus an osteogenic process will be initiated. This therapeutic target can be an alternative treatment for CGCL. The number of mitotic cells has been associated with CGCL aggressiveness. Some authors have demonstrated that aggressive lesions have a higher number of mitotic figures.7-9 However, in our study, we did not find differences in the number of mitotic cells between aggressive and nonaggressive lesions. This difference may be due to different methodologies applied; we investigated the number of mitotic figures in 10 highpower fields, whereas the other authors investigated

ORIGINAL ARTICLE Martins et al. 393

the same parameter within 25 high-power fields. Consistent with our findings, O’Malley10 investigated cell proliferation in CGCLs, using Ki-67 protein expression, and did not find a difference between the two forms in terms of positive cells among 10 fields. Our results provide new information about bone repair capacity in CGCLs through the investigation of OC expression. Due to the limited number of cases and bone repair markers evaluated, more studies should be conducted to assess the role of osteogenic cells in CGCLs and also to provide evidence to justify treatments that induce bone repair through the activation of OC-positive cells. In summary, our findings support that the positive expression of GR and CTR in CGCLs can inform choices of alternative treatments with intralesional glucocorticoids and calcitonin in aggressive or nonaggressive CGCLs. This type of therapy may reduce or avoid the defects caused by surgical treatment.

CONCLUSIONS GR, CTR, and OC expression levels are not different between the nonaggressive and aggressive forms of CGCL; however, the number of multinucleated giant cells may be helpful in distinguishing these types of lesions. The presence of GR and CTR gives a biologic justification for the use of glucocorticoids and calcitonin as alternative treatment approaches for aggressive and nonaggressive CGCLs. The reduced OC expression observed in our study may be due to the osteolytic behavior of the lesion. We believe that OC-positive cells may promote bone repair when stimulated by osteogenic agents. More studies about osteogenic cells in CGCLs should be conducted to help elucidate their role. REFERENCES 1. Nogueira RL, Faria MH, Osterne RL, et al. Central giant cell lesion of the jaws: study of CCND1 gene amplification and p16 INK4 a protein levels. J Mol Histol. 2013;44:527-534. 2. Nogueira RL, Faria MHG, Osterne RLV, Cavalcante RB, Ribeiro RA, Rabenhorst SHB. Glucocorticoid and calcitonin receptor expression in central giant cell lesions: implications for therapy. Int J Oral Maxillofac Surg. 2012;41: 994-1000. 3. Regezi JA, Pogrel MA. Comments on the pathogenesis and medical treatment of central giant cell granulomas. J Oral Maxillofac Surg. 2004;62:116-118. 4. Vered M, Buchner A, Dayan D. Immunohistochemical expression of glucocorticoid and calcitonin receptors as a tool for selecting therapeutic approach in central giant cell granuloma of the jawbones. Int J Oral Maxillofac Surg. 2006;35:756-760. 5. Chuong R, Kaban LB, Kozakewich H, Perez-Atayde A. Central giant cell lesions of the jaws: a clinicopathologic study. J Oral Maxillofac Surg. 1986;44:708-713. 6. de Lange J, van den Akker HP, van den Berg H. Central giant cell granuloma of the jaw: a review of the literature with emphasis on

ORAL AND MAXILLOFACIAL PATHOLOGY 394 Martins et al.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18. 19.

20.

21.

22.

23.

24.

therapy options. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2007;104:603-615. Reddy V, Saxena S, Aggarwal P, Sharma P, Reddy M. Incidence of central giant cell granuloma of the jaws with clinical and histological confirmation: an archival study in Northern India. Br J Oral Maxillofac Surg. 2012;50:668-672. Ficarra G, Kaban LB, Hansen LS. Central giant cell lesions of the mandible and maxilla: a clinicopathologic and cytometric study. Oral Surg Oral Med Oral Pathol. 1987;64:44-49. Kruse-Losler B, Diallo R, Gaertner C, Mischke KL, Joss U, Kleinheinz J. Central giant cell granuloma of the jaws: a clinical, radiologic, and histopathologic study of 26 cases. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2006;101:346-354. O’Malley M, Pogrel MA, Stewart JC, Silva RG, Regezi JA. Central giant cell granulomas of the jaws: phenotype and proliferation-associated markers. J Oral Pathol Med. 1997;26: 159-163. Rawashdeh MA, Bataineh AB, Al-Khateeb T. Long-term clinical and radiological outcomes of surgical management of central giant cell granuloma of the maxilla. Int J Oral Maxillofac Surg. 2006;35:60-66. Nogueira RL, Teixeira RC, Cavalcante RB, Ribeiro RA, Rabenhosrt SH. Intralesional injection of triamcinolone hexacetonide as an alternative treatment for central giant-cell granuloma in 21 cases. Int J Oral Maxillofac Surg. 2010;39:1204-1210. Allon DM, Anavi Y, Calderon S. Central giant cell lesion of the jaw: nonsurgical treatment with calcitonin nasal spray. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2009;107:811-818. De Lange J, Van den Akker HP. Clinical and radiological features of central giant-cell lesions of the jaw. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2005;99:464-470. Borges HO, Machado RA, Vidor MM, Beltrão RG, Heitz C, Filho MS. Calcitonin: a non-invasive giant cells therapy. Int J Pediatr Otorhinolaryngol. 2008;72:959-963. Kaban LB, Mulliken JB, Ezekowitz RA, Ebb D, Smith PS, Folkman J. Antiangiogenic therapy of a recurrent giant cell tumor of the mandible with interferon alfa-2 a. Pediatrics. 1999;103: 1145-1149. Carlos R, Sedano HO. Intralesional corticosteroids as an alternative treatment for central giant cell granuloma. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2002;93:161-166. Pogrel MA. Calcitonin therapy for central giant cell granuloma. J Oral Maxillofac Surg. 2003;61:649-653:discussion 653-654. Flanagan AM, Nui B, Tinkler SM, Horton MA, Williams DM, Chambers TJ. The multinucleate cells in giant cell granulomas of the jaw are osteoclasts. Cancer. 1988;62:1139-1145. de Lange J, Rosenberg AJ, van den Akker HP, Koole R, Wirds JJ, van den Berg H. Treatment of central giant cell granuloma of the jaw with calcitonin. Int J Oral Maxillofac Surg. 1999;28:372-376. da Silva NG, Carreira AS, Pedreira EN, Tuji FM, Ortega KL, de Jesus Viana Pinheiro J. Treatment of central giant cell lesions using bisphosphonates with intralesional corticosteroid injections. Head Face Med. 2012;8:23. Rubio-Correa I, Manzano-Solo de Zaldívar D, GonzálezGarcía R, et al. Giant cell granuloma of the maxilla. Global management, review of literature and case report. J Clin Exp Dent. 2012;4:e129-e131. da Silva Sampieri MB, Yaedú RY, Santos PS, et al. Central giant cell granuloma: treatment with calcitonin, triamcinolone acetonide, and a cystic finding 3 years and 6 months after the primary treatment. Oral Maxillofac Surg. 2013;17:229-234. de Lange J, van den Akker HP, van den Berg H, Richel DJ, Gortzak RA. Limited regression of central giant cell granuloma by interferon alpha after failed calcitonin therapy: a report of 2 cases. Int J Oral Maxillofac Surg. 2006;35:865-869.

OOOO September 2015 25. de Lange J, van den Akker HP, Veldhuijzen van Zanten GO, Engelshove HA, van den Berg H, Klip H. Calcitonin therapy in central giant cell granuloma of the jaw: a randomized double-blind placebocontrolled study. Int J Oral Maxillofac Surg. 2006;35:791-795. 26. de Lange J, van Rijn RR, van den Berg H, van den Akker HP. Regression of central giant cell granuloma by a combination of imatinib and interferon: a case report. Br J Oral Maxillofac Surg. 2009;47:59-61. 27. Goldman KL, Marshall MK, Alessandrini E, Bernstein ML. Complications of alpha-interferon therapy for aggressive central giant cell lesion of the maxilla. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2005;100:285-291. 28. Schutz P, El-Bassuoni KH, Munish J, Hamed HH, Padwa BL. Aggressive central giant cell granuloma of the mandible. J Oral Maxillofac Surg. 2010;68:2537-2544. 29. Tobon-Arroyave SI, Franco-González LM, Isaza-Guzmán DM, et al. Immunohistochemical expression of RANK, GRalpha and CTR in central giant cell granuloma of the jaws. Oral Oncol. 2005;41:480-488. 30. Schreuder WH, Coumou AW, Kessler PA, de Lange J. Alternative pharmacologic therapy for aggressive central giant cell granuloma: denosumab. J Oral Maxillofac Surg. 2014;72:1301-1309. 31. Liu B, Yu F, Li TJ. Multinucleated giant cells in various forms of giant cell containing lesions of the jaws express features of osteoclasts. J Oral Pathol Med. 2003;32:367-375. 32. Itonaga I, Hussein I, Kudo O, et al. Cellular mechanisms of osteoclast formation and lacunar resorption in giant cell granuloma of the jaw. J Oral Pathol Med. 2003;32:224-231. 33. Elias LS, Costa RF, Carvalho MA, et al. Markers of bone remodeling in neoplastic and bone-related lesions. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2010;110:624-631. 34. Park YK, Yang MH, Kim YW, Park HR. Osteocalcin expression in primary bone tumorsdin situ hybridization and immunohistochemical study. J Korean Med Sci. 1995;10:263-268. 35. Razzaque MS. Osteocalcin: a pivotal mediator or an innocent bystander in energy metabolism? Nephrol Dial Transplant. 2011;26:42-45. 36. Dewsnup NC, Susaria SM, Abulikemu M, Faquin WC, Kaban LB, August M. Immunohistochemical evaluation of giant cell tumors of the jaws using CD34 density analysis. J Oral Maxillofac Surg. 2008;66:928-933. 37. Peacock ZS, Jordan RC, Schmidt BL. Giant cell lesions of the jaws: does the level of vascularity and angiogenesis correlate with behavior? J Oral Maxillofac Surg. 2012;70:1860-1866. 38. Tobón-Arroyave SI, Mideros-Simarra SM, Castaño-Ramírez LM, Flórez-Moreno GA, Isaza-Guzmán DM. Overexpression of matrix metalloproteinase (MMP)-1 and -9 in central giant cell lesions of the jaws: implications for clinical behavior. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2010;110:755-763. 39. Vered M, Shohat I, Bucnher A, Dayan D, Taicher S. Calcitonin nasal spray for treatment of central giant cell granuloma: clinical, radiological, and histological findings and immunohistochemical expression of calcitonin and glucocorticoid receptors. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2007;104:226-239. 40. Pondel M. Calcitonin and calcitonin receptors: bone and beyond. Int J Exp Pathol. 2000;81:405-422. 41. Quinn JM, Morfis M, Lam MH, et al. Calcitonin receptor antibodies in the identification of osteoclasts. Bone. 1999;25:1-8. 42. Huang L, Teng XY, Cheng YY, Lee KM, Kumta SM. Expression of preosteoblast markers and Cbfa-1 and Osterix gene transcripts in stromal tumour cells of giant cell tumour of bone. Bone. 2004;34:393-401. 43. Steensma MR, Tyler WK, Shaber AG, et al. Targeting the giant cell tumor stromal cell: functional characterization and a novel therapeutic strategy. PLoS One. 2013;8:e69101.

OOOO Volume 120, Number 3 44. Kashyap B, Reddy SP, Desai R, Puranik RS, Vanaki SS. Computer assisted histomorphologic comparison and the expression of AgNORs in the central and peripheral giant cell lesions of the oral cavity and giant cell tumor of the long bone. J Oral Maxillofac Pathol. 2014;18:S54-S59. 45. Resnick CM, Margolis J, Susaria SM, et al. Maxillofacial and axial/appendicular giant cell lesions: unique tumors or variants of the same disease? A comparison of phenotypic, clinical, and radiographic characteristics. J Oral Maxillofac Surg. 2010;68: 130-137.

ORIGINAL ARTICLE Martins et al. 395 Reprint requests: Prof. Elismauro Francisco Mendonça, DDS, MSc, PhD Department of Stomatologic Sciences Federal University of Goiás Praça Universitária s/n Setor Universitário CEP 74605-220 Goiânia, GO, Brazil [email protected]