International Journal of Clinical Oncology

, Volume 19, Issue 2, pp 247–253

Prognostic significance of syndecan-1 expression in squamous cell carcinoma of the tonsil

Authors

  • Sung Hak Lee
    • Department of Hospital Pathology, College of MedicineThe Catholic University of Korea
  • Eun Ji Choi
    • Department of Pathology, College of MedicineThe Catholic University of Korea
  • Min Sik Kim
    • Department of Otorhinolaryngology, College of MedicineThe Catholic University of Korea
  • Jun Wook Park
    • Department of Otorhinolaryngology, College of MedicineThe Catholic University of Korea
    • Department of Hospital Pathology, College of MedicineThe Catholic University of Korea
    • Department of Pathology, College of MedicineThe Catholic University of Korea
  • Chang Suk Kang
    • Department of Hospital Pathology, College of MedicineThe Catholic University of Korea
Original Article

DOI: 10.1007/s10147-013-0552-7

Cite this article as:
Lee, S.H., Choi, E.J., Kim, M.S. et al. Int J Clin Oncol (2014) 19: 247. doi:10.1007/s10147-013-0552-7

Abstract

Background

Syndecan-1 (SDC1) is reported to modulate several key processes of tumorigenesis and to show variable expression in many cancers. The cause of these variations in expression is not known to date. In this study, we compared SDC1 status with clinicopathologic parameters to evaluate the prognostic implications of SDC1 status on squamous cell carcinoma (SCC) of the tonsil.

Methods

In 56 cases of tonsillar SCC, we screened SDC1 expression using immunohistochemistry and analyzed the relationships between SDC1 expression and clinicopathological parameters. To identify the cause of the changes in SDC1 expression seen in tumors, we measured the gene dosage of SDC1 in tumor cells using fluorescent in situ hybridization.

Results

SDC1 expression was found in cancer cells in 36 cases (64.3 %) of tonsillar SCC. It was associated with lymph node metastasis (p = 0.010) and a positive surgical resection margin (p = 0.014). On the other hand, it was not significantly correlated with sex, age, smoking status, degree of differentiation, T stage, or distant metastasis. We could not find any copy-number variation of SDC1 in the cases showing increased SDC1 immunopositivity. In addition, strong SDC1 expression in the tumor cells predicted a shorter overall survival (p = 0.020, log-rank).

Conclusions

We showed that SDC1 expression is associated with N stage and the status of resection margin involvement in SCC of the tonsil. With respect to survival, there were unfavorable outcomes in cases with SDC1 positivity. More studies are needed to better understand the role of SDC1 in the progression and invasiveness of tonsillar SCC.

Keywords

TonsilSquamous cell carcinomaSyndecan-1ExpressionPrognosis

Introduction

The syndecans are a major family of cell-surface heparan sulfate proteoglycans. Their ectodomains are thought to interact with components of the extracellular matrix (ECM), such as collagens and fibronectin [1]. They act as anchors, stabilizing the morphology of epithelial cells by connecting the ECM to the intracellular cytoskeleton, and function as co-receptors that catalyze the interactions between ligands and their respective signaling receptors [1]. There are four different kinds of syndecans in vertebrates, each encoded by distinct genes [2]. Among the syndecans, which are expressed in various types of cells, syndecan-1 (SDC1) is predominantly expressed in epithelial cells and plasma cells [3].

SDC1 is known to modulate several key processes associated with tumorigenesis, such as tumor cell proliferation, apoptosis, angiogenesis, and metastasis, in many cancers [1]. Several studies on the relationship between SDC1 expression and cancers have been conducted. Generally, membranous SDC1 expression is decreased in cancers such as squamous cell carcinoma (SCC) of the head and neck or lung [4, 5]. However, it has proven rather difficult to gauge the clinical significance of the amount of SDC1 expression in the cancer cells. Low SDC1 expression in tumor cells correlates with tumor invasiveness, metastatic potential, and a worse prognosis for head and neck carcinoma, squamous-cell lung cancer, and cervical cancer [4, 6, 7]. On the contrary, high SDC1 expression is associated with high-grade urothelial carcinoma and a high recurrence rate [8]. In studies of multiple myeloma, SDC1-negative cells were highly tumorigenic in one report, but failed to show tumor cell engraftment in other reports [9, 10]. In mice, however, SDC1 deficiency was found to have a protective effect against the development of mammary tumor, as induced by the ectopic expression of the proto-oncogene Wnt1 [11]. High levels of soluble SDC1—the secreted ectodomain of SDC1—in serum are associated with a poor prognosis in cases of lung cancer and multiple myeloma [12, 13]. In addition, SDC1 expression in cases of gallbladder cancer was associated with more frequent lymph node metastasis [14].

Head and neck SCC is the fifth most common cancer worldwide, with more than 600,000 cases diagnosed each year. Head and neck SCC includes SCC of the larynx, tongue, oral cavity, tonsil, and hypopharynx. Among these head and neck SCCs, tonsil cancer is the second most common malignant primary head and neck cancer [15, 16]. Alcohol consumption, cigarette smoking, human papillomavirus, and various genetic alterations are associated with SCC of the tonsil [17]. Recently, the overall incidence of head and neck cancer has fallen with the decrease in tobacco and alcohol consumption. However, the incidence of oropharyngeal tumors, including SCC of the tonsil, has increased in several countries [1619].

The clinical course of tonsillar SCC is known to be aggressive, with frequent locoregional recurrence and early tumor infiltration, although there have been remarkable advances in the management of tonsil cancer over the past few decades [20]. As with most other solid tumors, the TNM classification predicts prognosis in tonsillar SCC patients. However, it is difficult to determine the characteristics of tumors by TNM staging alone, and accurate stratification for treatment modality selection and prognosis prediction is likewise hard to achieve based only on the anatomic stage. Therefore, if a more accurate and reliable molecular marker for predicting prognosis were to be identified, cases with poor prognosis could be treated aggressively, which would increase treatment success rates in SCC of the tonsil [21].

In the study described in the present paper, we compared SDC1 expression with several clinicopathologic parameters to evaluate the clinicopathologic significance of the expression of SDC1 and prognostic implications of SDC1 expression in SCC of the tonsil. To identify the cause of changes in SDC1 expression in this tumor, we measured the dosage of the SDC1 gene using fluorescent in situ hybridization (FISH).

Materials and methods

Selection of patients and tumor samples

A total of 56 patients (51 men and 5 women) with SCC of the tonsil were enrolled in this study and underwent surgery at Seoul St. Mary’s Hospital during the period from 1994 to 2010. No patients received preoperative chemotherapy or radiation therapy. Clinicopathologic parameters were analyzed through a retrospective review of the medical records and pathologic reports at our medical institution. The patients’ ages ranged between 43 and 79 (mean 56.7) years. There were 17 (30.4 %) who had never smoked and 39 (69.6 %) who were light or heavy smokers. In addition, there were 7 cases (12.5 %) of well-differentiated carcinoma, 46 cases (82.1 %) of moderately differentiated carcinoma, and 3 cases (5.4 %) of poorly differentiated carcinoma. Tumor-specific survival data (median 47.5 months; range 0.5–155 months) were available. Survival was calculated from the date of surgery until death or June 30th, 2012. Of the 56 patients for whom follow-up data were available, 13 cases (23.2 %) had a locoregional recurrence and 7 cases (12.5 %) had a distant metastasis. All specimens were collected from the patients according to the protocols (KC12SISI0121 and MC12SIMI0109) approved by the Institutional Review Board of the Catholic University of Korea.

Tissue microarray (TMA) construction and immunohistochemistry

Tissue microarrays were constructed from paraffin-embedded blocks after reviewing glass slides for the 56 cases of tonsillar SCC using a manual tissue arrayer (Beecher Instruments, Sun Prairie, WI, USA) with 2.0 mm punches. The TMAs were sectioned at a thickness of 4 μm. Sections from the TMA blocks were transferred to Probe On Plus slides (Fisher Scientific, Pittsburgh, PA, USA) and incubated for 2 h in a chamber at 56 °C (Agilent Technologies, Santa Clara, CA, USA). The sections were deparaffinized in xylene three times and rehydrated through 100, 90, 80, and 70 % ethanol as well as Tris-buffered saline (TBS, pH 7.4). The tissues were then boiled in 10 mM sodium citrate buffer (pH 6.0) using a microwave oven for 20 min. After treating the tissues with 3 % H2O2 in phosphate-buffered saline (PBS), the tissues were incubated with diluted (1:50) mouse monoclonal antibody B-A38 to SDC1 (Abcam, Cambridge, UK) at 4 °C overnight. Having incubated the tissue with diluted (1:100) biotinylated anti-mouse antibody (Abnova, Walnut, CA, USA) for an hour at room temperature, the signal was amplified using a TSA HRP system (PerkinElmer, Waltham, MA, USA). A liquid DAB + substrate chromogen system (Dako, Glostrup, Denmark) was used for visualization.

Membranous or cytoplasmic staining in cancer cells was considered positive. The immunoreactivity of SDC1 was scored by adding the staining intensity level (0, no stain; 1, weak to moderate; 2, strong) to the points assigned based on the percentage of stained tumor cells present (0, no stain; 1, 1–50 %; 2, >50 %). Total scores of three or more were considered positive. Two pathologists scored the immunoreactivity independently.

FISH

The copy number of SDC1 was measured by FISH, as done previously [7]. In short, an SDC1 FISH probe was made using a BioPrime Array CGH Genomic Labeling Module (Invitrogen, Carlsbad, CA, USA), BAC clone (RP11-202B22; Invitrogen), and Spectrum Orange-dUTP (Abbott Molecular, Abbott Park, IL, USA). An Aquarius Satellite enumeration probe of chromosome 2 (Cytocell, Cambridge, UK) was purchased as a reference probe. The hybridization location of the homemade SDC1 probe was verified using a metaphase spread of normal peripheral mononuclear cells (data not shown). Tissue processing and hybridization was done using Paraffin Pretreatment Kit I (Abbott Molecular) and ThermoBrite (Abbott Molecular). We counted the number of fluorescent spots in at least 100 nuclei of tumor cells in each case.

Statistical analysis

The Chi-square test or Fisher’s exact test was performed to evaluate associations between SDC1 expression and various clinicopathologic parameters. We used Cox’s multivariate proportional hazards model to determine the relative importance of the clinicopathologic parameters that show significant associations with SDC1 expression. The Kaplan–Meier method was used for survival analysis. To test for differences between the survival curves of the groups, we used the nonparametric log-rank test. A p value of less than 0.05 was considered statistically significant. We used R v.2.10 (R Foundation, Vienna, Austria) for statistical calculations and to produce graphs.

Results

Correlations between SDC1 expression and clinicopathological parameters

Normal squamous epithelial cells showed a membranous pattern of SDC1 expression. A membranous or cytoplasmic pattern of SDC1 expression was found in the cancer cells from 36 cases (64.3 %) of tonsillar SCC (Fig. 1). We compared the status of SDC1 expression with various clinicopathologic parameters. SDC1 expression was associated with lymph node metastasis (p = 0.010) and a positive surgical resection margin (p = 0.014). On the contrary, immunoreactivity of SDC1 was not significantly correlated with HPV status, sex, age, smoking status, degree of differentiation, or T stage (Table 1). In multivariate analysis, we found that M stage and SDC1 expression were independent prognostic factors for survival (Table 2). HPV infection had a beneficial effect according to multivariate analysis, and this finding is concordant with that in a previous report [22].
https://static-content.springer.com/image/art%3A10.1007%2Fs10147-013-0552-7/MediaObjects/10147_2013_552_Fig1_HTML.jpg
Fig. 1

Representative syndecan-1 immunohistochemical staining in a normal tonsillar epithelium and tonsillar SCC, showing b membranous and c cytoplasmic staining patterns (×400). d Focal immunoreactivity is seen in the stromal spindle cell component among tumor nests

Table 1

Relationships between SDC1 expression and clinicopathologic parameters of tonsillar SCC

Parameters

Immunohistochemistry of SDC1

p value

Positive (n = 36)

Negative (n = 20)

Sex

0.336

 Male

34 (60.7)

17 (30.4)

 

 Female

2 (3.6)

3 (5.4)

 

Age

 ≤55 years

18 (32.1)

10 (17.9)

1.000

 >55 years

18 (32.1)

10 (17.9)

 

Smoking history

0.386

 Present

27 (48.2)

12 (21.4)

 

 Absent

9 (16.1)

8 (14.3)

 

HPV infection (NA: 6)a

0.953

 Positive

12 (24.0)

7 (14.0)

 

 Negative

18 (36.0)

13 (26.0)

 

Resection margin involvement

0.031*

 Present

14 (25.0)

2 (3.6)

 

 Absent

22 (39.3)

18 (32.1)

 

Tumor size

0.728

 T1–2

28 (50.0)

17 (30.4)

 

 T3–4

8 (14.3)

3 (5.4)

 

N stage

0.045*

 0–1

12 (21.4)

13 (23.2)

 

 2–3

24 (42.9)

7 (12.5)

 

Recurrence

0.466

 Present

7 (12.5)

2 (3.57)

 

 Absent

29 (51.8)

18 (32.1)

 

Values are presented as number (%)

* Statistically significant

aHPV status was not available in 6 cases

Table 2

Results of a Cox’s multivariate proportional hazards regression model for SCC of the tonsil

Factor

Hazard ratio

95 % Confidence interval

Cox’s test p value

Metastasis

7.6

1.9–31.0

0.004*

SDC1 expression

5.3

1.3–21.5

0.020*

HPV infection

0.2

0.1–0.9

0.031*

Nodal status

2.4

0.7–8.6

0.189

Tumor size

0.5

0.1–2.3

0.357

Smoking

0.8

0.2–3.4

0.759

Margin involvement

0.9

0.2–3.8

0.834

Grade

0.7

0.1–4.1

0.654

* Statistically significant

Copy-number status of SDC1

We could not find any copy-number variation of SDC1 in the cases showing increased SDC1 immunopositivity (Fig. 2).
https://static-content.springer.com/image/art%3A10.1007%2Fs10147-013-0552-7/MediaObjects/10147_2013_552_Fig2_HTML.jpg
Fig. 2

Quantification of the syndecan-1 (SDC1) gene in SCC of the tonsil. FISH results for SDC1 (red dots) and centromere of chromosome 2 (green dots). There is no copy-number variation of SDC1 in tumor tissue

Survival analysis

We analyzed patient survival for the groups that showed different levels of SDC1 expression. Kaplan–Meier analysis indicated that SDC1-positive expression (p = 0.034, log-rank) was a predictor of shorter overall survival (Fig. 3).
https://static-content.springer.com/image/art%3A10.1007%2Fs10147-013-0552-7/MediaObjects/10147_2013_552_Fig3_HTML.gif
Fig. 3

Survival analysis of the groups with different levels of SDC1 expression. A statistically significant difference was noted between the positive group and the negative group (p = 0.034)

Discussion

SDC1 expression is frequently altered in human carcinomas, but the function of SDC1 in tumorigenesis is not clearly understood. In some human cancers, this molecule is downregulated, whereas increased expression is noted in others [1, 23].

To date, studies of head and neck SCC have suggested a correlation between loss of SDC1 expression and the process of malignant transformation. Decreased SDC-1 expression in tumor cells was correlated with tumor invasiveness, metastatic potential, and a worse prognosis [4]. In addition, there was a correlation between reduced SDC1 expression and histological dedifferentiation, with increased syndecan-1 expression in well-differentiated tumors [24]. However, cases of tonsillar SCC comprise <10 % of the total cases investigated in their studies [4, 24], so we cannot reliably predict the nature of tonsillar SCC based on the results for head and neck SCC. Besides, tonsillar cancer is a different disease entity from other head and neck SCCs. Because tonsillar cancer is associated with HPV infection, and HPV infection favors survival in tonsillar cancer cases, tonsillar cancer is expected to have a distinct disease progression pattern from other SCCs of the head and neck. Moreover, the incidence of HPV-associated tonsillar SCC has greatly increased over the last decade, so the effect of HPV on tonsillar SCC has increased accordingly [22, 25, 26]. As far as we know, no study aimed at determining the clinical impact of SDC1 expression in cases of the SCC of the tonsil has been performed, so we designed the present study to identify the clinical implications of SDC1 expression in tonsillar SCC.

In contrast to other SCCs originating from the head and neck, we observed that strong SDC1 staining in malignant cells within SCC of the tonsil was associated with aggressive clinical behavior. This finding contrasts sharply with those from other studies on head and neck SCC [4, 24, 27]. This finding suggests that tonsillar SCC has a different patient survival pattern from other head and neck SCCs. Why the prognosis of patients with tonsillar SCC based on SDC1 expression is completely opposite to those of other head and neck SCCs remains to be elucidated. Also, the contributions of HPV and SDC1 expression to patient survival is not known.

The regulation of SDC1 expression is not clearly documented. One possible mechanism regulating SDC1 expression is the shedding of the SDC1 ectodomain. The removal of heparan sulfate chains from SDC1 causes ectodomain shedding of SDC1 and increases the expression of SDC1 [28]. This is why we measured the intensity of SDC1 expression in addition to the propensity for SDC1 expression in the tumor tissues. A single scoring scheme of SDC1 staining may not accurately gauge the total amount of both forms of SDC1 protein.

According to Vered et al., SDC1 expression is reportedly lost in the majority of tongue SCC cases, which contrasts with its ubiquitous expression in adjacent benign squamous mucosa, and decreased expression of SDC1 is associated with downregulation of intercellular adherence and acquisition of mobility. However, when assessing the immunohistochemical staining results, they only considered membranous staining in cancer cells as indicating positive reactivity, unlike in the current study [29]. Because we took into consideration SDC1 in abnormal locations in tumor cells, we scored the cytoplasmic expression of SDC1 as well as membranous staining. Using this scoring scheme, we found that strong SDC1 expression was associated with poor prognosis. This may partly explain the difference between our results and previous reports. Also, tonsillar SCC may have different mechanism of tumorigenesis than tongue SCC.

SDC1 is a protein that participates in the regulation of cell migration, cell–cell and cell–matrix interactions, growth factors linked to heparin, and in the modulation of protease activity [1]. Therefore, it may have an effect on tumor cell invasion and the progression of tonsil cancers through these mechanisms. According to the studies of Mennerich et al. and Ito et al. [23, 30], SDC1 expression in stromal cells is partly associated with a decrease in the SDC1 expression of tumor epithelial cells. After it is released by epithelial cells, SDC1 may be trapped by stromal cells, which then creates a favorable microenvironment for accelerated tumor cell growth and angiogenesis, thus contributing to invasion and distant metastasis [31]. Similarly, Gotte et al. [32] revealed that overexpression of SDC1 protein in breast cancer cells was correlated with decreased stromal protein expression.

In the present study, we showed that the normal membranous distribution of SDC1 vanished in the tumor cells, and increased amounts of SDC1 are located in the cytoplasm of tumor cells with diminished membranous staining. Due to the transposition of SDC1 from the cell membrane to cytoplasm, the tumor cells may have insufficient functional SDC1 on their surfaces. The cells may therefore interact less with the extracellular matrix and move more freely than normal due to a lack of functional SDC1. This may facilitate the spread and motility of tumor cells and lead to invasion and distant metastasis. This loss of functional SDC1 may explain the significant correlation with resection margin involvement by infiltrative growth pattern and status of lymph node metastasis in the present study. However, strong immunoreactivity does not necessarily mean a large amount of functional SDC1. To understand the cellular effects of altering the levels of cytoplasmic and membranous SDC1, further study is required.

Increased expression of SDC1 can be ascribed to a gene mutation or an alteration of the regulatory mechanism without any change in the gene itself. As a mechanism of increased SDC1, Roh et al. revealed that marked overexpression of fascin was correlated with positive SDC1 expression in patients with gallbladder cancer [33]. Sometimes changes in the copy number of a gene result in alterations in gene expression. In an attempt to elucidate the possible mechanism of SDC1 expression, we screened the copy number of SDC1 by FISH and found no variations in its copy number. It seems that the alterations in SDC1 expression are caused by mechanisms other than changes in gene dosage.

In conclusion, we showed that SDC1 expression in epithelial neoplastic cells is associated with N stage, histological grade, and status of resection margin involvement in SCC of the tonsil. With respect to survival, there were unfavorable outcomes in cases with SDC1 positivity. The altered SDC1 expression observed in cases of tonsillar SCC is not caused by changes in the copy number of SDC1.

Acknowledgments

This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF), as funded by the Ministry of Education, Science and Technology (2009-0087026).

Conflict of interest

The authors declare that they have no conflict of interest.

Copyright information

© Japan Society of Clinical Oncology 2013