Human Pathology (2014) 45, 851–857

www.elsevier.com/locate/humpath

Original contribution

SOX2 expression in hypopharyngeal, laryngeal, and sinonasal squamous cell carcinoma☆,☆☆ Rocío González-Márquez MD, José Luis Llorente MD, PhD, Juan P. Rodrigo MD, PhD, Juana M. García-Pedrero PhD, César Álvarez-Marcos MD, PhD, Carlos Suárez MD, PhD, Mario A. Hermsen PhD ⁎ Department of Otolaryngology, IUOPA, Hospital Universitario Central de Asturias, Oviedo, Spain Received 11 July 2013; revised 8 December 2013; accepted 11 December 2013

Keywords: Squamous cell carcinoma; Sinonasal; Larynx; Hypopharynx; SOX2; Immunohistochemistry; Prognosis

Summary Squamous cell carcinoma (SCC) of the head and neck display high frequencies of DNA copy number gains at chromosomal region 3q26-27. Recently SOX2 has been postulated as a driver oncogene for these amplifications; however, its role as a prognostic marker is still a matter of debate. The aim of this study was to evaluate the involvement of SOX2 protein expression in three different sublocalizations of head and neck SCC and its possible role as prognostic marker. SOX2 expression was analyzed by immunohistochemistry in 102 pharyngeal, 67 laryngeal, and 51 sinonasal SCCs, and the relation to clinicopathological and follow-up data was studied by χ2 and Kaplan-Meier analysis. SOX2 expression was significantly (P = .002) more frequent in hypopharyngeal and laryngeal SCC (38%, 39/101) and (42%, 28/67), respectively, compared to sinonasal cancer SCC (14%, 7/51). SOX2 expression did not correlate to disease stage, T or N classification, lymph node metastasis, recurrence or clinical outcome in any of the three sublocalizations. These results indicate that SOX2 expression is a common event in hypopharynx and larynx, but not in sinonasal SCC. The absence of correlation to clinical outcome, may suggest a role for SOX2 in tumor initiation, but not in tumor progression. © 2014 Elsevier Inc. All rights reserved.

1. Introduction Squamous cell carcinoma (SCC) of the head and neck (HNSCC) is the sixth most common neoplasm in the world, representing 2% to 5% of all malignancies [1]. The great

☆

majority is located in oral cavity, larynx and pharynx; sinonasal SCC may be considered a rare tumor. Tobacco and alcohol, as well as human papilloma virus infection, are well known etiological factors, the latter especially in oropharyngeal cancer. HNSCC display a characteristic profile of

Disclosure: The authors declare that they are not aware of conflicts of interest, financial or otherwise. Funding sources: This study was supported by grant PI08-1599 and EMER07-048 of Fondos de Investigación Sanitaria (FIS) and RD12/0036/0015 of Red Temática de Investigación Cooperativa en Cáncer (RTICC), Spain, and the FEDER Funding Program from the European Union. ⁎ Corresponding author. Department Otorrinolaringología, Instituto Universitario de Oncología del Principado de Asturias (IUOPA), Edificio H Covadonga 1ª Planta Centro, lab 2, Hospital Universitario Central de Asturias, Celestino Villamil s/n, 33006 Oviedo, Spain. E-mail addresses: [email protected] (R. González-Márquez), [email protected] (J. L. Llorente), [email protected] (J. P. Rodrigo), [email protected] (J. M. García-Pedrero), [email protected] (C. Álvarez-Marcos), [email protected] (C. Suárez), [email protected] (M. A. Hermsen). ☆☆

0046-8177/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.humpath.2013.12.004

852 genetic aberrations, including losses at chromosomal regions 3p, 8p, 9p, 11qter, 17p, 18q, and gains at 3q, 5p, 7p, 7q, 8q, 11q13, 18p, and 20q [2-4]. The most frequent copy number gains and amplifications occur at 3q26-27. This region contains many genes, and this may explain the fact that up to the present time there is still no consensus on the driver gene. Copy number gains of the 3q26-27 region have been observed in SCC of many localizations [2-7]. In pulmonary and esophagus SCC, the SOX2 (sex-determining region Y-box 2) transcription factor has been identified as an oncogene and a driver gene for this locus [8-10]. SOX2 amplification and expression has been implicated in many tumor types, including breast, colon, ovarian, and nasopharyngeal cancer [11-14], but mostly in SCC of various localizations (lung, esophagus, head and neck). SOX2 has been reported to be amplified and overexpressed in 20% of squamous cell pulmonary cancer, being higher in smokers than in nonsmokers [10,15]. Hussenet et al [9] observed a frequency of SOX2 amplification similar to c-MYC, EGFR or ERRB2. Freier et al [16] reported copy number gains in 52% and SOX2 expression in 18% of oral SCC. The role of SOX2 as a prognostic marker is still a matter of debate. Some studies have shown that expression of SOX2 confers a better prognosis [17,18], but others found an association with worse clinical outcome or adverse clinical parameters, including recurrences, lymph node and distant metastasis, disease-free and overall survival [12-14]. Also in HNSCC the prognostic significance of SOX2 remains unclear. Therefore, the aim of our study was to analyze SOX2 protein expression in three different sublocalizations of HNSCC: pharynx, larynx and sinonasal SCC, and to evaluate its possible role as prognostic marker using clinicopathological and follow-up data.

2. Materials and methods 2.1. Patients and samples A total of 220 patients with histologically confirmed SCC were included in our study, after informed consent and approval of the Hospital Universitario Central de Asturias Hospital Ethical Committee and following the Helsinki Declaration guidelines. In detail, there were 3 groups of patients, classified by their primary tumor localization: 102 hypopharyngeal, 67 laryngeal, and 51 sinonasal cancers. All patients had a single primary tumor and received no treatment prior to surgery. No patient had distant metastases at the time of diagnosis. TNM classification was performed according to the 7th Edition of the International Union Against Cancer TNM classification system and histological grade according to World Health Organization (WHO) classification [19]. Sixty-five percent (137/213) of patients received postoperative radiotherapy; for 10 patients this information is missing. Follow-up information was obtained of all patients, until the last occurrence for patients still alive,

R. González-Márquez et al. until the time of death or until the time of lost contact. The median follow-up time was 31 months (range, 0-211). A summary of the relevant clinicopathological features for each tumor localization is shown in Table 1.

2.2. Immunohistochemistry A total of 10 tissue microarray (TMA) blocks were assembled from formalin-fixed, paraffin-embedded tissues as previously described [20]. Briefly, areas of interest rich in non-necrotic areas were identified on corresponding hematoxylin and eosin–stained sections, and were marked with 2-mm circles on the source paraffin block. The source block was cored, and a 1-mm core was transferred to the

Table 1

Clinical features of the three groups of patients

Total Patients Gender Male Female Age Average Range T classification T1-2 T3-4 N classification N0 N+ Disease stage I-II III-IV Histological grade Well Moderate Poor Follow-up (months) Mean Median Range Radiotherapy No Yes Lost Local recurrence No Yes Distant metastasis No Yes Patient status Alive Died of disease Died of other causes Lost

Hypopharynx

Larynx

Sinonasal

102

67

51

99 3

67 0

37 14

59 43-84

62 36-86

66 47-91

29 73

21 46

5 46

16 86

33 34

38 13

7 95

13 54

4 47

23 41 38

25 30 12

18 10 23

25 14 0-95

41 37 1-97

31 18 1-211

25 75 2

39 20 8

12 39 0

39 63

36 31

9 42

61 41

49 18

46 5

23 62 8 9

22 26 9 10

8 37 6 0

SOX2 in head and neck cancer

853

recipient master block using the Beecher Tissue Microarrayer (Beecher Instruments, Silver Springs, MD). Three cores from different areas of the same tissue block were arrayed for each tumor. Three-micrometer sections were stained with hematoxylin and eosin and reviewed by a pathologist to determine whether the samples represented the tumor. Each TMA block included three samples of normal epithelium as an internal control. These samples were obtained from children less than 5 years old who underwent tonsillectomy due to respiratory obstruction. Immunohistochemistry (IHC) was performed using antibody SOX2 (AB5603, Millipore Ibérica SA Madrid, Spain), applied in a dilution of 1:1000, with an incubation time of 25 minutes in citrate buffer pH 9.0. Reactions were performed using an automatic staining workstation (Dako autostainer, DakoCytomation, Glostrup, Denmark) with the Envision system and diaminobenzidine chromogen as substrate. Negative controls were prepared by omitting the primary antibody. The IHC results were independently evaluated by two observers (R.G.M. and M.A.H.) without knowledge of the clinicopathological outcomes of the patients. Slides with indeterminate evaluation were re-evaluated, and a consensus was reached. Staining were scored as positive when N5% of cells showed nuclear expression [17,21,22].

2.3. Statistical analysis Univariate analysis by Pearson χ2 and Fisher’s exact tests was used for comparison between categorical variables. For time-to-event analysis, Kaplan-Meier curves were plotted. Differences between survival times were analyzed by the log-rank method. All tests were 2-sided. The values of P ≤ .05 were considered statistically significant.

3. Results 3.1. Clinicopathological features and follow-up Follow-up information on local recurrence, distant metastasis and patient status according to hypopharyngeal, laryngeal or sinonasal localization is presented in Table 1. During the course of follow-up, local recurrences were found in 63 (62%) of 102 of hypopharyngeal, 31 (46%) of 67 of laryngeal, and 42 (82%) of 51 of sinonasal cancer patients. Distant metastasis occurred in 41 (40%) of 102 hypopharynx, 18 (27%) of 67 larynx and 5 (10%) of 51 sinonasal tumors. The overall 5-year survival rates were 28%, 57%, and 20% for all hypopharynx, larynx, and sinonasal cancer patients, respectively (Fig. 1). The median disease-free time was 9 months for hypopharyngeal, 30 months for laryngeal, and 6 months for sinonasal SCC (Fig. 1). The most frequent causes of death were local recurrences and distant metastases. For all three sublocaliza-

Fig. 1 Overall (A) and disease-free (B) Kaplan–Meier survival curves of 102 hypopharyngeal, 67 laryngeal and 51 sinonasal SCC patients.

tions, both overall and disease-free survival correlated to disease stage and lymph node metastasis at the time of diagnosis, and to a lesser extent to T classification and histological grade (Table 2).

3.2. SOX2 expression in relation to clinicopathological features and follow-up Positive staining for SOX2, mainly localized in the nucleus, was observed in cancer cells for all sublocalizations (Fig. 2). All normal control samples showed absence of SOX2 staining, except for a weak cytoplasmatic staining in the basal layer. Hypopharyngeal and laryngeal SCC showed significantly more positive cases than sinonasal cancer, with respectively 39 (38%) of 102 and 28 (42%) of 67 versus 7 (14%) of 51 cases (P = .002). With regard to the clinicopathological data of each of the three subgroups separately, SOX2 expression did not

854 Table 2

R. González-Márquez et al. Overall and disease-free survival analysis by the by the log-rank method in relation to clinicopathological features

Disease stage I + II vs III + IV P Log rank T classification T1 + T2 vs T3 + T4 P Log rank N classification N0 vs N+ P Log rank Histological grade Well vs moderate vs poor P Log rank

Hypopharynx (n = 102)

Larynx (n = 67)

Overall Survival

Disease-free Survival

Overall Survival

.051 3.822

.095 2.792

.014 6.006

.005 7.854

.372 .798

.039 4.254

.188 .731

.529 .397

.020 5.455

.027 4.866

.150 2.069

.036 4.419

.026 4.963

.056 3.645

.000 12.478

.003 8.939

.821 .051

.197 1.667

.804 .435

.730 .629

.040 6.453

.200 3.215

.666 .814

.189 3.329

correlate to disease stage, T classification, lymph node metastasis, or the development of recurrence or distant metastasis (Table 3). Only the laryngeal tumors showed an association between SOX2 expression and histological grade (Table 3). Finally, there were no significant variations in overall or disease-free survival with regard to SOX2 expression in all sublocalizations (Fig. 3).

Sinonasal (n = 51) Disease-free Survival

Overall Survival

Disease-free Survival

4. Discussion Frequent chromosomal arm 3q26-27 copy number gains and amplifications in SCC of various localizations, have led to the postulate of SOX2 as driver gene [8-10]. SOX2 is a critical transcription regulator of normal stem cell function in embryonic and neural stem cells. During embryogenesis,

Fig. 2 Representative photomicrographs of immunohistochemical staining of SOX2 expression. A, No expression in normal sinonasal mucosa. Nuclear expression in sinonasal (B), laryngeal (C) and hypopharyngeal (D) SCC. Original magnification ×100.

SOX2 in head and neck cancer Table 3

855

Relation between clinicopathological features and SOX2 expression for the three HNSCC sublocalizations SOX2

Disease stage T classification N classification Histological grade

Local recurrence Distant metastasis

I + II III + IV T1-2 T3-4 N− N+ Well Moderate Poor No Yes No Yes

Hypopharynx (n = 102)

Larynx (n = 67)

−

+

−

+

6 57 17 46 10 53 16 28 19 23 40 39 24

1 38 12 27 6 33 7 13 19 16 23 22 17

9 30 13 26 21 18 20 12 7 21 18 29 10

4 24 8 20 12 16 5 18 5 15 13 20 8

P .246 .822 1

SOX2 is expressed in the developing foregut and lung [23] and maintains its expression in adult squamous cells of esophagus, trachea and airways tissues. It is also involved in stem cell differentiation in brain, retina and inner ear [24-26]. Cooperating with other factors such as OCT4, SOX2 is in fact able to reinduce pluripotency in terminally differentiated somatic cells [27,28]. Between hypopharynx, larynx, and sinonasal SCC, significantly different frequencies of copy number gains in the 3q26-27 region have been reported. Using multiple ligationdependent probe amplification (MLPA) and microarray

.265

.679 .679

Sinonasal (n = 51) P .533 .792 .46 .011

1 .788

−

+

3 41 4 40 34 10 16 8 20 8 36 39 5

1 6 1 6 4 3 2 2 3 1 6 7 0

P .457 .538 .352 .8

1 1

comparative genomic hybridization (CGH) analysis on sinonasal SCC, López et al [29] found 9% and 14% of gains at 3q, respectively, while a recent fluorescence in situ hybridization (FISH) study of various tumor types within the sinonasal cavities reported 38% of SOX2 copy number gains [30]. Reports on SCC of larynx, pharynx and oral cavity generally demonstrated frequencies up to 86% [2-4,6]. In this immunohistochemical study of SOX2 expression we found 38% to 42% of SOX2 expression in hypopharynx and larynx tumors, and a much lower frequency of 14% in sinonasal carcinoma. These numbers are lower than those found in

Fig. 3 Overall survival survival curves of 102 hypopharyngeal (A), 67 laryngeal (B) and 51 sinonasal (C) SCC patients according to SOX2 immunopositivity. Disease-free survival survival curves of 102 hypopharyngeal (D), 67 laryngeal (E) and 51 sinonasal (F) SCC patients according to SOX2 immunopositivity.

856 three previous studies on the same tumors, being 79% in hypopharynx and 55% in larynx [31,32]. Schröck et al showed a correlation between SOX2 expression and gene amplification (38%) but did not give an exact figure on protein expression [30]. The lower frequency in sinonasal SCC may be related to an important difference in etiology, because sinonasal SCC, in contrast to other HNSCC subsites, is not related to tobacco [29]. The majority of SOX2 expression studies, performed on many different tumor types, have indicated a relation to adverse clinical outcome [8,12,14,33,34]. Two studies, both on pulmonary SCC, claimed better prognosis for cases that showed SOX2 expression [17,18]. We did not find SOX2 expression related to clinical outcome for any of the three HNSCC subsites. In addition, there was no association to clinicopathological parameters known to be markers for survival, for example T-stage, N-stage, or local recurrence. In a large series of larynx SCC, Tang et al demonstrated a correlation between SOX2 expression and adverse clinical parameters as disease, T classification, lymph node metastasis and recurrence, as well as worse survival [32]. Two other studies on hypopharynx and sinonasal SCC however, did not find a prognostic value [30,31], which is in agreement with the present study. It is possible that these contradictory results are related to methodological differences in SOX2 immunopositivity scoring. Three approaches have been used in the literature: a 5% cutoff [17,21,22], a semiquantitative scoring of proportion and intensity and categorization in 2 or 3 classes [14,15,31-33], and an arbitrary evaluation of positivity by image processing software [18,30]. There may also be important clinical differences in the studied series of patients. For example, the proportion of cases with disease stage III and IV in our series was 93%, 81% and 92% for hypopharynx, larynx and sinonasal SCC, respectively, while other studies included 75% in hypopharynx [31] and 59% for larynx [32]. Even greater differences can be seen with regard to lymph node metastasis. Our series comprised cases with lymph node metastasis in 84% of hypopharynx, 51% larynx and 25% sinonasal SCC, considerably higher than the 22%, 25% and 18% respectively, published previously [30-32]. Another explanation may be that SOX2 copy number gain and expression are an early tumor-initiating event and therefore important for tumor development but not involved in conveying an aggressive or metastasizing phenotype [16]. SOX2 has been implicated in stem cell pluripotency and differentiation [27,28]. In the last years, there is an increasing interest in the cancer stem cell theory, attributing these cells a role in local recurrence. In addition, SOX2 has been proposed to play a role in the epithelium-mesenchymal transition in HNSCC, and therefore in invasion and metastases [35]. Our data on 3 different sublocalizations of HNSCC do not support a role for SOX2 in metastasis. However, we did find more frequent SOX2 expression in cases with poorer differentiation, possibly in line with a role of SOX2 in cancer stem cells.

R. González-Márquez et al. In conclusion, our results show that SOX2 expression is a common event in hypopharynx and larynx, but not in sinonasal SCC. In addition, SOX2 expression is not related to disease-free or overall survival, or to clinical indicators of poor outcome, such as lymph node metastasis, recurrences, or distant metastases.

Acknowledgments The authors thank Aitana Vallina and Eva Allonca for the tissue microarray preparation and the immunohistochemical experiments.

References [1] Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin 2005;55:74-108. [2] Hermsen M, Alonso Guervos M, Meijer G, et al. Chromosomal changes in relation to clinical outcome in larynx and pharynx squamous cell carcinoma. Cell Oncol 2005;27:191-8. [3] Jarvinen AK, Autio R, Kilpinen S, et al. High-resolution copy number and gene expression microarray analyses of head and neck squamous cell carcinoma cell lines of tongue and larynx. Genes Chromosomes Cancer 2008;47:500-9. [4] Mitelman F, Johansson B, Mertens F. Database of Chromosome Aberrations and Gene Fusions in Cancer. Available at http://cgap.nci. nih.gov/Chromosomes/Mitelman. Accessed on December 2, 2013. [5] Comtesse N, Keller A, Diesinger I, et al. Frequent expression of the genes FXR1, CLAPM1 and EIF4G located on amplicon 3q26-27 in squamous cell carcinoma of the lung. Int J Cancer 2007;120:2538-44. [6] Lin SC, Liu CJ, Ko SY, Chang HC, Liu TY, Chang KW. Copy number amplification of 3q26-27 oncogenes in microdissected oral squamous cell carcinoma and oral brushed samples from areca chewers. J Pathol 2005;206:417-22. [7] Heselmeyer K, Macville M, Schrock E, et al. Advanced-stage cervical carcinomas are defined by a recurrent pattern of chromosomal aberrations revealing high genetic instability and a consistent gain of chromosome arm 3q. Genes Chromosomes Cancer 1997;19:233-40. [8] Gen Y, Yasui K, Zen Y, et al. SOX2 identified as a target gene for the amplification at 3q26 that is frequently detected in esophageal squamous cell carcinoma. Cancer Genet Cytogenet 2010;202:82-93. [9] Hussenet T, Dali S, Exinger J, et al. SOX2 is an oncogene activated by recurrent 3q26.3 amplifications in human lung squamous cell carcinomas. PLoS One 2010;5:e8960. [10] Bass AJ, Watanabe H, Mermel CH, et al. SOX2 is an amplified lineage-survival oncogene in lung and esophageal squamous cell carcinomas. Nat Genet 2009;41:1238-42. [11] Rodriguez-Pinilla SM, Sarrio D, Moreno-Bueno G, et al. Sox2: a possible driver of the basal-like phenotype in sporadic breast cancer. Mod Pathol 2007;20:474-81. [12] Neumann J, Bahr F, Horst D, et al. SOX2 expression correlates with lymph-node metastases and distant spread in right-sided colon cancer. BMC Cancer 2011;11:518. [13] Zhang J, Chang DY, Mercado-Uribe I, Liu J. Sex-determining region Y-box 2 expression predicts poor prognosis in human ovarian carcinoma. HUM PATHOL 2012;43:1405-12. [14] Wang X, Liang Y, Chen Q, et al. Prognostic significance of SOX2 expression in nasopharyngeal carcinoma. Cancer Invest 2012;30:79-85. [15] Yuan P, Kadara H, Behrens C, et al. Sex determining region Y-Box 2 (SOX2) is a potential cell-lineage gene highly expressed in the

SOX2 in head and neck cancer

[16]

[17]

[18]

[19]

[20]

[21] [22]

[23]

[24] [25]

pathogenesis of squamous cell carcinomas of the lung. PLoS One 2010;5:e9112. Freier K, Knoepfle K, Flechtenmacher C, et al. Recurrent copy number gain of transcription factor SOX2 and corresponding high protein expression in oral squamous cell carcinoma. Genes Chromosomes Cancer 2010;49: 9-16. Brcic L, Sherer CK, Shuai Y, Hornick JL, Chirieac LR, Dacic S. Morphologic and clinicopathologic features of lung squamous cell carcinomas expressing Sox2. Am J Clin Pathol 2012;138:712-8. Wilbertz T, Wagner P, Petersen K, et al. SOX2 gene amplification and protein expression are associated with better outcome in squamous cell lung cancer. Mod Pathol 2011;24:944-53. Thompson L. World Health Organization classification of tumours: pathology and genetics of head and neck tumours. Ear Nose Throat J 2006;85:74. López F, Llorente JL, Oviedo CM, et al. Gene amplification and protein overexpression of EGFR and ERBB2 in sinonasal squamous cell carcinoma. Cancer 2012;118:1818-26. Lu Y, Futtner C, Rock JR, et al. Evidence that SOX2 expression is oncogenic in the lung. PLoS One 2010;5:e11022. Sholl LM, Barletta JA, Yeap BY, Chirieac LR, Hornick JL. Sox2 protein expression is an independent poor prognostic indicator in stage I lung adenocarcinoma. Am J Surg Pathol 2010;34:1193-8. Que J, Okubo T, Goldenring JR, et al. Multiple dose-dependent roles for Sox2 in the patterning and differentiation of anterior foregut endoderm. Development 2007;134:2521-31. Boyer LA, Lee TI, Cole MF, et al. Core transcriptional regulatory circuitry in human embryonic stem cells. Cell 2005;122:947-56. Kiernan AE, Pelling AL, Leung KK, et al. Sox2 is required for sensory organ development in the mammalian inner ear. Nature 2005;434:1031-5.

857 [26] Taranova OV, Magness ST, Fagan BM, et al. SOX2 is a dosedependent regulator of retinal neural progenitor competence. Genes Dev 2006;20:1187-202. [27] Masui S, Nakatake Y, Toyooka Y, et al. Pluripotency governed by Sox2 via regulation of Oct3/4 expression in mouse embryonic stem cells. Nat Cell Biol 2007;9:625-35. [28] Wang J, Rao S, Chu J, et al. A protein interaction network for pluripotency of embryonic stem cells. Nature 2006;444:364-8. [29] López F, Llorente JL, García-Inclán C, et al. Genomic profiling of sinonasal squamous cell carcinoma. Head Neck 2011;33:145-53. [30] Schröck A, Göke F, Wagner P, et al. Sex determining region Y-box 2 (SOX2) amplification is an independent indicator of disease recurrence in sinonasal cancer. PLoS One 2013;8:e59201. [31] Ge N, Lin HX, Xiao XS, et al. Prognostic significance of Oct4 and Sox2 expression in hypopharyngeal squamous cell carcinoma. J Transl Med 2010;8:94. [32] Tang XB, Shen XH, Li L, Zhang YF, Chen GQ. SOX2 overexpression correlates with poor prognosis in laryngeal squamous cell carcinoma. Auris Nasus Larynx 2013;40:481-6. [33] Du L, Yang Y, Xiao X, et al. Sox2 nuclear expression is closely associated with poor prognosis in patients with histologically nodenegative oral tongue squamous cell carcinoma. Oral Oncol 2011;47: 709-13. [34] Michifuri Y, Hirohashi Y, Torigoe T, et al. High expression of ALDH1 and SOX2 diffuse staining pattern of oral squamous cell carcinomas correlates to lymph node metastasis. Pathol Int 2012;62: 684-9. [35] Chen C, Wei Y, Hummel M, et al. Evidence for epithelialmesenchymal transition in cancer stem cells of head and neck squamous cell carcinoma. PLoS One 2011;6:e16466.