Ⅰ.Introduction
Calcifying cystic odontogenic tumor (CCOT) has been recently classified by WHO, which was previously known as calcifying odontogenic cyst (COC)1,2). COC was first reported by Gorlin in 1962, thereafter, great controversy was made in clinical and histopathological presentations of COC with the characteristic nature of ghost cells. However, many authors still prefer to diagnose in the term of COC, when the lesion exhibits unique cystic lining of calcifying odontogenic epithelium with no feature of tumorous growth3-6).
CCOT is considered as a developmental odontogenic cyst in the jaw7), because it is characterized with frequent calcification and ossification associated with ghost cells8,9). Its incidence is usually peaked at the second decade, and it is more affected in maxilla than the mandible, with a predilection for the canine-premolar region10). However, many authors investigated its diverse histopathologic features and biological behavior, particularly focused on the ghost cells in comparison with ameloblastoma, dentinogenic ghost cell tumor (DGCT), and ghost cell odontogenic carcinoma (GCOC) 6,11,12). In the series of immunohistochemical detection using antisera of keratins (PKK1, KL1, TK, Ck6, Ck13, Ck14, Ck18, and Ck19), involucrin, carcinoembryonic antigen, epithelial membrane antigen, and S-100 protein in ghost cells, it was suggested that ghost cells were “keratinizing” odontogenic cells showing aberrant differentiation, metaplastic with altered synthesis of homogenous acellular materials, and that these cells finally became transformed into calcifying epithelium containing dystrophic calcified masses resembling to the altered differentiation of hair follicles13-15).
Ghost cells are anucleate cells with homogeneous pale eosinophilic cytoplasm, and also showed strong reaction of Notch1 and Jagged1. As Notch signaling is an evolutionarily conserved mechanism that enables adjacent cells to adopt different fates, the Notch1-Jagged1 signaling might serve as the main transduction mechanism in cell fate decision for ghost cells16). However, these anucleated ghost cells showed rare mitotic activity in the immunostainings of PCNA, Ki-67, and p5317-19).
The present study is aimed to detect the pathogenetic expression of ghost cells in a case of CCOT, which showed the conspicuous neoplastic features with complex odontoma, through immunohistochemical staining using different antisera relevant to cellular differentiation and protection. The results were discussed with the review of literature.
Ⅱ.Case report
A nineteen years old male patient showed a cystic lesion in left maxillary canine to premolar area (#23-#25). This lesion was asymptomatic, and found during his routine radiological check in local clinic. In the radiological observation pantomogram showed a round radiolucent image in left anterior maxilla, which was expanding as tumorous growth, resulted in the displacement of canine and second premolar (#23 and #25). There was absent of first premolar (#24) (Fig. 1 A). CT also disclosed that the lesion was bulging out the labial cortical plate, and also contained multiple granular radiopaque materials (Fig. 1 B). Therefore, this lesion was primarily diagnosis as adenomatoid odontogenic tumor (AOT).
The lesion was surgically enucleated, and a cystic fibrous tissue together with abnormal teeth was removed and examined pathologically. The biopsy specimen was fixed in 10% neutral buffered formalin, processed routinely, and embedded in paraffin. Histologic sections in 4 μm thickness were mounted on glass slides, stained with hematoxylin and eosin. And serial microsections were also prepared for immunohistochemical staining using antisera of proliferating cell nuclear antigen (PCNA), p53, Ki-67 (DAKO, Denmark), β-catenin, Wnt1, growth arrest and DNA damage 45 (GADD45), microtubule-associated proteins 1A/1B light chain 3A (LC3), B-cell lymphoma 2 (BCL2), heat shock protein-70 (HSP-70), p38, and matrix metalloprotease-9 (MMP-9) (Santa Cruz Biotech. USA). The immunohistochemical reaction protocols differed according to the target antigen and manufacturers’ protocols. But briefly, after deparaffinization and rehydration of the tissue sections in xylene followed by ethanol, sections were incubated with 0.5% hydrogen peroxide in phosphate-buffered saline for 30 minutes. Primary anti-human (rabbit/mouse/goat) polyclonal antibodies were applied to each microsection using the triple sandwich indirect immunohistochemical methods20). The usage of biopsy specimens filed in the Department of Oral Pathology, GWNUDH was approved by our institutional review board (IRB2016-011).
In the microscope observation the abnormal teeth were confirmed as complex odontoma due to their abortive tooth structure. Even though there appeared conspicuous cyst wall lined by thin odontogenic epithelium, the lesion also showed the solid tumor epithelium gradually transforming into ghost cells. The abortive teeth were composed of enamel matrix and dentin surrounded by fibrous tissue, and the enamel matrix showed irregular incremental lines (Fig. 2 A1-A3). With the histological observation of tumorous odontogenic epithelium including many ghost cells, which were closely associated with abortive teeth, the lesion was finally diagnosed as CCOT associated with complex odontoma.
High magnification view showed abundant formation of ghost cells, admixed with spindle-shaped epithelial cells. The ghost cells were swollen with spotty eosinophilic materials in their cytoplasms, and their nuclei were gradually disappeared (Fig. 2 A4-A6). Some ghost cells became calcified and stained with hematoxylin. The dystrophic calcification began from the periphery of ghost cell clusters, where was demarcated by strong hematoxylin staining (Fig. 2 B1-B3). The cyst wall was partly thickened and composed of myxoid connective tissue (Fig. 2 C1). In high magnification there appeared many calcifying granules in the fibrous connective tissue. Some calcifying granules still exhibited weak image of nuclei in the center, therefore, it could be presumed that the calcifying granules were originated from the odontogenic mesenchymal cells which were prematurely calcified (Fig. 2 C2 and C3).
In the immunohistochemistry β-catenin staining was strong in the tumor epithelium transforming into ghost cells (Fig. 3 A1-A3). Some ghost cells were conspicuously positive for GADD45 (Fig. 3B1-B3). The tumor epithelium was strongly positive for LC3 (Fig. 3 C1-C3). Especially the strong LC3 staining was found in the spotty materials in the cytoplasms of ghost cells (Fig. 3 C4). The calcifying granules (arrows) in the connective tissue were also strongly positive for LC3 (Fig. 3 C5 and C6). And the ghost cells were slightly positive for MMP-9 (Fig. 3 D). Otherwise, the immunostainings of HSP-70 (Fig. 3 E), p38 (Fig. 3 F), BCL2 (Fig. 3 G), and Wnt1 (Fig. 3 H) were rarely positive.
Ⅲ.Discussion
Although WHO classification has been known since 2005, many clinical cases still favored to diagnose the calcifying cystic lesion as COC rather than CCOT due to its benign nature with rare ghost cell formation2,3,8,21,22). However, some calcifying cystic lesions producing tumorous epithelium with frequent ghost cell formation were unanimously agreed to be diagnosed as CCOT due to their aggressive growth and increased recurrence.
The present cystic lesion in anterior maxilla contained a solid tumor tissue as well as an epithelial-lined cystic space. The tumorous tissue was closely associated with complex odontoma, and showed multiple clusters of spindle to polygonal epithelial cells transforming into ghost cells. And like the prerious report17)there also found rare expression of proliferation biomarkers, i.e., PCNA, Ki-67, and p53. Thus, it was presumed that the present case was slowly growing benign tumor with cystic expansion and many calcification foci. With the above histological and immunohistochemical observation the present case was finally diagnosed as CCOT differentially from COC and AOT.
It was reported that in COCs and ameloblastomas, the expression of MMP-9 mRNA and protein was detected in tumor cells as well as in stromal cells15). NF-kappaB may minimally affect the progression and local invasiveness of CCOT, DGCT and GCOC. GCOC show significantly higher proliferative activity than CCOT and DGCT12). And the TGF- β/SMAD signaling pathway regulates diverse cellular functions, including tooth development, and is involved in numerous pathological processes such as tumorigenesis. TGF-β/SMAD signaling pathway is activated in ameloblastoma, AOT, and CCOT. The statistically significant reduced TGF-β 1/SMAD immunoexpression in ameloblastoma compared to AOT/CCOT could be associated with the more aggressive biological behavior of ameloblastoma including increased cell proliferation and reduced apoptosis and differentiation23).
MMPs are responsible for extracellular matrix remodeling and, together their inhibitors and inducer, determinate the level of its turnover in pathological processes, leading to an auspicious microenvironment for tumor development. It was also reported that in the epithelial layer of CCOT, immunostaining for MMPs, tissue inhibitor of matalloproteinases (TIMPs), reversion-inducing-cystein-rich protein with kazal motifs (RECK) and extracellular matrix metalloproteinase inducer (EMMPRIN) was found in basal, suprabasal spindle and stellate cells surrounding ghost cells and ghost cells themselves, except for MMP-9 and TIMP-2 which were only expressed by ghost cells24). The MMP-9 expression in stroma is associated with invasive ability of the CCOT, DGCT and GCOC12). Ribeiro et al found that for MMP-2, there was predominant in stromal tissues, whereas for MMP-26 immunostaining was varied25). The presence of these metalloproteinases in stromal cells reveals the active participation of these cells in the degradation of the extracellular matrix, contributing to the growth of the tumor studied.
The present CCOT showed strong positive reaction of β -catenin in the tumor epithelium including ghost cells, but it was rarely positive for Wnt1. And the β-catenin immunostain was not found in the nuclei of tumor epithelial cells. Therefore, it was presumed that the overexpression of β-catenin was derived from the altered differentiation of epithelial cells rather than the oncogenic signaling of Wnt1 activation or β-catenin mutation.
The tumor epithelium of CCOT was rarely positive for BCL2, HSP-70, and p38, which were essential biomarkers of anti-apoptosis, cellular protection, and cellular stress, respectively. Whereas GADD45 and LC3 were strongly positive in the epithelium of CCOT, particularly condensed in ghost cells. As GADD45 and LC3 play important role of cell growth arrest/DNA damage signaling and autophage activation, respectively, it was presumed that the ghost cells of CCOT might undergo dormant cell state through altered cytodifferentiation stimulated by severe growth arrest, DNA damage signaling, and abundant autophage formation. On the other hand, it was also supposed that the ghost cells of CCOT would be relatively free from anti-apoptotic cascade and cellular stress due to their sparse expression of BCL2, HSP-70, and p38.
