Journal of Oral and Maxillofacial Pathology

: 2008  |  Volume : 12  |  Issue : 1  |  Page : 8--15

Correlation of clinico-pathologic features and AgNOR counts between aggressive and nonaggressive central gaint cell lesions

Aarti Mahajan, SM Ganvir, VK Hazarey 
 Department of Oral Pathology, Govt. Dental College and Hospital, Ajni, Nagpur, Maharashtra, India

Correspondence Address:
Aarti Mahajan
«DQ»Savali«DQ», Rameshwar Nagar, Cidco-Ambad Link Road, Nashik - 422 010


The aim of this study was to review all the cases of central giant cell lesions (CGCLs) of the jaws which appeared in archives of Government Dental College and Hospital, Nagpur, during the period 1978-2004. CGCLs which were termed «SQ»aggressive«SQ» based on clinical behavior were compared with «SQ»non-aggressive«SQ» lesions in order to evaluate if any histologic difference existed between them. In addition, a study of silver (Ag) nucleolar organizing regions (AgNOR) counts was performed to delineate lesions of varying behavior. A total of 34 cases were examined for clinical and radiologic features, and 27 cases were evaluated for histopathology. We identified statistically significant differences between aggressive and nonaggressive lesions with respect to certain histological characteristics such as shape of mononuclear cells, shape of giant cells, and number of nuclei per giant cell. It was also found that there was a definite and statistically significant correlation between the aggressiveness of the lesion and the AgNOR count.

How to cite this article:
Mahajan A, Ganvir S M, Hazarey V K. Correlation of clinico-pathologic features and AgNOR counts between aggressive and nonaggressive central gaint cell lesions.J Oral Maxillofac Pathol 2008;12:8-15

How to cite this URL:
Mahajan A, Ganvir S M, Hazarey V K. Correlation of clinico-pathologic features and AgNOR counts between aggressive and nonaggressive central gaint cell lesions. J Oral Maxillofac Pathol [serial online] 2008 [cited 2022 Jul 1 ];12:8-15
Available from:

Full Text


The central giant cell lesions (CGCLs) of the jaws are distinctive benign osteolytic lesions that occur in the mandible or maxilla. They were first described in the jaws by Jaffe and designated giant cell reparative granuloma. This term however has been discarded by most oral and maxillofacial pathologists, who use the term giant cell granuloma or the more noncommittal term 'giant cell lesions'.

The question whether the giant cell tumors which most often occur in the epiphyses of long bones occur in the jaws has been argued for many years and still is unresolved. [1]

Although most of the jaw lesions are asymptomatic and slow growing, which may be discovered during a routine radiographic examination, a minority of cases are characterized by pain, rapid expansion, cortical perforation, root resorption, and a tendency to recur after treatment. Based on these characteristics, the lesions are classified as 'aggressive' or 'nonaggressive'.

Histomorphologic studies have shown that majority of jaw lesions can be distinguished from giant cell tumors on histologic grounds, though a smaller number of jaw lesions fall within the histologic profile of giant cell tumors of long bones. Few cases of 'true' giant cell tumors of the jaws have been described in literature. [2],[3]

pathologists have attempted, with varying results, to identify histopathologic parameters in order to predict clinical behavior and prognosis of central giant cell granulomas. [4],[5] Immunohistochemical studies have shown that no phenotypic differences were found between aggressive and nonaggressive lesions. [6]

Several studies have suggested that AgNOR enumeration is useful in estimating the proliferative potential of neoplasms in the practice of surgical pathology. These nucleolar organizer regions (NORs) are loops of DNA that transcribe for ribosomal RNA and are present within the nucleolus (at cellular interphase) and are located in the acrocentric chromosomes 13, 15, 21 and 22. Since these NORs are associated with argyrophylic, nonhistonic acidic proteins, they can be stained with a silver-staining technique. This simple technique gives the pathologist objective data to evaluate.

This study intended to present clinicopathological findings from 34 cases of central giant cell lesions. An attempt was made to correlate histologic features with clinical behavior. We further questioned whether AgNOR quantification is useful in delineating lesions exhibiting variable behavior.

 Materials and Methods

For the period 1978-2004, the files of the Department of Oral Pathology and Microbiology, Government Dental College and Hospital, Nagpur, were examined for lesions coded as central giant cell granuloma, central giant cell tumor, or central giant cell lesions among a total of 9000 specimens. Cases of aneurysmal bone cyst, hyperparathyroidism, and cherubism were not included in the present study.

A total of 34 cases were accepted as examples of central giant cell lesions. The case histories of these 34 patients were analyzed for age, sex, localization of the lesion, signs and symptoms, radiological appearances, treatment, and recurrences.

Based on clinical criteria, the 31 cases were categorized as aggressive and nonaggressive. Aggressive lesions were characterized by the presence of pain, rapid growth, root resorption, cortical perforation, and a tendency to recur after treatment. [7]

Nonaggressive lesions were characterized by the absence of symptoms or the presence of minimal symptoms, slow growth, the absence of root resorption or cortical perforation, and low recurrence rate.

Standard H and E - stained histologic sections were available in 27 of the 34 cases. The histological sections were reviewed independently by each of us without the knowledge of clinical behavior of the lesions. The histological features of the mononuclear cells and stroma and giant cells were assessed. Each feature was graded on a 0 to 3+ scale. The features assessed are listed in [Table 1].

The predominant shape of the mononuclear cells was evaluated categorically as predominantly spindle, predominantly ovoid, or mixed. A high level of agreement was obtained between the observers.

In order to enumerate AgNOR numbers in the 2 clinical groups (aggressive and nonaggressive CGCL), 4 cases representative of each group were selected, and 4-μm sections were made from each block. These were stained according to the AgNOR method, which is briefly as follows. 4mm thick sections were cut from formalin-fixed paraffin blocks. Following dewaxing and rehydration through descending grades of alcohol, the slides were washed in de-ionized water for 10 min. The slides were then dipped in freshly prepared silver colloidal solution at room temperature for 45 min and then kept in the dark for incubation.The slides were then washed thoroughly in de-ionized water for 15 to 20 min. The sections were washed in running de-ionized water for 20 to 30 min. They were then dehydrated through ascending grades of alcohol for 2 min each in 50%, 70%, 90%, and absolute alcohol respectively. The sections were then cleared in xylene and mounted in DPX mounting medium.

Routine light microscopy was utilized, and the AgNORs were evaluated under ×1000 magnification and oil immersion. One hundred nuclei from mononuclear cells and one hundred nuclei from multinucleated giant cells were observed for each case, and the AgNOR dots were counted. The counting protocol suggested by Crocker et al. [8] was followed. The data was expressed as mean ± standard deviation, and the results were evaluated by using the one-way ANOVA analysis. The Rank-Sum Mann-Whitney test was employed to correlate histologic findings with clinical behavior.


A total of 34 cases with central giant cell lesions of the jaws were included in this study. [Table 2] briefly illustrates the patient-case demographics.

Age and gender

The range of age was from 9 to 55 years, the mean age was 23 years, and 71% of the lesions appeared before the age of 30 years. A 56% female predilection was noted.


Twenty-three (67%) cases were in the mandible, and 11 cases were in the maxilla. The posterior mandible was more involved in the mandibular lesions, and anterior maxilla was more involved in the maxilla. The mandibular condyle was involved in 1 case.

Radiographic findings

Radiographic findings varied from small lesions with well-delineated borders to large multilocular lesions with noncorticated borders. One (2.94%) case showed evidence of root resorption, 3 (8.82%) lesions showed displaced roots, and 3 (8.82%) lesions showed cortical perforation [Table 3].

Clinical behavior and symptoms

Rapid enlargement was seen in 12 (35.29%) cases, 7 (20.58%) cases exhibited pain, 1 lesion exhibited paresthesia, mobility was seen in 1 case, and recurrence was observed in 2 cases.

From a compilation of clinical and radiologic data, a conclusion was arrived at that 18 lesions, i.e., 52.94%, exhibited aggressive and 16 (47.05%) cases exhibited nonaggressive behavior as defined by Kaban et al. [7] Lange JD et al. [9] have shown that clinically and radiologically, a difference between aggressive and nonaggressive lesions can be made.

Histologic features

All lesions showed certain common histologic features such as presence of multinucleated giant cells and stroma composed of ovoid- to spindle-shaped mononuclear cells [Figure 1].

Significant differences among histologic features

The comparison and differences in the histologic features between aggressive and nonaggressive lesions have been discussed in [Table 4],[Table 5],[Table 6].


The mean AgNOR count of mononuclear and multinuclear cells in aggressive and nonaggressive lesions is given in [Table 7].


Age, gender, and location

Central giant cell lesions (CGCLs) are relatively uncommon lesions, with 34 cases in the present study accounting for 0.37% of all the pathological specimens sent to our department. Waldron and Shafer [10] found a 0.17% incidence of this lesion in their review of 20,000 oral biopsies. Motamedi et al. [11] found an incidence of 9.29% of CGCLs in their study.

The mean age of occurrence in the present study was 23 years, with 71% of the lesions seen before the age of 30 years. These findings are similar to those of Whittaker and Waldron, [12] who observed a mean age of 23 years and 64% occurrence before the age of 30 years.

Mandible was more commonly involved (67%), which is the most common location for intraoral giant cell lesions. Motamedi et al. [11] found a mandibular involvement of 70.58% in their study. The posterior mandible was more commonly involved than the anterior mandible, which is in contrast to the traditional view that CGCLs predominantly affect the deciduous tooth - bearing area. Kaffe et al., [13] however, have found posterior mandible involvement in 53% of cases. In the maxilla, anterior maxilla was more commonly affected.

One case of CGCL also presented in the mandibular condyle, which is a rare presentation. There are only a few case reports in literature regarding the involvement of this region [14] .

Radiologic features, recurrence, and aggressiveness

The present study used the criteria suggested by Kaban et al. , [7] to separate the cases that were considered on clinical and radiologic grounds to be potentially aggressive lesions from those that were considered to be nonaggressive.

Compiling together the lesions exhibiting root resorption, cortical perforation, rapid expansion, pain and paresthesia, and a tendency to recur after treatment, 18 (52.94%) of the 34 lesions exhibited aggressive behavior [Table 3].

Histologic features

Various studies have been attempted to ascertain which, if any, histologic features could be used to separate those lesions that were aggressive from those which exhibited indolent behavior. Losler et al. [5] proved that there is a significant increase in large giant cells, fractional surface area, and mitotic activity in aggressive lesions. Kaban et al. [7] showed a higher relative size index of giant cells in aggressive lesions. Ficarra et al. [4] employed computer-assisted image analysis and found higher number of giant cells and greater fractional surface area in clinically aggressive lesions. Auclair et al., [15] using similar cytometric methods, were unable to find any significant histologic difference between recurrent and nonrecurrent lesions. Liu et al. [16] found it difficult to predict the clinical behavior by its histopathologic patterns only. Vered et al. [17] showed that though myofibroblasts were an integral component of stromal cells in CGCL, the mean core density could not distinguish between aggressive and nonaggressive lesions.

Out of the 27 cases that were histologically evaluated, 16 belonged to the group of aggressive lesions and 11 fell in the category of nonaggressive lesions.

Statistically significant differences were observed in the two groups in the following histopathological features, i.e., number of nuclei per multinucleated giant cell, shape of giant cells and shape of mononuclear cells, presence of fresh hemorrhage, and foci of chronic inflammatory cell infiltration [Table 4],[Table 5],[Table 6].

The presence of giant cells with more than 20 nuclei occurred in a larger proportion of aggressive lesions [Figure 2], which can be correlated with the findings of Auclair et al. [15] and Franklin et al., [18] who found higher number of nuclei in giant cell tumors as compared to the corresponding number in central giant cell granulomas of the axial skeleton [Table 6].

The finding of the presence of odd-shaped giant cells in nonaggressive lesions in the present study is in consonance with the corresponding finding by Whitaker and Waldron, [19] who speculated that these odd-shaped giant cells were truly functional as in a 'reactive'-type process whereas a true neoplasm would have few-to-no 'functional' cells. A categoric evaluation of the shape of mononuclear cells in the present study led us to the conclusion that in aggressive lesions, the shape was predominantly ovoid as opposed to predominantly spindle or a mixture of ovoid and spindle [Table 5]. Ovoid or plump stromal cells have been noted as a feature of giant cell tumors of the long bones.

Greater stromal hemorrhage and lesser chronic inflammatory cell infiltration was observed in aggressive lesions as opposed to those in nonaggressive lesions [Table 6].

AgNOR findings

The AgNORs were distinctly stained in black as dots or blobs that were either rounded or elongated while the rest of the nucleus stained brown. They were identified in the nuclei of both mononuclear and multinuclear cells [Figure 3],[Figure 4],[Figure 5]. The difference with regard to the mean AgNOR count in the mononuclear cells of aggressive (2.11 ± 0.17) and nonaggressive (1.45 ± 0.17) lesions was statistically significant ( P value = 0.0013). In the multinucleated giant cells too, the mean number of AgNORs was higher in aggressive lesions (1.74 ± 0.14) as compared to that in nonaggressive lesions (1.29 + 0.10) ( P value = 0.0021) [Table 7]. Similar studies by Whitaker et al. [19] have found higher AgNOR count in aggressive lesions as compared to that in nonaggressive lesions.

Various studies have correlated AgNOR count with clinical behavior in various types of neoplasms. [20],[21] In a study of breast carcinoma, it was found that clinical behavior correlated better with AgNOR counts rather than with histological parameters.

It has been proposed that AgNORs can be equated with proliferative activity. It has been found that AgNORs are larger and less in number in benign lesions, whereas they are smaller and numerous in number in malignant lesions.

Argyrophilia of the NOR proteins acts as a marker of ribosomal DNA and possibly the level by transcription and therefore could provide useful information about the structure of the nucleolus and nucleolar activity in hyperplastic and neoplastic conditions.

Therefore, this procedure acts as an adjunct to routine histopathology to differentiate central giant cell lesions that carry a higher risk of subsequent malignancy from those that have a benign outcome. We favor the concept that CGCLs of the small bones, CGCLs of the jaw, and CGCLs of the axial skeleton form a spectrum with CGCLs of the small bones representing the benign end of the spectrum.

A higher AgNOR count in CGCLs would therefore be indicative of a more aggressive course and would alert the surgeon to follow a more radical approach towards the management of such a lesion, thereby administering a more appropriate treatment therapy for CGCL and thereby ensuring a better outcome of treatment.

This study has found a correlation of a few histological parameters and AgNOR enumeration with the biological behavior of these lesions. However, use of image analysis would have enabled us to study dispersion pattern and size of NORs and make an assessment of histological parameters more accurately.


1Neville BW, Damm DD, Allen CM, Bouquot JE.: Oral and maxillofacial pathology, 2 nd ed. Philadelphia: Saunders; 2002.
2Shklar G, Meyer I. Giant cell tumors of the mandible and maxillae. Oral Surg Oral Med Oral Pathol 1967;14:809-27.
3Mintz G. Abrams AM, Carlsen GD, Melrose RJ, Fister HW. Primary malignant giant cell tumor of the mandible Report of a case and review of the literature. Oral Surg Oral Med Oral Pathol 1981;51:164-71.
4Ficarra 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;64:44-9.
5Kruse-Losner B, Diallo R, Gaertner C, Mischke KL, Joos U, Kleinheinz J. Central giant cell granuloma of the jaws: A clinical, radiologic and histopathologic study of 26 cases. Oral Surg Oral Med Oral Path Oral Radiol Endod 2006;101:346-54.
6Pogrel MA, O'Malley M, Stewart JC, Silva RG, Regezi JA. Central giant cell granulomas of the jaws: Phenotype and proliferation associated markers. J Oral Pathol Med 1997;26:159-63.
7Kaban BL Chuong R, Kozakewich H, Atayde AP. Central giant cell lesions of the jaws: A clinicopathologic study. J Oral Maxillofac Surg 1986;44:708-13.
8Crocker J, Boldy DA, Egan MJ. How should we count AgNORs? Proposal for a standardized approach. J Pathol 1989;158:185-8.
9Lange JD, Akker Hp, Berg HV. Central giant cell granuloma of the jaw: A review of the literature with emphasis on therapy options. Oral Surg Oral Med Oral Path Oral Radiol Endod 2007;104:603-15.
10Waldron LA, Shafer WG. The central reparative giant cell granuloma of the jaws. Am J Clin Pathol 1966;45:437-47.
11Motamedi MH, Eshqhyar N, Jafari SM, Lassaemi E, Navi F, Abbas FM, et al . Peripheral and central giant cell granulomas of the jaws: A demographic study. Oral Surg Oral Med Oral Path Oral Radiol Endod 2007;103):e3q-43.
12Whitaker BS, Waldron CA. Central giant cell lesions of the jaws: A clinical, radiologic and histopathologic study. Oral Surg Oral Med Oral Pathol 1993;75:199-208.
13Kaffe I, Ardekian I, Taicher S, Littner MM, Buchner A. Radiologic features of central giant cell granuloma of the jaws. Oral Surg Oral Med Oral Pathol 1996;81:720-6.
14Abu-El-Naaj I, Ardekian L, Liberman R, Peled M. Central giant cell granuloma of the mandibular condyle: A rare presentation. J Oral Maxillofac Surg 2002;60:939-41.
15Auclair PL, Cuenin P, Kratochvil LF, Slater LJ, Ellis GL. A clinical and histomorphologic comparison of the central giant cell granuloma and the central giant cell tumor. Oral Surg Oral Med Oral Pathol 1998;66:197-208.
16Liu B, Yu SF, Wu YT, Pang SZ. Central giant cell lesions of the jaws: A clinicopathologic study of 31 cases. Zhonghua.Kou Qiang Yi Xue Za Zhi 2005;40:67-9.
17Vered M, Nasrallah W, Buchner A, Dayan D. Stromal myofibroblasts in Central giant cell granulomas of the jaws cannot distinguish between aggressive and non aggressive lesions. J Oral Pathol Med 2007;36:495-500.
18Franklin OD, Craig GT, Smith CJ. Quantitative analysis of histologic parameters in giant cell lesions of the jaws and long bones. Histopathology 1979;3:511-22.
19Whitaker SB, Vigneshwaran N, Budnick SD, Waldron CA. Giant cell lesions of the jaws: Evaluation of nucleolar organizer regions in lesions of varying bheaviour. J Oral Pathol 1993;22:402-5.
20Eusebi V, Caltani MG, Lamovec J, Trere D, Ceccarelli C, Veromesi C, et al. Prognostic relevance of silver stained nucleolar proteins in sarcomatoid carcinomas of the breast. Ultrastruct Pathol 1991;15:203-14.
21Ohno T, Tanaka T, Takeuchi S, Matsunga T, Mori H. Nuclear organizer regions in bone tumors. Clin Orthop Relat Res 1991;272:287-91.