- Email: [email protected]
Is maspin immunolocalization a tool to differentiate central low-grade mucoepidermoid carcinoma from glandular odontogenic cyst?
ARTICLE IN PRESS Acta histochemica 112 (2010) 161—168
www.elsevier.de/acthis
Is maspin immunolocalization a tool to differentiate central low-grade mucoepidermoid carcinoma from glandular odontogenic cyst? Marilena Vered, Irit Allon, Amos Buchner, Dan Dayan Department of Oral Pathology and Oral Medicine, School of Dental Medicine, Tel Aviv University, Tel Aviv 69978, Israel Received 7 August 2008; received in revised form 13 October 2008; accepted 27 October 2008
KEYWORDS Maspin; Low-grade mucoepidermoid carcinoma; Glandular odontogenic cyst; Human
Summary Mucoepidermoid carcinoma (MEC) of the salivary glands has a low-grade variant (LGMEC), which may be found within the jawbones. LGMEC shares a number of histopathological similarities with glandular odontogenic cysts (GOC) of the jawbones. Maspin has been identified in several benign and malignant salivary gland neoplasms. We investigated the immunolocalization of maspin in LGMEC and GOC and evaluated its potential to distinguish between these two entities. Cases of LGMEC (n ¼ 6), GOC (n ¼ 8) and various odontogenic cysts with marked mucous metaplasia (OCMM, n ¼ 7), which served as controls, were immunohistochemically labeled for the binding of an antibody directed against maspin. Immunomorphometry was performed separately for maspin-immunopositive epithelial cells and epithelial-mucous cells in either their nuclear or cytoplasmic compartments. Results were presented as the volume fraction (Vv) of each element. The Vv of the maspinimmunopositive epithelial-mucous cytoplasm and nuclei was significantly higher in LGMEC than in GOC and OCMM (po0.001 and p ¼ 0.026, respectively). In the epithelial cells, no significant differences were observed among the lesions (p40.05). It is suggested that the high levels of maspin in the epithelial-mucous cells (in both cytoplasm and nuclei) in LGMEC may serve as a tool to distinguish it from GOC. This may be useful especially in equivocal cases and in small incisional biopsy samples. & 2008 Elsevier GmbH. All rights reserved.
Corresponding author. Tel.: +972 3640 9305; fax: +972 3640 9250.
E-mail addresses: [email protected] (M. Vered), [email protected] (I. Allon), [email protected] (A. Buchner), [email protected] (D. Dayan). 0065-1281/$ - see front matter & 2008 Elsevier GmbH. All rights reserved. doi:10.1016/j.acthis.2008.10.006
ARTICLE IN PRESS 162
Introduction The odontogenic epithelium and the salivary gland epithelia share a common origin, namely, the epithelium of the stomatodeum (Ten Cate, 1994). Both begin their embryonic development as a thickening within the epithelial lining of the stomatodeum during weeks 5–6 of gestation. Although each of these epithelial types develops into completely different and highly specific functioning organs, the neoplastic cells within the various pathologic lesions derived from both the odontogenic and salivary gland epithelia may show common cytomorphological features, such as basaloid, columnar-to-cuboidal cells with a palisading appearance, clear cells and epithelial-mucous secreting cells. A rare type of developmental odontogenic cyst, glandular odontogenic cyst (GOC), may demonstrate an aggressive biological behavior (Waldron, 2002). It is generally accepted that such a cyst has an odontogenic origin; however, it shows gland-like features reminiscent of salivary gland tumors, thus reflecting the common origin of the odontogenic and salivary gland epithelia (Ellis and Auclair, 1996; Waldron, 2002). On rare occasions, salivary gland tumors arise centrally within the jawbones. Their origin is attributed to neoplastic transformation of odontogenic epithelium, mainly dentigerous cysts (Ellis and Auclair, 1996), or to salivary gland tissue entrapped within the jawbones dating from the embryonic development (Bouquot et al., 2000). The most common central salivary gland tumor is mucoepidermoid carcinoma (MEC) of the low-grade type (LGMEC), a malignant neoplasm, with low metastatic potential (Ellis and Auclair, 1996). Central LGMEC of the jawbones is indistinguishable in both architecture and immunohistochemical phenotype of the cells from its salivary gland counterpart (Pires et al., 2004), and has histological features that overlap with those of GOC (Ellis and Auclair, 1996; Waldron, 2002). Differentiation of LGMEC from GOC bears both treatment and prognostic implications. Mammary serine protease inhibitor (maspin) was originally described in breast myoepithelium by Zou et al. (1994), and has since been detected in various normal glandular tissues, such as the prostate, pancreas, ovary and salivary glands, as well as in several benign and malignant types of epithelial neoplasms (Futscher et al., 2002). The tumor suppressor qualities of maspin have been reported in breast (Zou et al., 1994), prostate (Pierson et al., 2002) and oral cavity carcinoma (Xia et al., 2000; Vered et al., 2008). The inhibitory role
M. Vered et al. of maspin has been described in different stages of tumor formation and progression, including degradation of basal laminae following serine protease activities, connective tissue invasion, tumorinduced angiogenesis and development of metastases (Solomon et al., 2006). Maspin is a useful prognosticator of tumor progression and patient survival, e.g., in oral and breast cancer (Sabbatini et al., 2000; Xia et al., 2000). Its presence is related to sensitization of tumor cells to apoptosis (Lockett et al., 2006). The purpose of this study was to examine the immunohistochemical localisation of maspin in LGMEC, in comparison with GOC, and to assess its use as a marker to distinguish between these entities.
Materials and methods Study cases Cases of GOC (n ¼ 8), LGMEC (n ¼ 6) and various odontogenic cysts (radicular and dentigerous) with mucous metaplasia (OCMM, n ¼ 7), which served as a control group, were retrieved from the archival files of the Department of Oral Pathology and Oral Medicine, School of Dental Medicine, Tel-Aviv University. All samples were formalin-fixed and paraffin wax-embedded. Five micrometer-thick sections were cut and stained with hematoxylin and eosin, and mucicarmine (Mayer’s Mucicarmine, BioOptica, Milano, Italy) using conventional protocols.
Immunohistochemistry Three micrometer-thick sections were cut and mounted on positively charged microscope slides (OptiplusTM, Biogenex, San Ramon, CA, USA). These were then labelled immunohistochemically for binding of an anti-maspin antibody. After dewaxing and rehydrating by conventional methods, antigen retrieval was performed by incubating with NuclearDecloakerTM, pH ¼ 9.5 (Biocare Medical, Walnut Creek, CA, USA), in a pressure cooker for 12 min. Sections were then incubated with a mouse monoclonal primary antibody directed against maspin (EAW24, diluted 1:75, incubated overnight at 4 1C; Novocastra, Newcastle upon Tyne, UK) and binding was revealed using a Broad Spectrum Poly HRP conjugate ready-to-use kit (Zymed, San Francisco, CA, USA), employed according to manufacturer’s instructions. Sections were reacted with AEC substrate-chromagen kit (Zymed, San Francisco,
ARTICLE IN PRESS Maspin in central mucoepidermoid carcinoma and glandular odontogenic cyst CA, USA) employed according to manufacturer’s instructions, counterstained in Mayer’s hematoxylin (Pioneer Research Chemicals, Colchester, UK) and mounted in GVA mounting medium (Zymed, San Francisco, CA, USA).
Histomorphometry Two independent investigators (MV and IA) conducted the assessment. Immunomorphometry was first performed on GOC. Epithelial cells and epithelial-mucous cells were counted separately. Both cytoplasmic and nuclear immunolabelling were taken into consideration. The extent of maspin immunolabelling was assessed using a modified point-counting method (Dayan et al., 2000). At 400 magnification, a standard field defined by a 100-square grid (Olympus, Tokyo, Japan), comprising 121 intersections between each horizontal and vertical line was mounted on the microscope and placed over the cystic epithelium with one border tangentially touching the epithelial–connective tissue interface. In each examined field, intersections positioned beyond the cystic epithelium (i.e., cyst lumen) were deleted from the total number of intersections per field (i.e., 121). Point-counting was conducted on the following maspin-immunopositive elements: epithelial cell cytoplasm (EC) and nucleus (EN); epithelialmucous cell cytoplasm (EMC) and nucleus (EMN). When an intersection hit one of these elements, the specified element scored one point. A minimum of 300 intersections was counted in each section. Then, the sum of points overlying each specified element was calculated, divided by the total number of counted intersections in that section and multiplied by 100. This was defined as the volume fraction (Vv) of each maspin-immunopositive element. The procedure was applied to all sections in the GOC group to allow calculation of the mean Vv of each examined element in this group. The same approach was used to calculate the mean Vv of the examined elements in LGMEC and OCMM groups.
163
Results Histology Two oral pathologists (MV and DD) confirmed the original diagnoses by examining hematoxylin and eosin-stained sections from the cases. GOC cases selected for the study fulfilled the histopathological diagnostic criteria established by the World Health Organization (WHO), including cysts lined by non-keratinized stratified epithelium (partly squamous), which has a surface layer of acidophilic cuboidal or columnar cells, often forming irregular papillary projections (Kaplan et al., 2005). The rest of the cyst lining contains epithelial cells characterized by an eosinophilic cytoplasm and large, usually centrally located, nuclei and distinct polygonal cell outlines. Occasionally, the luminal surface of the epithelial cells exhibit cilia. Among the epithelial cells, varying numbers of epithelial–mucous cells are interposed. These are identified by their abundant, foamy cytoplasm and small, eccentric rather pyknotic nuclei, large cell size and round, cuboidal, ovoid, columnar or goblet shapes. Within the epithelial layer, there may be crypts/cyst-like/glandular-like spaces lined by cells, similar to those at the epithelial surface. Mucin may be demonstrable within the intraepithelial spaces and within the main cyst cavity (Figure 1). Histologically, LGMEC shows a cystic architecture with epithelial lining that varies in thickness and that features a cellular composition
Statistical analysis Inter-observer variations were assessed by the non-parametric Wilcoxon test. Differences in the mean Vv of the various elements among study groups were statistically analyzed using one-way ANOVA. Statistical significance was set at po0.05. The Statistical Package for the Social Sciences (SPSS 14) software was used for computations.
Figure 1. Glandular odontogenic cyst (GOC) showing slightly thickened cyst lining composed of cuboidal-tocolumnar epithelial cells and mucus-secreting cells (finely granular, bluish cytoplasm), the latter also seen as part of the lining of the intraepithelial gland-like structures. Delicate cilia can be seen on the luminal surface of the epithelial cells (hematoxylin and eosin, Scale bar ¼ 200 mm).
ARTICLE IN PRESS 164
M. Vered et al.
Figure 2. Low-grade mucoepidermoid carcinoma (LGMEC) showing a fragment of a cystic structure lined by epithelial cells, among which mucus-producing cells are interposed. Intraepithelial gland-like structures are present. Cytologically, the neoplastic cells are benign without atypia or mitotic figures (hematoxylin and eosin, Scale bar ¼ 200 mm).
comprising of an admixture of epithelial/epidermoid and mucus-producing cells with occasional formation of gland-like intra-epithelial microcysts or papillary-cystic architecture (Figure 2). Often, the architectural and cellular composition in GOC and LGMEC are remarkably similar (Figure 3). Based on the histological similarities and common immunohistochemical profiles of central MEC and salivary gland MEC (Pires et al., 2004), the LGMEC cases selected in this study were of minor salivary gland origin and fulfilled the histopathological criteria established for the low-grade type of MEC (Goode and El-Naggar, 2005). The OCMM consisted of developmental (e.g., dentigerous) and inflammatory (e.g., radicular) cysts that demonstrated prominent mucous metaplasia of the lining epithelium.
Immunohistochemistry and mucicarmine staining In all cases of GOC, histochemical staining with mucicarmine revealed positively stained epithelial– mucous cells and mucous-containing pools within the lining epithelium (Figure 4a). The epithelial cells did not stain with mucicarmine. Positive immunohistochemical labelling of maspin was found within the cyst lining in all cases. Immunopositivity in the epithelial cells was usually diffuse in both the cytoplasm and nuclei, and absent in the epithelial–mucous cells (Figure 4b). In all LGMEC cases, similar to GOC, mucicarmine
Figure 3. GOC versus LGMEC Architectural overlapping between the lesions includes a similar thickness of the lining epithelium and arrangement of the epithelial and epithelial–mucous cells (hematoxylin and eosin, Scale bar ¼ 200 mm).
staining was found in the epithelial–mucous cells and mucous-containing pools (Figure 5a). However, unlike in GOC, numerous epithelial–mucous cells in LGMEC demonstrated maspin immunopositivity in both the cytoplasm and nuclear compartments (Figure 5b). Most of the epithelial cells in LGMEC were maspin-immunopositive in both cytoplasmic and nuclear compartments. All OCMM cases demonstrated maspin immunopositivity in a pattern similar to that of GOC.
Immunomorphometry Inter-observer variations were statistically insignificant (p40.05). Results are summarized in Table 1. The present quantitative measurements further support the immunomorphological differences found between LGMEC and GOC, as previously described. The mean Vv of EMC was significantly
ARTICLE IN PRESS Maspin in central mucoepidermoid carcinoma and glandular odontogenic cyst
165
Figure 4. Glandular odontogenic cyst (GOC) (a) showing extensive mucicarmine staining (purple) of the epithelial-mucous cells (Scale bar ¼ 300 mm). (b) The same area as Figure 1a, showing positive maspin immunolabelling in the epithelial cells, but not in the epithelial–mucous cells (Scale bar ¼ 300 mm).
Figure 5. Low-grade mucoepidermoid carcinoma (LGMEC) (a) showing a fragment of a cystic structure lined by epithelial and epithelial–mucous cells; the latter are stained with mucicarmine (Scale bar ¼ 300 mm). (b) The same area as illustrated in Figure 2a, showing maspin-immunopositive cytoplasm and nuclei of the epithelial–mucous cells (Scale bar ¼ 300 mm).
higher in LGMEC (16.5%710.8%) than in GOC and OCMM (1.572% and 171%, respectively), (po0.001). The mean Vv of EMN in LGMEC (1.7%, 41.5%) was significantly higher compared to GOC and OCMM (0.370.5% and 0.470.7%, respectively) (p ¼ 0.026). No significant differences were found in the mean Vv of the maspin-immunopositive epithelial cells (in both cytoplasmic and nuclear compartments) among the different lesions (p40.05). No significant differences were found in the mean Vv of the epithelial and epithelial– mucous cells between GOC and OCMM (p40.05).
genic lesions (i.e., lateral periodontal cyst/ botryoid cyst), and between GOC and central LGMEC (Sciubba et al., 2001), has prompted many researchers to examine a wide panel of markers that could possibly aid in differentiating between these lesions. This is important because of the different treatment needs for patients with a developmental odontogenic cyst compared to lowgrade carcinoma. Since GOC is a relatively rare lesion of the jawbones, only small series and sporadic cases are reported in the current literature. From the available data, the cytokeratin profile, cell cycle proteins (i.e., p53) and proliferation markers (i.e., Ki-67 and PCNA) in GOC odontogenic lesions (cysts and tumors) and LGMEC (central and salivary gland origin) show overlapping features (Pires et al., 2004; Kaplan et al., 2005; Shen et al., 2006). The subtle differences in these studies might reflect variations in epithelial
Discussion The overlap in the histopathological features between GOC and other developmental odonto-
ARTICLE IN PRESS 166
M. Vered et al.
Table 1. Mean Vv (7SD) of the maspin-immunopositive components in the various types of examined lesions.
EMC EMN EC EN
LGMEC
GOC
OCMM
p value
16.5710.8 1.771.5 13.3712.6 8.9710.5
1.572 0.370.5 26.4719.9 12.375.5
171.1 0.771.1 25.4714.4 15.876.2
po0.001 p ¼ 0.026 NS NS
LGMEC – low-grade mucoepidermoid carcinoma; GOC – glandular odontogenic cyst; OCMM – odontogenic cyst with mucous metaplasia. EMC – epithelial–mucous cell cytoplasm; EMN – epithelial–mucous cell nuclei; EC – epithelial cell cytoplasm; EN – epithelial cell nuclei. NS – not significant.
maturation. No parameter demonstrated a consistent pattern that supports a definite separation between odontogenic lesions, GOC, and LGMEC. The results of our study clearly showed immunomorphological differences regarding maspin between LGMEC and GOC. These differences were further supported by quantitative measurements. While the epithelial–mucous cells in LGMEC showed quite extensive maspin immunolabelling, in GOC, maspin immunoreactivity was restricted to only a small fraction of these cells. The pattern of maspin immunolabelling in the epithelial and epithelial– mucous cells in OCMM was similar to that in GOC. These findings may be of substantial aid in the diagnosis of cases that present as large cystic lesions within the jawbones, where an incisional biopsy is performed. In cases where diagnosis of GOC or central LGMEC cannot be made based on the morphological features alone, especially when the biopsy sample is small, extensive maspin immunolabelling in the epithelial–mucous cells (in both cytoplasm and nuclei) may be used as a tool to favor a diagnosis of LGMEC rather than GOC. The present results should be further validated in studies comprising larger series of lesions. Several types of benign and malignant salivary gland tumors are maspin immunopositive (de Navarro et al., 2004; Nakashima et al., 2006); however, in respect to MEC, the present results were the first to show that the low-grade variant of this tumor was maspin immunopositive. The extent of maspin immunolabelling in LGMEC in the epithelial–mucous cells can be explained in terms of signals that are assumed to induce maspin expression in association with a more differentiated cell phenotype (Lockett et al., 2006). The epithelial–mucous cells in LGMEC represent a higher state of differentiation compared to other cell types often present in this tumor; the epithelial/
epidermoid and intermediate cells. The present results are in accordance with a recently published study that found that intermediate grade MECs (characterized by a smaller number of epithelial– mucous cells), together with other malignant salivary gland tumors of a similar grade, demonstrate loss of maspin expression (Schwarz et al., 2008). In regard to the nuclear immunolocalisation of maspin in GOC, we assumed that this was part of the pattern of immunolabelling in states of stress associated with the inflammatory infiltrate, which was usually found at varying degrees of severity in the vicinity of the cystic epithelium. In general, in stress-related conditions such as inflammation, maspin immunolabelling was observed at both cytoplasmic and nuclear compartments (Cao et al., 2005; Lockett et al., 2006). It is suggested that the significantly higher nuclear immunolabelling of maspin in LGMEC compared to GOC, reflects the fact that the neoplastic transformation is considered as the highest state of cellular stress. Nevertheless, the relatively high nuclear immunopositivity of maspin in LGMEC was compatible with less aggressive types of malignancies and favorable clinical outcomes (Mohsin et al., 2003; Dietmaier et al., 2006; Marioni et al., 2008). The general picture emerging from this study and the literature indicate a histological continuum of mucous cell containing-lesions starting from inflammatory and developmental odontogenic cysts with mucous cell metaplasia (Takeda et al., 2005), to inflammatory and developmental odontogenic cysts with focal GOC-like areas, to GOC developing within odontogenic cysts (Takeda et al., 2005), GOC, central LGMEC developing within odontogenic cysts (Tran et al., 2004; Holsinger et al., 2002; Sittitavornwong et al., 2006) and finally to central LGMEC. These lesions share some histopathological features, but have a wide range of biological behavior that may differ among lesions of the same type, and among different types of lesions. This may reflect the pluripotential character of the odontogenic epithelium, which has the ability to develop diverse types of lesions. In light of this, no definite separation can be made between central LGMEC and GOC, but a more frequent immunopositivity of the epithelial–mucous cells favors a diagnosis of LGMEC.
Acknowledgements The study was supported by the Ed and Herb Stein Chair in Oral Pathology, Tel Aviv University, Tel
ARTICLE IN PRESS Maspin in central mucoepidermoid carcinoma and glandular odontogenic cyst Aviv, Israel. The study was presented at the 1st International Congress on Oral Cancer organized by the International Academy of Oral Oncology (IAOO); May 2007, Amsterdam, The Netherlands.
References Bouquot JE, Gnepp DR, Dardick I, Hietanen JH. Intraosseous salivary tissue: jawbones examples of choristomas, hamartomas, embryonic rests, and inflammatory entrapment: another histogenetic source for intraosseous adenocarcinoma. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2000;90:205–17. Cao D, Wilentz RE, Abbruzzese JL, Ho L, Maitra A. Aberrant expression of maspin in idiopathic inflammatory bowel disease is associated with disease activity and neoplastic transformation. Int J Gastrointest Cancer 2005;36:39–46. Dayan D, Vered M, Paz T, Buchner A. Aging of human palatal salivary glands: a histomorphometric study. Exp Gerontol 2000;35:85–93. de Navarro RL, Martins MT, de Araujo VC. Maspin expression in normal and neoplastic salivary gland. J Oral Pathol Med 2004;33:435–40. Dietmaier W, Bettstetter M, Wild PJ, Woenckhaus M, Ru ¨mmele P, Hartmann A, et al. Nuclear Maspin expression is associated with response to adjuvant 5-fluorouracil based chemotherapy in patients with stage III colon cancer. Int J Cancer 2006;118: 2247–54. Ellis GL, Auclair PL. Central mucoepidermoid carcinoma. In: Rosai J, editor. Atlas of tumor pathology. Third series, Fascicle 17. Tumors of the salivary glands. Washington: AFIP; 1996. p. 172–5. Futscher BW, Oshiro MM, Wozniak RJ, Holtan N, Hanigan CL, Duan H, et al. Role for DNA methylation in the control of cell type-specific maspin expression. Nat Genet 2002;31:175–9. Goode RK, El-Naggar AK. Mucoepidermoid carcinoma. In: Barnes L, Eveson JW, Reichart P, Sidransky D, editors. WHO classification of tumours. Pathology and genetics – Head and neck tumours. Lyon: IARC; 2005. p. 219–20. Holsinger FC, Owens JM, Raymond AK, Myers JN. Central mucoepidermoid carcinoma of the mandible: tumorigenesis within a keratocyst. Arch Otolaryngol Head Neck Surg 2002;128:718–20. Kaplan I, Anavi Y, Manor R, Sulkes J, Calderon S. The use of molecular markers as an aid in the diagnosis of glandular odontogenic cyst. Oral Oncol 2005;41: 895–902. Lockett J, Yin S, Li X, Meng Y, Sheng S. Tumor suppressive maspin and epithelial homeostasis. J Cell Biochem 2006;97:651–60. Marioni G, Staffieri C, Staffieri A, De Filippis C, Blandamura S. Maspin tumour-suppressing activity in head and neck squamous cell carcinoma: emerging evidence and therapeutic perspectives. Acta Otolaryngol 2008;10:1–5.
167
Mohsin SK, Zhang M, Clark GM, Craig Allred D. Maspin expression in invasive breast cancer: association with other prognostic factors. J Pathol 2003;199: 432–5. Nakashima D, Uzawa K, Kasamatsu A, Koike H, Endo Y, Saito K, et al. Protein expression profiling identifies maspin and stathmin as potential biomarkers of adenoid cystic carcinoma of the salivary glands. Int J Cancer 2006;118:704–13. Pierson CR, McGowen R, Grignon D, Sakr W, Dey J, Sheng S. Maspin is up-regulated in premalignant prostate epithelia. Prostate 2002;53:255–62. Pires FR, Chen SY, da Cruz Perez DE, de Almeida OP, Kowalski LP. Cytokeratin expression in central mucoepidermoid carcinoma and glandular odontogenic cyst. Oral Oncol 2004;40:545–51. Sabbatini R, Federico M, Morselli M, Depenni R, Cagossi K, Luppi M, et al. Detection of circulating tumor cells by reverse transcriptase polymerase chain reaction of maspin in patients with breast cancer undergoing conventional-dose chemotherapy. J Clin Oncol 2000; 18:1914–20. Schwarz S, Ettl T, Kleinasser N, Hartman A, Reichert T-E, Driemel O. Loss of maspin expression is a negative prognostic factor in common salivary gland tumors. Oral Oncol 2008;44:563–70. Sciubba JJ, Fantasia JE, Kahn LB. Glandular odontogenic cyst. In: Rosai J, editor. Atlas of tumor pathology. Third series, Fascicle 29. Tumors and cysts of the jaws. Washington: AFIP; 2001. p. 31–4. Shen J, Fan M, Chen X, Wang S, Wang L, Li Y. Glandular odontogenic cyst in China: report of 12 cases and immunohistochemical study. J Oral Pathol Med 2006;35:175–82. Sittitavornwong S, Koehler JR, Said-Al-Naief N. Glandular odontogenic cyst of the anterior maxilla: case report and review of the literature. J Oral Maxillofac Surg 2006;64:740–5. Solomon LA, Munkarah AR, Schimp VL, Arabi MH, Morris RT, Nassar H, et al. Maspin expression and localization impact on angiogenesis and prognosis in ovarian cancer. Gynecol Oncol 2006;101:385–9. Takeda Y, Oikawa Y, Furuya I, Satoh M, Yamamoto Y. Mucous and ciliated cell metaplasia in epithelial linings of odontogenic inflammatory and developmental cysts. J Oral Sci 2005;47:77–81. Ten Cate AR. Structure of the oral tissues. In: Ten Cate AR, editor. Oral histology. Developmental, structure and function. St. Louis: Mosby; 1994. p. 45–57. Tran P-T, Cunningham CJ, Baughman RA. Glandular odontogenic cyst. J Endod 2004;30:182–4. Vered M, Allon I, Dayan D. Maspin, p53, p63 and Ki-67 in epithelial lesions of the tongue: from hyperplasia through dysplasia to carcinoma. J Oral Pathol Med 2008 [in press]. Waldron CA. Odontogenic cysts and tumors. In: Neville BW, Damm DD, Allen CM, Bouquot JE, editors. Oral and maxillofacial pathology. Philadelphia, PA: WB Saunders; 2002. p. 607–8.
ARTICLE IN PRESS 168 Xia W, Lau Y-K, Hu M C-T, Li L, Johnston DA, Sheng S, et al. High tumoral maspin expression is associated with improved survival of patients with oral squamous cell carcinoma. Oncogene 2000;19:2390–403.
M. Vered et al. Zou Z, Anisowicz A, Hendrix MJ, Thor A, Neveu M, Sheng S, et al. Maspin, a serpin with tumor-suppressing activity in human mammary epithelial cells. Science 1994;263:526–9.
www.elsevier.de/acthis
Is maspin immunolocalization a tool to differentiate central low-grade mucoepidermoid carcinoma from glandular odontogenic cyst? Marilena Vered, Irit Allon, Amos Buchner, Dan Dayan Department of Oral Pathology and Oral Medicine, School of Dental Medicine, Tel Aviv University, Tel Aviv 69978, Israel Received 7 August 2008; received in revised form 13 October 2008; accepted 27 October 2008
KEYWORDS Maspin; Low-grade mucoepidermoid carcinoma; Glandular odontogenic cyst; Human
Summary Mucoepidermoid carcinoma (MEC) of the salivary glands has a low-grade variant (LGMEC), which may be found within the jawbones. LGMEC shares a number of histopathological similarities with glandular odontogenic cysts (GOC) of the jawbones. Maspin has been identified in several benign and malignant salivary gland neoplasms. We investigated the immunolocalization of maspin in LGMEC and GOC and evaluated its potential to distinguish between these two entities. Cases of LGMEC (n ¼ 6), GOC (n ¼ 8) and various odontogenic cysts with marked mucous metaplasia (OCMM, n ¼ 7), which served as controls, were immunohistochemically labeled for the binding of an antibody directed against maspin. Immunomorphometry was performed separately for maspin-immunopositive epithelial cells and epithelial-mucous cells in either their nuclear or cytoplasmic compartments. Results were presented as the volume fraction (Vv) of each element. The Vv of the maspinimmunopositive epithelial-mucous cytoplasm and nuclei was significantly higher in LGMEC than in GOC and OCMM (po0.001 and p ¼ 0.026, respectively). In the epithelial cells, no significant differences were observed among the lesions (p40.05). It is suggested that the high levels of maspin in the epithelial-mucous cells (in both cytoplasm and nuclei) in LGMEC may serve as a tool to distinguish it from GOC. This may be useful especially in equivocal cases and in small incisional biopsy samples. & 2008 Elsevier GmbH. All rights reserved.
Corresponding author. Tel.: +972 3640 9305; fax: +972 3640 9250.
E-mail addresses: [email protected] (M. Vered), [email protected] (I. Allon), [email protected] (A. Buchner), [email protected] (D. Dayan). 0065-1281/$ - see front matter & 2008 Elsevier GmbH. All rights reserved. doi:10.1016/j.acthis.2008.10.006
ARTICLE IN PRESS 162
Introduction The odontogenic epithelium and the salivary gland epithelia share a common origin, namely, the epithelium of the stomatodeum (Ten Cate, 1994). Both begin their embryonic development as a thickening within the epithelial lining of the stomatodeum during weeks 5–6 of gestation. Although each of these epithelial types develops into completely different and highly specific functioning organs, the neoplastic cells within the various pathologic lesions derived from both the odontogenic and salivary gland epithelia may show common cytomorphological features, such as basaloid, columnar-to-cuboidal cells with a palisading appearance, clear cells and epithelial-mucous secreting cells. A rare type of developmental odontogenic cyst, glandular odontogenic cyst (GOC), may demonstrate an aggressive biological behavior (Waldron, 2002). It is generally accepted that such a cyst has an odontogenic origin; however, it shows gland-like features reminiscent of salivary gland tumors, thus reflecting the common origin of the odontogenic and salivary gland epithelia (Ellis and Auclair, 1996; Waldron, 2002). On rare occasions, salivary gland tumors arise centrally within the jawbones. Their origin is attributed to neoplastic transformation of odontogenic epithelium, mainly dentigerous cysts (Ellis and Auclair, 1996), or to salivary gland tissue entrapped within the jawbones dating from the embryonic development (Bouquot et al., 2000). The most common central salivary gland tumor is mucoepidermoid carcinoma (MEC) of the low-grade type (LGMEC), a malignant neoplasm, with low metastatic potential (Ellis and Auclair, 1996). Central LGMEC of the jawbones is indistinguishable in both architecture and immunohistochemical phenotype of the cells from its salivary gland counterpart (Pires et al., 2004), and has histological features that overlap with those of GOC (Ellis and Auclair, 1996; Waldron, 2002). Differentiation of LGMEC from GOC bears both treatment and prognostic implications. Mammary serine protease inhibitor (maspin) was originally described in breast myoepithelium by Zou et al. (1994), and has since been detected in various normal glandular tissues, such as the prostate, pancreas, ovary and salivary glands, as well as in several benign and malignant types of epithelial neoplasms (Futscher et al., 2002). The tumor suppressor qualities of maspin have been reported in breast (Zou et al., 1994), prostate (Pierson et al., 2002) and oral cavity carcinoma (Xia et al., 2000; Vered et al., 2008). The inhibitory role
M. Vered et al. of maspin has been described in different stages of tumor formation and progression, including degradation of basal laminae following serine protease activities, connective tissue invasion, tumorinduced angiogenesis and development of metastases (Solomon et al., 2006). Maspin is a useful prognosticator of tumor progression and patient survival, e.g., in oral and breast cancer (Sabbatini et al., 2000; Xia et al., 2000). Its presence is related to sensitization of tumor cells to apoptosis (Lockett et al., 2006). The purpose of this study was to examine the immunohistochemical localisation of maspin in LGMEC, in comparison with GOC, and to assess its use as a marker to distinguish between these entities.
Materials and methods Study cases Cases of GOC (n ¼ 8), LGMEC (n ¼ 6) and various odontogenic cysts (radicular and dentigerous) with mucous metaplasia (OCMM, n ¼ 7), which served as a control group, were retrieved from the archival files of the Department of Oral Pathology and Oral Medicine, School of Dental Medicine, Tel-Aviv University. All samples were formalin-fixed and paraffin wax-embedded. Five micrometer-thick sections were cut and stained with hematoxylin and eosin, and mucicarmine (Mayer’s Mucicarmine, BioOptica, Milano, Italy) using conventional protocols.
Immunohistochemistry Three micrometer-thick sections were cut and mounted on positively charged microscope slides (OptiplusTM, Biogenex, San Ramon, CA, USA). These were then labelled immunohistochemically for binding of an anti-maspin antibody. After dewaxing and rehydrating by conventional methods, antigen retrieval was performed by incubating with NuclearDecloakerTM, pH ¼ 9.5 (Biocare Medical, Walnut Creek, CA, USA), in a pressure cooker for 12 min. Sections were then incubated with a mouse monoclonal primary antibody directed against maspin (EAW24, diluted 1:75, incubated overnight at 4 1C; Novocastra, Newcastle upon Tyne, UK) and binding was revealed using a Broad Spectrum Poly HRP conjugate ready-to-use kit (Zymed, San Francisco, CA, USA), employed according to manufacturer’s instructions. Sections were reacted with AEC substrate-chromagen kit (Zymed, San Francisco,
ARTICLE IN PRESS Maspin in central mucoepidermoid carcinoma and glandular odontogenic cyst CA, USA) employed according to manufacturer’s instructions, counterstained in Mayer’s hematoxylin (Pioneer Research Chemicals, Colchester, UK) and mounted in GVA mounting medium (Zymed, San Francisco, CA, USA).
Histomorphometry Two independent investigators (MV and IA) conducted the assessment. Immunomorphometry was first performed on GOC. Epithelial cells and epithelial-mucous cells were counted separately. Both cytoplasmic and nuclear immunolabelling were taken into consideration. The extent of maspin immunolabelling was assessed using a modified point-counting method (Dayan et al., 2000). At 400 magnification, a standard field defined by a 100-square grid (Olympus, Tokyo, Japan), comprising 121 intersections between each horizontal and vertical line was mounted on the microscope and placed over the cystic epithelium with one border tangentially touching the epithelial–connective tissue interface. In each examined field, intersections positioned beyond the cystic epithelium (i.e., cyst lumen) were deleted from the total number of intersections per field (i.e., 121). Point-counting was conducted on the following maspin-immunopositive elements: epithelial cell cytoplasm (EC) and nucleus (EN); epithelialmucous cell cytoplasm (EMC) and nucleus (EMN). When an intersection hit one of these elements, the specified element scored one point. A minimum of 300 intersections was counted in each section. Then, the sum of points overlying each specified element was calculated, divided by the total number of counted intersections in that section and multiplied by 100. This was defined as the volume fraction (Vv) of each maspin-immunopositive element. The procedure was applied to all sections in the GOC group to allow calculation of the mean Vv of each examined element in this group. The same approach was used to calculate the mean Vv of the examined elements in LGMEC and OCMM groups.
163
Results Histology Two oral pathologists (MV and DD) confirmed the original diagnoses by examining hematoxylin and eosin-stained sections from the cases. GOC cases selected for the study fulfilled the histopathological diagnostic criteria established by the World Health Organization (WHO), including cysts lined by non-keratinized stratified epithelium (partly squamous), which has a surface layer of acidophilic cuboidal or columnar cells, often forming irregular papillary projections (Kaplan et al., 2005). The rest of the cyst lining contains epithelial cells characterized by an eosinophilic cytoplasm and large, usually centrally located, nuclei and distinct polygonal cell outlines. Occasionally, the luminal surface of the epithelial cells exhibit cilia. Among the epithelial cells, varying numbers of epithelial–mucous cells are interposed. These are identified by their abundant, foamy cytoplasm and small, eccentric rather pyknotic nuclei, large cell size and round, cuboidal, ovoid, columnar or goblet shapes. Within the epithelial layer, there may be crypts/cyst-like/glandular-like spaces lined by cells, similar to those at the epithelial surface. Mucin may be demonstrable within the intraepithelial spaces and within the main cyst cavity (Figure 1). Histologically, LGMEC shows a cystic architecture with epithelial lining that varies in thickness and that features a cellular composition
Statistical analysis Inter-observer variations were assessed by the non-parametric Wilcoxon test. Differences in the mean Vv of the various elements among study groups were statistically analyzed using one-way ANOVA. Statistical significance was set at po0.05. The Statistical Package for the Social Sciences (SPSS 14) software was used for computations.
Figure 1. Glandular odontogenic cyst (GOC) showing slightly thickened cyst lining composed of cuboidal-tocolumnar epithelial cells and mucus-secreting cells (finely granular, bluish cytoplasm), the latter also seen as part of the lining of the intraepithelial gland-like structures. Delicate cilia can be seen on the luminal surface of the epithelial cells (hematoxylin and eosin, Scale bar ¼ 200 mm).
ARTICLE IN PRESS 164
M. Vered et al.
Figure 2. Low-grade mucoepidermoid carcinoma (LGMEC) showing a fragment of a cystic structure lined by epithelial cells, among which mucus-producing cells are interposed. Intraepithelial gland-like structures are present. Cytologically, the neoplastic cells are benign without atypia or mitotic figures (hematoxylin and eosin, Scale bar ¼ 200 mm).
comprising of an admixture of epithelial/epidermoid and mucus-producing cells with occasional formation of gland-like intra-epithelial microcysts or papillary-cystic architecture (Figure 2). Often, the architectural and cellular composition in GOC and LGMEC are remarkably similar (Figure 3). Based on the histological similarities and common immunohistochemical profiles of central MEC and salivary gland MEC (Pires et al., 2004), the LGMEC cases selected in this study were of minor salivary gland origin and fulfilled the histopathological criteria established for the low-grade type of MEC (Goode and El-Naggar, 2005). The OCMM consisted of developmental (e.g., dentigerous) and inflammatory (e.g., radicular) cysts that demonstrated prominent mucous metaplasia of the lining epithelium.
Immunohistochemistry and mucicarmine staining In all cases of GOC, histochemical staining with mucicarmine revealed positively stained epithelial– mucous cells and mucous-containing pools within the lining epithelium (Figure 4a). The epithelial cells did not stain with mucicarmine. Positive immunohistochemical labelling of maspin was found within the cyst lining in all cases. Immunopositivity in the epithelial cells was usually diffuse in both the cytoplasm and nuclei, and absent in the epithelial–mucous cells (Figure 4b). In all LGMEC cases, similar to GOC, mucicarmine
Figure 3. GOC versus LGMEC Architectural overlapping between the lesions includes a similar thickness of the lining epithelium and arrangement of the epithelial and epithelial–mucous cells (hematoxylin and eosin, Scale bar ¼ 200 mm).
staining was found in the epithelial–mucous cells and mucous-containing pools (Figure 5a). However, unlike in GOC, numerous epithelial–mucous cells in LGMEC demonstrated maspin immunopositivity in both the cytoplasm and nuclear compartments (Figure 5b). Most of the epithelial cells in LGMEC were maspin-immunopositive in both cytoplasmic and nuclear compartments. All OCMM cases demonstrated maspin immunopositivity in a pattern similar to that of GOC.
Immunomorphometry Inter-observer variations were statistically insignificant (p40.05). Results are summarized in Table 1. The present quantitative measurements further support the immunomorphological differences found between LGMEC and GOC, as previously described. The mean Vv of EMC was significantly
ARTICLE IN PRESS Maspin in central mucoepidermoid carcinoma and glandular odontogenic cyst
165
Figure 4. Glandular odontogenic cyst (GOC) (a) showing extensive mucicarmine staining (purple) of the epithelial-mucous cells (Scale bar ¼ 300 mm). (b) The same area as Figure 1a, showing positive maspin immunolabelling in the epithelial cells, but not in the epithelial–mucous cells (Scale bar ¼ 300 mm).
Figure 5. Low-grade mucoepidermoid carcinoma (LGMEC) (a) showing a fragment of a cystic structure lined by epithelial and epithelial–mucous cells; the latter are stained with mucicarmine (Scale bar ¼ 300 mm). (b) The same area as illustrated in Figure 2a, showing maspin-immunopositive cytoplasm and nuclei of the epithelial–mucous cells (Scale bar ¼ 300 mm).
higher in LGMEC (16.5%710.8%) than in GOC and OCMM (1.572% and 171%, respectively), (po0.001). The mean Vv of EMN in LGMEC (1.7%, 41.5%) was significantly higher compared to GOC and OCMM (0.370.5% and 0.470.7%, respectively) (p ¼ 0.026). No significant differences were found in the mean Vv of the maspin-immunopositive epithelial cells (in both cytoplasmic and nuclear compartments) among the different lesions (p40.05). No significant differences were found in the mean Vv of the epithelial and epithelial– mucous cells between GOC and OCMM (p40.05).
genic lesions (i.e., lateral periodontal cyst/ botryoid cyst), and between GOC and central LGMEC (Sciubba et al., 2001), has prompted many researchers to examine a wide panel of markers that could possibly aid in differentiating between these lesions. This is important because of the different treatment needs for patients with a developmental odontogenic cyst compared to lowgrade carcinoma. Since GOC is a relatively rare lesion of the jawbones, only small series and sporadic cases are reported in the current literature. From the available data, the cytokeratin profile, cell cycle proteins (i.e., p53) and proliferation markers (i.e., Ki-67 and PCNA) in GOC odontogenic lesions (cysts and tumors) and LGMEC (central and salivary gland origin) show overlapping features (Pires et al., 2004; Kaplan et al., 2005; Shen et al., 2006). The subtle differences in these studies might reflect variations in epithelial
Discussion The overlap in the histopathological features between GOC and other developmental odonto-
ARTICLE IN PRESS 166
M. Vered et al.
Table 1. Mean Vv (7SD) of the maspin-immunopositive components in the various types of examined lesions.
EMC EMN EC EN
LGMEC
GOC
OCMM
p value
16.5710.8 1.771.5 13.3712.6 8.9710.5
1.572 0.370.5 26.4719.9 12.375.5
171.1 0.771.1 25.4714.4 15.876.2
po0.001 p ¼ 0.026 NS NS
LGMEC – low-grade mucoepidermoid carcinoma; GOC – glandular odontogenic cyst; OCMM – odontogenic cyst with mucous metaplasia. EMC – epithelial–mucous cell cytoplasm; EMN – epithelial–mucous cell nuclei; EC – epithelial cell cytoplasm; EN – epithelial cell nuclei. NS – not significant.
maturation. No parameter demonstrated a consistent pattern that supports a definite separation between odontogenic lesions, GOC, and LGMEC. The results of our study clearly showed immunomorphological differences regarding maspin between LGMEC and GOC. These differences were further supported by quantitative measurements. While the epithelial–mucous cells in LGMEC showed quite extensive maspin immunolabelling, in GOC, maspin immunoreactivity was restricted to only a small fraction of these cells. The pattern of maspin immunolabelling in the epithelial and epithelial– mucous cells in OCMM was similar to that in GOC. These findings may be of substantial aid in the diagnosis of cases that present as large cystic lesions within the jawbones, where an incisional biopsy is performed. In cases where diagnosis of GOC or central LGMEC cannot be made based on the morphological features alone, especially when the biopsy sample is small, extensive maspin immunolabelling in the epithelial–mucous cells (in both cytoplasm and nuclei) may be used as a tool to favor a diagnosis of LGMEC rather than GOC. The present results should be further validated in studies comprising larger series of lesions. Several types of benign and malignant salivary gland tumors are maspin immunopositive (de Navarro et al., 2004; Nakashima et al., 2006); however, in respect to MEC, the present results were the first to show that the low-grade variant of this tumor was maspin immunopositive. The extent of maspin immunolabelling in LGMEC in the epithelial–mucous cells can be explained in terms of signals that are assumed to induce maspin expression in association with a more differentiated cell phenotype (Lockett et al., 2006). The epithelial–mucous cells in LGMEC represent a higher state of differentiation compared to other cell types often present in this tumor; the epithelial/
epidermoid and intermediate cells. The present results are in accordance with a recently published study that found that intermediate grade MECs (characterized by a smaller number of epithelial– mucous cells), together with other malignant salivary gland tumors of a similar grade, demonstrate loss of maspin expression (Schwarz et al., 2008). In regard to the nuclear immunolocalisation of maspin in GOC, we assumed that this was part of the pattern of immunolabelling in states of stress associated with the inflammatory infiltrate, which was usually found at varying degrees of severity in the vicinity of the cystic epithelium. In general, in stress-related conditions such as inflammation, maspin immunolabelling was observed at both cytoplasmic and nuclear compartments (Cao et al., 2005; Lockett et al., 2006). It is suggested that the significantly higher nuclear immunolabelling of maspin in LGMEC compared to GOC, reflects the fact that the neoplastic transformation is considered as the highest state of cellular stress. Nevertheless, the relatively high nuclear immunopositivity of maspin in LGMEC was compatible with less aggressive types of malignancies and favorable clinical outcomes (Mohsin et al., 2003; Dietmaier et al., 2006; Marioni et al., 2008). The general picture emerging from this study and the literature indicate a histological continuum of mucous cell containing-lesions starting from inflammatory and developmental odontogenic cysts with mucous cell metaplasia (Takeda et al., 2005), to inflammatory and developmental odontogenic cysts with focal GOC-like areas, to GOC developing within odontogenic cysts (Takeda et al., 2005), GOC, central LGMEC developing within odontogenic cysts (Tran et al., 2004; Holsinger et al., 2002; Sittitavornwong et al., 2006) and finally to central LGMEC. These lesions share some histopathological features, but have a wide range of biological behavior that may differ among lesions of the same type, and among different types of lesions. This may reflect the pluripotential character of the odontogenic epithelium, which has the ability to develop diverse types of lesions. In light of this, no definite separation can be made between central LGMEC and GOC, but a more frequent immunopositivity of the epithelial–mucous cells favors a diagnosis of LGMEC.
Acknowledgements The study was supported by the Ed and Herb Stein Chair in Oral Pathology, Tel Aviv University, Tel
ARTICLE IN PRESS Maspin in central mucoepidermoid carcinoma and glandular odontogenic cyst Aviv, Israel. The study was presented at the 1st International Congress on Oral Cancer organized by the International Academy of Oral Oncology (IAOO); May 2007, Amsterdam, The Netherlands.
References Bouquot JE, Gnepp DR, Dardick I, Hietanen JH. Intraosseous salivary tissue: jawbones examples of choristomas, hamartomas, embryonic rests, and inflammatory entrapment: another histogenetic source for intraosseous adenocarcinoma. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2000;90:205–17. Cao D, Wilentz RE, Abbruzzese JL, Ho L, Maitra A. Aberrant expression of maspin in idiopathic inflammatory bowel disease is associated with disease activity and neoplastic transformation. Int J Gastrointest Cancer 2005;36:39–46. Dayan D, Vered M, Paz T, Buchner A. Aging of human palatal salivary glands: a histomorphometric study. Exp Gerontol 2000;35:85–93. de Navarro RL, Martins MT, de Araujo VC. Maspin expression in normal and neoplastic salivary gland. J Oral Pathol Med 2004;33:435–40. Dietmaier W, Bettstetter M, Wild PJ, Woenckhaus M, Ru ¨mmele P, Hartmann A, et al. Nuclear Maspin expression is associated with response to adjuvant 5-fluorouracil based chemotherapy in patients with stage III colon cancer. Int J Cancer 2006;118: 2247–54. Ellis GL, Auclair PL. Central mucoepidermoid carcinoma. In: Rosai J, editor. Atlas of tumor pathology. Third series, Fascicle 17. Tumors of the salivary glands. Washington: AFIP; 1996. p. 172–5. Futscher BW, Oshiro MM, Wozniak RJ, Holtan N, Hanigan CL, Duan H, et al. Role for DNA methylation in the control of cell type-specific maspin expression. Nat Genet 2002;31:175–9. Goode RK, El-Naggar AK. Mucoepidermoid carcinoma. In: Barnes L, Eveson JW, Reichart P, Sidransky D, editors. WHO classification of tumours. Pathology and genetics – Head and neck tumours. Lyon: IARC; 2005. p. 219–20. Holsinger FC, Owens JM, Raymond AK, Myers JN. Central mucoepidermoid carcinoma of the mandible: tumorigenesis within a keratocyst. Arch Otolaryngol Head Neck Surg 2002;128:718–20. Kaplan I, Anavi Y, Manor R, Sulkes J, Calderon S. The use of molecular markers as an aid in the diagnosis of glandular odontogenic cyst. Oral Oncol 2005;41: 895–902. Lockett J, Yin S, Li X, Meng Y, Sheng S. Tumor suppressive maspin and epithelial homeostasis. J Cell Biochem 2006;97:651–60. Marioni G, Staffieri C, Staffieri A, De Filippis C, Blandamura S. Maspin tumour-suppressing activity in head and neck squamous cell carcinoma: emerging evidence and therapeutic perspectives. Acta Otolaryngol 2008;10:1–5.
167
Mohsin SK, Zhang M, Clark GM, Craig Allred D. Maspin expression in invasive breast cancer: association with other prognostic factors. J Pathol 2003;199: 432–5. Nakashima D, Uzawa K, Kasamatsu A, Koike H, Endo Y, Saito K, et al. Protein expression profiling identifies maspin and stathmin as potential biomarkers of adenoid cystic carcinoma of the salivary glands. Int J Cancer 2006;118:704–13. Pierson CR, McGowen R, Grignon D, Sakr W, Dey J, Sheng S. Maspin is up-regulated in premalignant prostate epithelia. Prostate 2002;53:255–62. Pires FR, Chen SY, da Cruz Perez DE, de Almeida OP, Kowalski LP. Cytokeratin expression in central mucoepidermoid carcinoma and glandular odontogenic cyst. Oral Oncol 2004;40:545–51. Sabbatini R, Federico M, Morselli M, Depenni R, Cagossi K, Luppi M, et al. Detection of circulating tumor cells by reverse transcriptase polymerase chain reaction of maspin in patients with breast cancer undergoing conventional-dose chemotherapy. J Clin Oncol 2000; 18:1914–20. Schwarz S, Ettl T, Kleinasser N, Hartman A, Reichert T-E, Driemel O. Loss of maspin expression is a negative prognostic factor in common salivary gland tumors. Oral Oncol 2008;44:563–70. Sciubba JJ, Fantasia JE, Kahn LB. Glandular odontogenic cyst. In: Rosai J, editor. Atlas of tumor pathology. Third series, Fascicle 29. Tumors and cysts of the jaws. Washington: AFIP; 2001. p. 31–4. Shen J, Fan M, Chen X, Wang S, Wang L, Li Y. Glandular odontogenic cyst in China: report of 12 cases and immunohistochemical study. J Oral Pathol Med 2006;35:175–82. Sittitavornwong S, Koehler JR, Said-Al-Naief N. Glandular odontogenic cyst of the anterior maxilla: case report and review of the literature. J Oral Maxillofac Surg 2006;64:740–5. Solomon LA, Munkarah AR, Schimp VL, Arabi MH, Morris RT, Nassar H, et al. Maspin expression and localization impact on angiogenesis and prognosis in ovarian cancer. Gynecol Oncol 2006;101:385–9. Takeda Y, Oikawa Y, Furuya I, Satoh M, Yamamoto Y. Mucous and ciliated cell metaplasia in epithelial linings of odontogenic inflammatory and developmental cysts. J Oral Sci 2005;47:77–81. Ten Cate AR. Structure of the oral tissues. In: Ten Cate AR, editor. Oral histology. Developmental, structure and function. St. Louis: Mosby; 1994. p. 45–57. Tran P-T, Cunningham CJ, Baughman RA. Glandular odontogenic cyst. J Endod 2004;30:182–4. Vered M, Allon I, Dayan D. Maspin, p53, p63 and Ki-67 in epithelial lesions of the tongue: from hyperplasia through dysplasia to carcinoma. J Oral Pathol Med 2008 [in press]. Waldron CA. Odontogenic cysts and tumors. In: Neville BW, Damm DD, Allen CM, Bouquot JE, editors. Oral and maxillofacial pathology. Philadelphia, PA: WB Saunders; 2002. p. 607–8.
ARTICLE IN PRESS 168 Xia W, Lau Y-K, Hu M C-T, Li L, Johnston DA, Sheng S, et al. High tumoral maspin expression is associated with improved survival of patients with oral squamous cell carcinoma. Oncogene 2000;19:2390–403.
M. Vered et al. Zou Z, Anisowicz A, Hendrix MJ, Thor A, Neveu M, Sheng S, et al. Maspin, a serpin with tumor-suppressing activity in human mammary epithelial cells. Science 1994;263:526–9.