Introduction
Calcifying epithelial odontogenic tumour (CEOT), also known as Pindborg tumour, is a rare benign odontogenic tumour that is characterized by the presence of amyloid material that may become calcified [
1,
2]. It is uncommon, accounting for only approximately 1% of all odontogenic tumours. Histologically, the classic form shows sheets and strands of variable size, polyhedral epithelial cells with well-defined borders, and distinct intercellular bridges. Nuclear pleomorphism and hyperchromatism are common. Round eosinophilic homogeneous deposits or amyloid substances are frequently found and often calcified. A pattern of concentric calcification called Liesegang rings is common, but it is not always present [
3].
Interestingly, despite the small number of documented cases of CEOT, several variants of this tumour have been reported, including the presence of clear cells [
3‐
5], Langerhans cells [
6,
7], myoepithelial cells [
3], bone, and cementum-like components [
8]. There have also been reports of CEOT combined with adenomatoid odontogenic tumours (AOTs) [
3] or ameloblastoma [
9].
Asano et al. [
6] and Takata et al. [
7] described the Langerhans cell variant of the intraosseous CEOT in 2 Japanese patients. In both cases, the tumour chiefly consisted of scattered small islands of epithelial cells. In some nests, a few clear cells tested positive for S-100 protein, lysozome, MT 1, LN-3, and OKT 6 antibodies but not for keratin antibody. Almost no calcification of homogenous eosinophilic materials was observed. Ultrastructurally, the S-100 positive cells were identified as Langerhans cells based on the finding of rod- and racket-shaped Birbeck’s granules. Three more Asian cases were also reported later [
10,
11].
Due to the limited cases, the biological behaviour and histogenesis of the Langerhans cell variant of CEOT are not yet fully understood, and the relationship between normal CEOT and the Langerhans cell variant remains to be clarified. Here, we report 8 cases of CEOT, including 2 cases of the Langerhans cell variant, from our own file and review the literature with special reference to the Langerhans cell variant.
Material and methods
Cases diagnosed as CEOT or cysts with CEOT-like hyperplasia from January 1985 to December 2010 were retrieved from the files of the Department of Oral Pathology, Peking University School of Stomatology. The study protocol was approved by the Ethical Committee for Human Experiments of Peking University School of Stomatology (IRB00001052-11041). Standard hematoxylin and eosin-stained sections were reviewed and the lesions were reclassified according to the World Health Organization Classification of Odontogenic Tumours [
12].
Eight cases of CEOT were confirmed: 6 cases had typical histopathological features similar to the description of CEOT, while the other 2 cases showed unusual histopathological features similar to the description of the Langerhans cell variant of CEOT in the literature. After confirmation by immunohistochemical staining, the 2 cases were classified as the Langerhans cell variant of CEOT. Clinical data including age, sex, anatomic site, duration, radiographic features, clinical impression, treatment, and available follow-up data were recorded. Cases of the Langerhans cell variant of CEOT reported in the English language literature were reviewed.
Two cases of the Langerhans cell variant were analysed by immunohistochemistry. Four-μm-thick tissue sections were deparaffinised and rehydrated. Endogenous peroxidase activity was blocked in 3% H2O2. For antibodies against human S-100, CD 1a, and langerin, antigen retrieval was achieved by treating the sections with a boiling solution of tris-ethylenediaminetetraacetic acid (pH 8.0) for 10 min in a microwave. Except for the anti-langerin antibody (Maixin; Fuzhou, Fujian, China), the primary antibodies were purchased from Zhongshan Biotechical Company (Beijing, China). The colour was developed in freshly made diaminobenzidine. The sections were washed briefly in running tap water and lightly stained with Mayer’s hematoxylin. Negative controls were prepared by omission of the relevant primary antibody.
To compare the distribution of Langerhans cells in the 2 cases of the Langerhans cell variant of CEOT, 2 cases of typical CEOTs (Table
1, cases 1 and 4) were also analysed by immunohistochemistry. Sections of a case of Langerhans cell histiocytosis were used as positive control. The numbers of langerin-positive cells and epithelial tumour cells from 5 randomly selected areas of each section under high power magnification (× 400) in each case were scored to obtain the ratio of Langerhans cells to epithelial tumour cells.
Table 1
The clinical information of conventional CEOT
1 | 22/M | 10 | Max, R6 to tubera | Swelling | Swelling, no loose teeth | Yes | Unilobular, radiolucent and radiopaque | No | Enucleation, 7 year after surgery, NER |
2 | 20/M | 12 | Mand, L6-8 | Swelling | Swelling, no loose teeth | No | Unilobular, radiolucent | No | Enucleation, 7 year after surgery, NER |
3 | 39/M | 2 | Mand, L3-5 | Swelling | Swelling, loose teeth | No | Multilobular, radiolucent | No, displace of 4,5 | Enucleation, lost to follow-up |
4 | 37/F | 0 | Mand, L6 to retramolar region | No symptom | Swelling, no loose teeth | No | Unilobular, radiolucent and radiopaque | No | Segmental resection, 4 year after surgery, NER |
5 | 41/F | 3 | Mand, L4-6 | Swelling | Swelling, no loose teeth | Yes | Unilocular, radiolucent | No | Curretage, 2 year after surgery, NER |
6 | 12/F | 1 | Mand, L4-R5 | Swelling | Swelling, no loose teeth | Yes | Unilocular, radiolucent and radiopaque | No | Currettage, no follow-up information |
Discussion
CEOT is an uncommon odontogenic tumour. Upon reviewing our own file of odontogenic tumours from 1985 to 2006, we found a total of 6 cases of odontogenic tumours among 1,309 cases (0.46%) [
13]. The male:female ratio of patients with CEOT was 1:1. The tumours were primarily in the mandible (maxilla:mandible ratio, 1:2), and approximately 82% of the cases were located in the premolar and molar regions [
3]. The clinical features of the present series were in general agreement with those of earlier reports, i.e. equal sex distribution and predilection for the mandible, especially the premolar and molar regions.
When first described [
1], CEOT was considered a locally invasive tumour, and some subsequent publications supported this concept. This view was based upon evidence suggestive of bone marrow involvement from radiographs and histological sections [
2,
14]. However, Vap et al [
14] believed that the tumour is an expansile lesion that does not extend into the intertrabecular spaces as does the infiltrative solid ameloblastoma. Franklin and Pindborg [
15] reported a recurrence rate of 14% that was primarily attributed to inadequate treatment. A few cases of malignant transformation have been reported [
16,
17]. It is now accepted that CEOT is a less aggressive tumour than ameloblastoma that responds well to conservative surgery. Four cases of CEOT with follow-up data in the present series showed no signs of recurrence following surgery, which is in keeping with the notion.
As the first few reported cases of CEOT were all associated with an unerupted tooth, Pindborg [
2] initially believed that CEOT develops from the reduced enamel epithelium of an unerupted tooth. However, as reports of cases without an associated unerupted tooth and cases of a peripheral variant started to accumulate, it became evident that sources other than reduced enamel epithelium should also be considered regarding CEOT histogenesis.
In the present series, 3 of 6 cases were associated with an impacted tooth. This is consistent with the literature that approximately half of CEOT cases are associated with an unerupted tooth. A few reports of a dentigerous cyst or dental follicle showed focal CEOT-like hyperplasia [
18,
19], which suggests that CEOT associated with the crowns of impacted teeth may indeed originate from the reduced enamel epithelium.
Our review of the English literature revealed 5 cases, together with our 2 new cases, for a total of 7 cases of the Langerhans cell variant of CEOT (Table
2). All of the patients were Asian. The patient age range at first presentation was 38–58 years (median, 44 years). Three patients were male and 4 were female. Six tumours occurred in the maxilla, while 1 occurred in the mandible. Five of the 6 maxillary tumours involved the anterior region. This feature is in contrast to the findings of conventional CEOT. Jaw swelling was noted in 4 cases. Two patients complained of loosened teeth and 1 patient showed no obvious symptoms. During an intraoral examination, bony depression of the tumour area was noted in 3 cases, a unique feature not seen in conventional CEOT.
Table 2
The clinical information of reported cases of Langerhans cell variant of CEOT
Present series | 58/M | 3 | Max, R molar to L3 | Swelling | Swelling, loose teeth | No | Mutilobular radiolucent | R3-6 | Partial resection of the maxilla, 10 years NER |
Present series | 40/F | 48 | Max, R incisor to L premolar | Loose teeth, pain | Depression over the anterior maxilla, loss of alveolar bone | No | Unilobular radiolucent | L1-2 | Currettage, 5 years NER |
Asano M (1990) | 44/F | Several years | Max, L1-R6 | Swelling | Swelling | No | Unilobular, radiolucent | R1-3 | Partial resection of the maxilla, no follow-up information |
Takata T (1993) | 58/M | 6 | Max, L3-5 | Loose teeth, loss of palatal and buccal alveolar bone | No swelling, loss of bone, loose teeth | No | Unilobular, radiolucent | L3-5 | Enucleated,10 years NER |
Li Wang (2006) | 38/M | N.S. | Mand, R4 to ramus | Swelling, pain | N.S. | N.s | Well-difined radiolucencent | N.S. | Partial resection of the mandible, 2.5 year NER |
| 39/F | 24 | Max, L premolar, gingiva | Swelling in gingiva | Swelling in gingiva | N.s | N.S. | N.S. | Resected, 2 year NER |
Wang YP (2007) | 52/F | 0 | Max, R anterior region | No symptom | Depression over the anterior maxilla | No | Unilobular, radiolucent | No | Partial maxillectomy 23-16, No follow-up information |
Of the 6 cases with radiographic records, well-defined unilocular radiolucency was noted in 5 cases and multilocular radiolucency was noted in one. The radiopaque area was not evident in any cases. Root resorption was detected in 4 cases. Three patients were initially treated by enucleation or curettage and the other 4 cases underwent partial jaw resection. Five patients were followed for 2–10 years and no recurrence was recorded. Thus, the Langerhans cell variant of CEOT may show similar behaviours to those of conventional CEOT.
Histologically, the Langerhans cell variant of CEOT is unusual. Although Congo red-positive homogenous amyloid material could be seen, there was no calcification. The tumour islands were composed of small numbers of epithelial cells with clear cells suggestive of Langerhans cells. Langerin is the protein expressed in Langerhans cells that is localized in the Birbeck granules present within the cytoplasm. Langerin staining, accompanied by S100 and CD1a staining, confirmed the presence of Langerhans cells in the tumour.
With regard to histogenesis, none of the 7 cases summarized here were associated with an unerupted tooth; thus, the Langerhans cell variant of CEOT is less likely to arise from a reduced enamel epithelium. The tumour was mainly located within the jaw bone with apparent root resorption of the apical part of the involved tooth. Other tissue origins like the rests of Malassez should be considered.
Another interesting finding of the present study was that Langerhans cells were detected not only in the Langerhans cell variant of CEOT but also in the epithelial nests of the conventional type. Similar reports have been presented by other groups [
20]. Langerhans cells are bone marrow-derived cells that migrate into the oral epithelium and serve as antigen-presenting cells. Because both oral and odontogenic epithelia originate from the same oral ectoderm, it is possible that Langerhans cells may also migrate into the epithelial nests of odontogenic tumours. Thus, Langerhans cells may not exclusively exist in the Langerhans cell variant of CEOT. However, the number of Langerhans cell increased dramatically in the Langerhans cell variant of CEOT. The increased number of Langerhans cells may be associated with inflammation. Interestingly, 6 of the 7 cases of Langerhans cell variant here had a moderate number of lymphocytes and plasma cell infiltration. Asono [
6] postulated that Langerhans cells may play an important role in antigen presentation or regression of CEOT.
Although Langerhans cells increases are due to immunogenic reactions and do not exist exclusively in this kind of CEOT, we continue to use the term “Langerhans cell variant of CEOT” to describe it since a limited number of cases have been reported. At this point, the use of another name may not be helpful for accumulating data about this kind of CEOT with special clinical and histopathological characteristics. A more suitable name may be adopted after more cases have been reported.
Differentiating conventional CEOT from the Langerhans cell variant is not difficult. The conventional type affects the premola and molar region as an asymptomatic slow-growing expansile mass and may feature an unerupted tooth. In contrast, lesions of the Langerhans cell variant have a predisposition for the anterior maxilla and a possibly overlying bone depression. Histopathologically, in the conventional type, the tumour consists of islands and sheets of polyhedral epithelial cells, eosinophilic homogenous amyloid substance, and calcified tissue. In the Langerhans cell variant, very small islands and cords of neoplastic cells with abundant amyloid substance and no calcification are characteristic features.
In conclusion, the Langerhans cell variant of CEOT is rare but consistent. To date, it has only been reported in Asian people, mainly in the anterior portion of the maxilla. This jaw tumour could present features of bone depression with loss of teeth and alveolar bone. The lesion comprises small epithelial islands and amorphous eosinophilic globules of amyloid materials with no apparent calcification. These features are somewhat different from those of conventional CEOT and should alert the pathologists to the differential diagnosis.
Competing interest
The authors declare that they have no competing interests.
Authors’ contributions
Study concepts and study design: TJL, Data acquisition: YC, TTW, Quality control of data and algorithms: TJL, Data analysis and interpretation: YC, TTW, Manuscript preparation: YC, TTW, Manuscript editing: TJL, Manuscript review: TJL. YC and TTW: co-first author, TJL: corresponding author. All authors read and approved the final manuscript.