Journal of

Oral Pathology & Medicine

doi: 10.1111/j.1600-0714.2009.00768.x J Oral Pathol Med (2009) 38: 639643 2009 The Authors. Journal compilation 2009 Blackwell Munksgaard All rights reserved interscience.wiley.com/journal/jop

Evaluation of myobroblasts in oral epithelial dysplasia and squamous cell carcinoma

S. Etemad-Moghadam1, M. Khalili2, F. Tirgary3, M. Alaeddini1
Dental Research Center, Tehran University of Medical Sciences, Iran, 2Department of Oral and Maxillofacial Pathology, Faculty of Dentistry, Tehran University of Medical Sciences, Iran; 3Department of Surgical Pathology, Cancer Institute, Imam Hospital, Tehran, Iran
1

BACKGROUND: Carcinogenesis is accompanied by a number of changes in the adjacent stroma including the appearance of myobroblasts. The purpose of this study was to evaluate and compare the presence of myobroblasts in normal mucosa, oral epithelial dysplasia, and different grades of oral squamous cell carcinoma. METHODS: The study sample consisted of three groups, including 40 oral squamous cell carcinomas, 15 dysplasias, and 15 sections of normal oral epithelium. Vimentin, desmin, and alpha-smooth muscle actin were used to identify myobroblasts. RESULTS: The percentage and intensity of alpha-smooth muscle actin were examined, and positive immunostaining was observed in the myobroblasts of all squamous cell carcinomas; however these cells did not stain in the dysplasias or normal epithelium specimens. The presence of myobroblasts was signicantly higher in oral squamous cell carcinomas compared to both, dysplasias and normal mucosa cases (P < 0.001). A signicant difference was not observed between the different grades of oral squamous cell carcinoma (P = 0.2). CONCLUSIONS: These ndings show the presence of myobroblasts in the stroma of oral squamous cell carcinoma but not dysplasia and normal mucosa, suggesting further investigation to clarify the role of myobroblasts in the carcinogenesis of this tumor. J Oral Pathol Med (2009) 38: 639643 Keywords: oral squamous cell carcinoma; myobroblast; dysplasia; stroma

Introduction
Squamous cell carcinoma (SCC) is the most frequent type of oral malignancy and is associated with a high mortality rate (1). Transformation of normal oral mucosa to squamous dysplasia and ultimately SCC represents a complicated process involving numerous etiologic factors (2). Approximately 1020% of oral dysplasias develop carcinomatous features and they eventually invade beyond the basement membrane (3). There is an increasing amount of interest in the molecular and biological events that occur during the transition of dysplastic epithelium to SCC (2, 4). Over the past decade, several studies have shown that the microenvironment or stroma of neoplastic tissues plays an active role in tumor progression. Concurrent with the conversion of non-diseased epithelial tissue to pre-cancerous epithelium to carcinoma, the stroma also changes from normal to primed to activated or tumor associated (5, 6). Remodeling of the extracellular matrix or stromagenesis is initiated by tumor cells, while stromal cells are responsible for the organization of this process (5). Fibroblasts are considered as one of the most important mesenchymal cells involved in tumor progression (5, 7). For example, broblasts were shown to induce epithelial proliferation in smokeless-tobaccolesions, following stimulation by tobacco (8). Transdifferentiation of broblasts to myobroblasts is a crucial and early event in tumorigenesis, which is mediated by growth factors and cytokines expressed by tumor cells (57, 9). Myobroblasts secrete numerous growth factors and inammatory mediators that stimulate epithelial cell proliferation (10). Therefore, these cellular elements play an important role in tumoral invasion and use a combination of dierent factors in the course of neoplastic growth and development. An invasion-promoting role of myobroblasts has been shown in numerous aggressive and malignant lesions (1115). In addition, decreased CD34+ brocytes along with an increase in smooth muscle actin (SMA)-positive myobroblasts has been observed in invasive oral SCCs (OSCCs) (16). Similarly, recent

Correspondence: Mojgan Alaeddini, Dental Research Center, Dental Faculty, Tehran University of Medical, Sciences, Ghods St., Enghelab Ave., P.O. Box: 14155-5583, Postal code: 14174 Tehran Iran. Tel: +98 21 88986677, Fax: +98 21 88986688, E-mail: malaeddini@ sina.tums.ac.ir Accepted for publication February 4, 2009

Myobroblasts in oral cancer

Etemad-Moghadam et al.

640

investigations have demonstrated the induction of myobroblasts by OSCC-derived factors which in turn can promote carcinomatous proliferation resulting in neoplastic growth (17, 18). However, the number of studies evaluating the role of myobroblasts in OSCC remains limited. The aim of this investigation was to compare the distribution of myobroblasts in normal oral mucosa, oral epithelial dysplasia, and dierent histological grades of OSCC.

was used for vimentin. The primary antibodies were replaced by non-immune mouse serum at the same dilutions for negative controls. Immunohistochemical analysis Stromal spindle cells positive for a-SMA and vimentin, but negative for desmin, were regarded as myobroblasts (11). Immunostaining was assessed by the evaluation of the staining intensity and percentage of a-SMA, according to the method proposed by Tuxhorn et al. (21). The percentage of immunopositive cells in the noninammatory and non-endothelial stromal cells immediately adjacent to the carcinomatous islands and normal and dysplastic epithelium was recorded as: 0% = no positive cells, 1 = 133% positive cells, 2 = 3466% positive cells, and 3 = 67100% positive cells. Staining intensity was considered 0 when there was no staining; 1, in parts where positivity was observed only at a magnication of 400; 2, in cases where staining was obvious at 100, but not 40; and 3, in elds where immunopositive cells were seen even at 40. In most specimens, there was either no or minimal background staining when examining dierent areas at high magnication (400). Multiplication of the percentage and intensity scores comprised the staining index of each specimen. This index was classied as zero (0), low (12), moderate (34); and high (69). All hematoxylin eosin- and immunohistochemicalstained sections were examined by two observers and disagreements were resolved utilizing a double-headed microscope. Statistical analysis Dierences in the presence of myobroblasts between groups were analyzed using KruskalWallis and Mann Whitney tests. The Bonferroni method was used to adjust the P-value in multiple comparisons. P-values of less than 0.05 were considered to be statistically signicant.

Materials and methods

Samples Medical records of primary OSCC and epithelial dysplasia patients admitted to our institutions from 1997 to 2007 were retrieved. Normal oral mucosa was obtained from patients undergoing tooth extraction for orthodontic purposes after signing written consents. None of the subjects indicated any sign of inammatory gingival or periodontal disease. According to the reviewed charts, formalin-xed paran-embedded blocks of all cases with excisional biopsies excluding the lip were collected from the pathology archives. Diagnoses were conrmed by two oral pathologists based on the WHO classication of cancer and precancerous lesions (19), using hematoxylin eosin stained slides. Intraepithelial dysplasias were recorded as mild, moderate, and severe (19), and the SCC cases were histologically graded as well-, moderately- and highlydierentiated (20). Inammation was minimal in the normal tissues and all were devoid of pathologic conditions. The protocol for this study has been approved by the ethics committee of Tehran University of Medical Sciences. Staining procedure Immunostaining was performed according to the manufacturers instructions, using primary mouse monoclonal antibodies against a-SMA (clone 1A4; DAKO, Denmark), desmin (clone D33; DAKO), and vimentin (clone V9; DAKO). In brief, 4-lm sections were mounted on 3-aminopropyltriethoxysilane coated slides; de-waxed in xylene and hydrated in graded alcohol. Endogenous peroxidase activity was blocked by incubation in H2O2 (6%) for 10 min followed by rinsing in distilled water. For heat-induced antigen retrieval (microwave oven, 56C, 20 min), a-SMA and desmin were treated with 10 mM tris buer, 1 mM EDTA at a pH of 9; and vimentin with 10 mM citrate buer at a pH of 6. All slides were incubated with a-SMA (1:100) and desmin (1:150) for 1 h and with vimentin (1:50) for 120 min at room temperature. The sections were washed in phosphate buer saline (PBS) for 10 min and stained with the streptavidinbiotin peroxidase detection kit (DAKO), followed by rinsing with PBS and development in 3,3-diaminobenzidine. Mayers hematoxylin was employed for counterstaining. Positive and negative controls were run simultaneously with the study specimens. Positive controls were obtained from the normal colon tissue for a-SMA and desmin, and human tonsil
J Oral Pathol Med

Results
A total of 70 cases including 40 oral squamous cell carcinomas (29 males and 11 females, with an age range of 4277 years), 15 oral dysplasias (ve males and 10 females, between 29 and 61 years), and 15 sections of normal oral epithelium (six males and nine females, aged 1827) were evaluated in this study. Squamous cell carcinomas were subdivided into lowgrade (35%), intermediate-grade (40%), and highgrade (25%) specimens. The intraepithelial dysplasias consisted of 33.4% mild, 40% moderate, and 26.6% severe cases. Blood vessels present within the connective tissue of the immunostained sections served as positive internal control and stained with an intensity of three in all cases. These internal controls conrmed that antigenic expression was correctly maintained in the study sample. There were no myobroblasts in the stroma of normal mucosa (Fig. 1) and oral dysplasia (Fig. 2) specimens, resulting in a staining index of zero in these cases. All SCCs demonstrated immunostaining

Myobroblasts in oral cancer

Etemad-Moghadam et al.

641

Figure 1 A representative section of normal oral epithelium showing negative immunostaining for a-SMA in stromal cells. Strong positive staining of blood vessel walls and RBC extravasation is seen (original magnication, 200).

Figure 3 Expression of a-SMA in a representative section of oral Squamous cell carcinoma (OSCC). Note the high staining index of myobroblasts in the connective tissue (original magnication, 200).

Table 1 Immunohistochemical staining index for alpha-smooth muscle actin in normal, dysplastic epithelium and OSCC a-SMAa staining index Zero Normal oral epithelium Oral dysplasia Oral squamous cell carcinoma 15 15 0 Low 0 0 8 Moderate 0 0 19 High 0 0 13 P-value <0.001*

*KruskalWallis test. a Alpha-smooth muscle actin.

Discussion
Considering that genetic and epigenetic factors are capable of aecting the entire tissue (epithelium and lamina propria), it would be logical to assume that carcinogenesis and tumor progression result from a defective response of both compartments (22). Several studies have conrmed the important role of the carcinomatous stroma in tumorigenesis, invasion and metastasis (5, 7, 9). However, the exact mechanism by which dierent stromal cell types such as myobroblasts can inuence neoplastic cells remains unclear. Albini and Sporn (2) have even suggested a revision of the nomenclature of malignant epithelial tumors, i.e. carcinoma. Accordingly, the altered epithelial cells of SCC would not be solely responsible for carcinogenesis, and dierent stromal factors participate in its development via communication with the epithelial elements. Transdierentiation of broblasts to myobroblasts is considered an important event that occurs in the stroma of several invasive carcinomas (11, 12, 15, 23). According to the results obtained in this study, the number of myobroblasts was signicantly higher in OSCC compared to normal and dysplastic epithelium, which were both devoid of myobroblasts. These
J Oral Pathol Med

Figure 2 Negative immunostaining for a-SMA in the stromal cells of a representative dysplasia biopsy specimen. Note the strong positive staining of blood vessel walls (original magnication, 200).

for a-SMA with dierent intensities (Fig. 3). The staining index was low in eight (20%), moderate in 19 (47.5%), and high in 13 (32.5%) of the squamous cell carcinomas (Table 1). A statistically signicant dierence in the presence of myobroblasts was found between the SCCs, oral dysplasias and normal mucosa specimens (Kruskal Wallis test; P < 0.001). The presence of myobroblasts was signicantly higher in the SCCs compared to both, dysplasia and normal mucosa cases (MannWhitney test; adjusted P < 0.001). There was no signicant dierence between the dysplasia and normal mucosa samples (P = 1). The distribution of myobroblasts was not signicantly dierent between the three histological grades of oral squamous cell carcinoma (P = 0.2).

Myobroblasts in oral cancer

Etemad-Moghadam et al.

642

ndings were in agreement with those reported by Zidar et al. (11) who also found a lack of myobroblasts in normal and dysplastic laryngeal epithelium. The presence of myobroblasts in dysplastic and carcinomatous oral epithelium has not been extensively investigated. We were able to nd only four studies published in the English literature that evaluated the transdierentiation of broblasts to myobroblasts in OSCC (1618, 24). In two of the articles, non-neoplastic epithelium adjacent to invasive SCCs was considered as normal tissue and both failed to demonstrate myobroblasts in the approximal connective tissue. These investigations showed an increased amount of myobroblasts in the stroma of SCCs, which was in accordance with the results obtained in this study. Neither of the two papers had included oral epithelial dysplasia in their samples (16, 18). Lewis et al. (24) demonstrated the presence of myobroblasts in the vicinity of invasive SCC but not in benign mucosal polyps. These cells were also absent in the stroma distant from carcinomatous epithelial islands. Dysplasia was not included in their study sample. Kellermann et al. (17) in a correspondence article reported the prognostic signicance of myobroblasts in SCC of the tongue, normal controls and premalignant leukoplakias with histological dysplasia. Similar to the current investigation, no myobroblasts were found in the stroma of normal mucosa and epithelial dysplasia; however they were detected at the invasive front of the SCCs. It is noteworthy that the increase in myobroblasts found in SCCs may be due to an inducing eect of the carcinomatous component. Epithelialstromal interactions, dierent growth factors released by malignant epithelial cells or numerous other processes may be responsible for the appearance of myobroblasts. In this study, the presence of myobroblasts did not show a signicant dierence between the three histologic grades. This was in agreement with the results obtained by Kellermann et al. (18) who were also unable to nd a correlation between SCC dierentiation and the observation of myobroblasts. These ndings may suggest that the transdierentiation of myobroblasts is induced somewhere in the invasive stage of SCC, and further loss of tumoral dierentiation (increased grade) would not aect the number of these cells. Statistical evaluation was not performed for the dierent grades of intraepithelial dysplasias due to the small number of cases. According to the tissue organization eld theory, cells are normally in a proliferative state and do not tend to be quiescent. Thus mutated epithelial stromal cells and disturbed stromalepithelial interactions may be equally responsible for the induction of carcinogenesis (5, 25), emphasizing the importance of the neoplastic microenvironment in oncogenesis. Numerous factors have been studied in OSCC including hepatocyte growth factor (HGF) (24, 26, 27) and transforming growth factor-b (24), which are secreted by myobroblasts (10) and related to their dierentiation (9), respectively. Additional investigation on these myobroblast-related factors in dierent stages

of carcinogenesis may help to clarify how and to what extent these cells contribute to carcinogenesis. A major issue in this study is that there was no way of knowing whether the examined dysplasias would have transformed into SCCs or remained as dysplastic entities. Another problem was that we were unable to obtain samples of carcinoma in-situ and superciallyinvasive SCC. This would have been helpful to draw probable conclusions as to where and when myobroblasts are rst observed. In summary, considering the lack of myobroblasts in normal and dysplastic oral epithelium and their appearance in SCC, it seems that the genetically altered epithelium (carcinomatous epithelium) may have an inductive eect on the adjacent stroma to produce myobroblasts. However, more sophisticated techniques are suggested to further clarify the exact mechanism by which these important cellular elements exert their eects on stromal and epithelial tissue compartments. In the event that our ndings are conrmed by future investigations, therapeutic targeting of myobroblasts, their byproducts or factors responsible for their transdierentiation from broblasts may be benecial to OSCC patients.

References
1. Neville BW, Day TA. Oral cancer and precancerous lesions. CA Cancer J Clin 2002; 52: 195215. 2. Kuer R, Lombardi T. Premalignant lesions of the oral mucosa. A discussion about the place of oral intraepithelial neoplasia (OIN). Oral Oncol 2002; 38: 12530. 3. Johnstone S, Logan RM. Expression of vascular endothelial growth factor (VEGF) in normal oral mucosa, oral dysplasia and oral squamous cell carcinoma. Int J Oral Maxillofac Surg 2007; 36: 2636. 4. Warnakulasuriya S. Lack of molecular markers to predict malignant potential of oral precancer. J Pathol 2000; 190: 4079. 5. Beacham DA, Cukierman E. Stromagenesis: the changing face of broblastic microenvironments during tumor progression. Semin Cancer Biol 2005; 15: 32941. 6. Amatangelo MD, Bassi DE, Klein-Szanto AJ, Cukierman E. Stroma-derived three-dimensional matrices are necessary and sucient to promote desmoplastic dierentiation of normal broblasts. Am J Pathol 2005; 167: 47588. 7. Liotta LA, Kohn EC. The microenvironment of the tumour-host interface. Nature 2001; 411: 3759. 8. Dabelsteen S, Christensen S, Gron B, Bardow A, Dabelsteen E. HGF is released from buccal broblasts after smokeless tobacco stimulation. Oral Oncol 2005; 41: 50914. 9. De Wever O, Mareel M. Role of tissue stroma in cancer cell invasion. J Pathol 2003; 200: 42947. 10. Powell DW, Miin RC, Valentich JD, Crowe SE, Saada JI, West AB. Myobroblasts. I. Paracrine cells important in health and disease. Am J Physiol 1999; 277: C19. 11. Zidar N, Gale N, Kambic V, Fischinger J. Proliferation of myobroblasts in the stroma of epithelial hyperplastic lesions and squamous carcinoma of the larynx. Oncology 2002; 62: 3815. 12. Tomas D, Kruslin B. The potential value of (Myo)broblastic stromal reaction in the diagnosis of prostatic adenocarcinoma. Prostate 2004; 61: 32431.

J Oral Pathol Med

Myobroblasts in oral cancer

Etemad-Moghadam et al.

13. Adegboyega PA, Miin RC, Dimari JF, Saada JI, Powell DW. Immunohistochemical study of myobroblasts in normal colonic mucosa, hyperplastic polyps, and adenomatous colorectal polyps. Arch Pathol Lab Med 2002; 126: 82936. 14. Vered M, Shohat I, Buchner A, Dayan D. Myobroblasts in stroma of odontogenic cysts and tumors can contribute to variations in the biological behavior of lesions. Oral Oncol 2005; 41: 102833. 15. Surowiak P, Suchocki S, Gyory B, et al. Stromal myobroblasts in breast cancer: relations between their occurrence, tumor grade and expression of some tumour markers. Folia Histochem Cytobiol 2006; 44: 1116. 16. Barth PJ, Schenck ZU, Schweinsberg T, Ramaswamy A, Moll R. CD34+ brocytes, alpha-smooth muscle antigenpositive myobroblasts, and CD117 expression in the stroma of invasive squamous cell carcinomas of the oral cavity, pharynx, and larynx. Virchows Arch 2004; 444: 2314. 17. Kellermann MG, Sobral LM, Da Silva SD, et al. Myobroblasts in the stroma of oral squamous cell carcinoma are associated with poor prognosis. Histopathology 2007; 51: 84953. 18. Kellermann MG, Sobral LM, Da Silva SD, et al. Mutual paracrine eects of oral squamous cell carcinoma cells and normal oral broblasts: induction of broblast to myobroblast transdierentiation and modulation of tumor cell proliferation. Oral Oncol 2008; 44: 50917. 19. Pindborg JJ, Reichart PA, Smith CJ, Van Der Waal I. Histological typing of cancer and precancer of the oral mucosa. In: Pindborg JJ, Reichart PA, Smith CJ, Van Der Waal I, eds. World Health Organization International Histological Classication of Tumors, 2nd edn. Berlin Heidelberg, New York: Springer-Verlag, 1997; 929. 20. Neville BW, Damm DD, Allen CM, Bouquot JE. Oral & Maxillofacial Pathology. Philadelphia, PA: W.B. Saunders, 2002.

21. Tuxhorn JA, Ayala GE, Smith MJ, Smith VC, Dang TD, Rowley DR. Reactive stroma in human prostate cancer: induction of myobroblast phenotype and extracellular matrix remodeling. Clin Cancer Res 2002; 8: 291223. 22. Albini A, Sporn MB. The tumour microenvironment as a target for chemoprevention. Nat Rev Cancer 2007; 7: 139 47. 23. Orimo A, Tomioka Y, Shimizu Y, et al. Cancer-associated myobroblasts possess various factors to promote endometrial tumor progression. Clin Cancer Res 2001; 7: 3097 105. 24. Lewis MP, Lygoe KA, Nystrom ML, et al. Tumourderived TGF-beta1 modulates myobroblast dierentiation and promotes HGF SF-dependent invasion of squamous carcinoma cells. Br J Cancer 2004; 90: 82232. 25. Weaver VM, Gilbert P. Watch thy neighbor: cancer is a communal aair. J Cell Sci 2004; 117: 128790. 26. Daly AJ, Mcilreavey L, Irwin CR. Regulation of HGF and SDF-1 expression by oral broblasts - implications for invasion of oral cancer. Oral Oncol 2008; 44: 64651. 27. Uchida D, Kawamata H, Omotehara F, et al. Role of HGF c-met system in invasion and metastasis of oral squamous cell carcinoma cells in vitro and its clinical signicance. Int J Cancer 2001; 93: 48996.

643

Acknowledgements
This study was supported by a grant (no. 132 2365) from Tehran University of Medical Sciences and Health Services. The authors would like to thank the Department of Oral and Maxillofacial Pathology, Faculty of Dentistry, Tehran University of Medical Sciences and Cancer Institute, Imam Hospital for providing paran blocks.

Conict of interest statement

None declared.

J Oral Pathol Med