SEARCH

SEARCH BY CITATION

Keywords:

  • squamous cell carcinoma;
  • Verrucous carcinoma;
  • E-cadherin;
  • oral cancer;
  • methylation

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Background.

This study aims to compare the alterations in the methylation profiles of E-cadherin in oral cancer, especially in tumors with lowest metatastic potential.

Methods.

Nine oral verrucous carcinomas (VCs), 20 oral well-differentiated squamous cell carcinomas without lymph node involvement (SCC-pN0), and 17 with lymph node involvement (SCC-pN+) were analyzed using methylation-specific polymerase chain reaction and immunohistochemical expression of E-cadherin gene.

Results.

The immunohistochemical expression of E-cadherin in VC was significantly higher (p = .016) when compared with SCC-pN0 and SCC-pN+ groups. The E-cadherin gene methylation was not correlated with its abnormal immunohistochemical expression in VC and SCC-pN0. All tumors of the SCC-pN+ group with unmethylated E-cadherin gene showed significant loss of E-cadherin immunoexpression (p = .044).

Conclusions.

The E-cadherin gene methylation presence in tumors with lowest invasive and metastatic potential, such as VC, suggests the early involvement of this epigenetic event in the multistep progression of the oral carcinogenesis. © 2007 Wiley Periodicals, Inc. Head Neck, 2008

Abnormalities of cell adhesion molecules such as downregulation of E-cadherin are associated with tumor invasion, metastasis, and, consequently, poor prognosis in patients with oral squamous cell carcinoma (SCC).1–4 The proposed mechanisms responsible for E-cadherin loss in oral SCC include deletions, mutations, and CpG island hypermethylation of the E-cadherin gene.1–4

Mutations of the CDH1 gene encoding E-cadherin are rare or absent in oral cancer, and CDH1 gene mutations that compromise the nonadhesive function of E-cadherin have been observed only in human gastric carcinoma cell lines, lobular breast cancer, and familial gastric cancer.4, 5 In addition, according to the authors,4, 5 the loss of E-cadherin expression is heterogeneous in cancers and may be modulated by the tumor microenvironment without involving irreversible genetic alterations.

The CpG island hypermethylation is a potential epigenetic means of inactivating the E-cadherin gene contributing to the dynamic, phenotypic heterogeneity, which drives metastatic progression of the oral SCC.4–6

To date, several studies have demonstrated that the E-cadherin gene is hypermethylated and silenced in oral cancer.6–11 However, most of these studies have focused on the E-cadherin loss in highly invasive and metastatic SCCs.6, 7, 9, 11 On the other hand, it has been suggested that reduced E-cadherin expression resulting from CpG methylation promoter occurs early during malignant progression, prior to the oral cancer tumor invasion.11 If the CpG island hypermethylation of the E-cadherin gene contributes to the development of oral cancer, then this methylation change might be found in nonmetastatic oral SCC as verrucous carcinoma (VC).

In the present study, to address the changes in the methylation profiles of E-cadherin in oral cancer, especially in nonmetastatic tumor, we investigated the CpG methylation of the E-cadherin gene promoter and its immunohistochemical expression in oral VC and oral well-differentiated SCC with (SCC-pN+) and without (SCC-pN0) lymph node involvement.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Patients and Tissues

This study was based on the analysis of 46 patients who underwent surgical treatment for primary oral SCC at the Head and Neck Surgery and Otorhinolaryngology Department of the A.C. Camargo Cancer Hospital, São Paulo, Brazil, from 1980 to 2000. Nine oral VC, 20 oral well-differentiated SCC without lymph node involvement (SCC-pN0), and 17 with lymph node involvement (SCC-pN+), confirmed by histopathological evaluation, were selected for analysis of the E-cadherin gene methylation profile and its immunohistochemical expression. The inclusion criteria were: (1) primary oral SCC located in the oral tongue, floor of the mouth, retromolar area, inferior or superior gingiva, soft or hard palate, buccal mucosa, and inferior or superior lip, confirmed by biopsy; (2) patients who did not undergo radiotherapy, chemotherapy, or other treatment prior to surgery; and (3) tumor tissue available for microscopic analysis. The exclusion criteria were: (1) patients with other simultaneous primary tumors; (2) unresectable tumors; (3) distant metastases at the time of admission; and (4) patients who refused surgical treatment. Clinical data of the patients with oral cancer were obtained from the medical records and included age, sex, ethnic group, tobacco and alcohol consumption, tumor location, extension, and infiltration pattern (ulcero-infiltrative or exophytic), T and N classification, treatment and clinical follow-up (recurrence, second primary tumor, and death).

Formalin-fixed and paraffin-embedded tumor tissues were taken from the Department of Pathology archive of A.C. Camargo Cancer Hospital, São Paulo, Brazil. The specimens of oral VC and well-differentiated SCC, embedded in paraffin, were stained for hematoxylin and eosin for the determination of the most representative highly invasive clones and largest epithelial malignant cells areas to the DNA extraction.

Three American Type Culture Collection (ATCC) cell lines were used for standardization and control of the polymerase chain reaction (PCR). The cell lines H 1299 and HCT 116 were used as a positive and MCF-7 as a negative control for E-cadherin promoter methylation, respectively. In addition, DNA was also extracted from 5 oral fibrous hyperplasias in order to analyze the methylation profile of the E-cadherin promoter in nonmalignant tissues.

E-Cadherin Promoter Methylation in Oral Squamous Cell Carcinomas.

The analysis of the E-cadherin promoter methylation pattern was undertaken at the Laboratory of Cancer Genetics of the Ludwig Institute for Cancer Research, São Paulo, Brazil. DNA from the cell lines was extracted by phenol and chloroform technique or alternatively the Perfect gDNA Blood Mini kit (Eppendorf, Westbury, NY) was used, following the manufacturer's instructions. DNA from MCF-7 was used as negative control for methylation (PMID: 9581841 and/or PMID: 12553040); H1299 cell line (PMID: 15126351) was used as a positive control for methylation. Genomic DNAs of the oral paraffin-embedded samples were extracted using the Nucleon HT kit (Amersham Biosciences, Pittsburgh, PA). The modification with sodium bisulfite of the DNA samples and the methylation-specific PCRs were performed as previously described.12 The sets of primers, described by Zöchbauer-Müller et al,13 were E-cadherin M-sense (5′-TTAGGTTAGAG GGTTATCGCGT-3′) and E-cadherin M-antisense (5′-TAACTAAAAATTCACCTACCGAC-3′) for the methylated sequence, and E-cadherin U-sense (5′-TAATTTTAGGTTAGAGGGTTATT-GT-3′) and E-cadherin U-antisense (5′-CACAA-CCAATCAAC AACACA-3′) for the unmethylated sequence. Briefly, 20 μL of DNA genomic solution was denatured by treatment with NaOH and modified by sodium bisulfite. DNA samples were then purified using the Wizard DNA Clean-up System kit (Promega, Madison, WI), according to the manufacturer's instructions and eluted into 45 μL water. Modification was completed by NaOH (final concentration 0.3M) treatment for 10 minutes at room temperature, followed by ethanol precipitation, and the DNA was resuspended in 25 μL water. PCR was used to amplify the modified DNAs. The PCR reaction was performed in a total volume of 25 μL containing 0.2 mM dNTPs (Invitrogen, Brazil), 2.0 mM MgCl2 (Invitrogen), 0.4 μM each primer, 5% DMSO, 1X Platinum Taq DNA polymerase buffer (Invitrogen) and 1.5u Platinum Taq DNA polymerase (Invitrogen)

The amplification involved 3 stages in which the annealing temperature was higher in the first 10 cycles and reduced in 2 degrees in the following stage (10 cycles) and other 2 degrees in the last 15 cycles. Different annealing temperatures are used for amplifying the methylated and unmethylated sequences. After activation of the Platinum Taq DNA polymerase at 94°C for 5 minutes, PCR was performed in a thermal cycle for 10 cycles consisting of denaturation at 94°C for 30 seconds, annealing at 64°C (E-cadherin M-primer set) or 62°C (E-cadherin U-primer set) for 1 minute, and extension at 72°C for 1 minute followed by 10 cycles at 94°C for 30 seconds, 62°C (E-cadherin M-primer set) or 60°C (E-cadherin U-primer set) for 1 minute and 72°C for 1 minute and 15 cycles at 94°C for 30 seconds, 60°C (E-cadherin M-primer set) or 58°C (E-cadherin U-primer set) for 1 minute and 72°C for 1 minute followed by a final 7-minute extension. Controls without DNA were carried out for each set of reaction. PCR products were loaded onto 8% polyacrylamide gels, stained with silver and visualized by UV illumination. The predicted sizes of the PCR product were 115 and 97 bp for methylated and unmethylated sequences, respectively.

E-Cadherin Immunoexpression in Oral Squamous Cell Carcinomas

Formalin-fixed 5-μm sections were used for immunohistochemical analysis of the E-cadherin antibody (Clone 36, Transduction Laboratories ref C20820, Lexington, KY). After antigen retrieval using 10 mM citrate buffer, pH 6.0, in a pressure cooker during 4 minutes, endogenous peroxidase activity was blocked by incubation in 3.0% hydrogen peroxide for 20 minutes. The sections were incubated overnight at 4°C with monoclonal E-cadherin primary antibody (dilution 1:700), in phosphate-buffered saline with bovine serum albumin to block nonspecific reaction. The antigen–antibody reaction was detected using streptavidin–biotin-based detection kit (StreptABComplex HRP; Duet, Mouse/Rabbit, Dako, ref K0492, Denmark) and visualized using 3.3′-diaminobenzidine tetrahydrochroride (DAB/Sigma, ref D-5637, USA). Sections were counterstained with Harris's hematoxylin before being dehydrated and cover-slipped. Dermatofibroma was used as external positive control and normal oral squamous epithelium when present in the tumoral section served as internal positive control. Negative control was prepared by omission of the primary antibody.

Quantitative computer-assisted analysis of 6 invasive front tumor fields (206.866,98 μm2) was performed to evaluate the complete membranous staining of E-cadherin in malignant epithelial cells. The positive tumors cells in each VC, SCC-pN0, and SCC-pN+ were counted in a 400× field, using a Sony Cyber-Shot camera attached to a light microscope (Axioskop 4.0, Zeiss) and recorded using Axiovision computer program system. A minimum of 500 cells was counted, and the percentage of positive malignant epithelial cells for E-cadherin membranous staining in the tumoral groups was determined. Based on the average value of the E-cadherin expression in epithelial malignant cells obtained from the oral SCC groups (VC, SCC-pN0, and SCC-pN+), the immunostaining scores were established, and the median value was shown to be best choice to have a group with 2 categories of score: score ≤36.5% = negative or weak complete membranous staining of E-cadherin; score >36.5% = positive complete membranous staining of E-cadherin. Because the small sample size of 46 cases, the median was shown to be the cut-off with better distribution between groups than other quartiles giving some statistical power to the applied tests.

Statistical Analysis

All statistical analyses were performed using the STATA 7.0 software (StataCorp., College Station, TX). To verify the association among the studied groups regarding the categorical variables, the chi-square test was used. Whenever at least 1 expected frequency was less than 5 in 2 × 2 tables, the Fisher exact test was used. p-values less than .05 were considered statistically significant. Fisher's exact test was performed to verify the correlation between the E-cadherin gene methylation status and their immunohistochemical expression in the oral SCC groups. The survival probabilities were estimated using the Kaplan–Meier method, and to compare survival curves, the log-rank test was used. The follow-up period was the time between the surgery date and the death or last patient information date.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

The main details concerning clinical features of our series of 46 patients with oral VC, SCC-pN0, and SCC-pN+ are summarized in Table 1.

Table 1. Oral carcinomas: clinical findings, treatment and patients outcome
Clinical featuresNo. of patients (%) by histologic group
VCSCC-pN0SCC-pN+
  1. Abbreviations: VC, verrucous carcinoma; SCC-pN0, well-differentiated squamous cell carcinoma without lymph node involvement; SCC-pN+, well-differentiated squamous cell carcinoma with lymph node involvement; NED, no evidence of disease; RD, recurrent disease; NRD, not related to disease.

Sex   
 Male6 (67)11 (55)12 (71)
 Female3 (33)9 (45)5 (29)
Ethnic group   
 White8 (89)18 (90)13 (76)
 Non-White1 (11)2 (10)4 (24)
Age   
 ≤58 years3 (30)8 (40)12 (71)
 >58 years6 (70)0 (0)0 (0)
Gum2 (22)8 (40)6 (35)
Palate2 (22)5 (25)0 (0)
Floor mouth0 (0)2 (10)10 (59)
Buccal mucosa1 (11)5 (25)1 (6)
Tobacco   
 Yes5 (56)11 (55)14 (82)
 No3 (33)8 (40)3 (18)
 Ignored1 (11)1 (5)0 (0)
Alcohol   
 Yes8 (89)10 (50)12 (71)
 No0 (0)9 (45)5 (29)
 Ignored1 (11)1 (5)0 (0)
T classification, clinical   
 T1-29 (100)10 (50)8 (47)
 T3-40 (0)10 (50)9 (53)
Neck dissection   
 Ipsilateral0 (0)7 (35)8 (47)
 Bilateral0 (0)1 (5)9 (53)
 No neck dissection9 (100)12 (60)0 (0)
Radiotherapy   
 Yes0 (0)7 (35)13 (76)
 No9 (100)13 (65)4 (24)
Recurrence   
 Yes3 (33)4 (20)8 (47)
 No6 (67)16 (80)9 (53)
Clinical outcome   
 Alive, NED4 (45)10 (50)2 (11)
 Dead of disease0 (0)5 (25)12 (71)
 Dead, NRD2 (22)5 (25)3 (18)
 Lost follow-up3 (33)0 (0)0 (0)
Total9 (100)20 (100)17 (100)

There was a predominance of white men in the 3 groups of oral well-differentiated SCC. The age of patients ranged from 41 to 83 years (mean, 67.8 years) for those with VC, from 28 to 82 years (mean, 57.1 years) for those with SCC-pN0, and from 35 to 73 years (mean, 53.7 years) for those with SCC-pN+. Clinically, the VCs occurred mainly in the lower lip. The gum and the floor of mouth were more affected by SCCs without (SCC-pN0) or with (SCC-pN+) lymph node involvement, respectively. Tobacco use and alcohol consumption were reported by most of the patients with oral VC, SCC-pN0, and SCC-pN+. Based on the Union Internationale Contre le Cancer (UICC) classification of oral cavity carcinomas, tumors were classified as clinical stage I–II (100% VC, 50% SCC-pN0, 29.4% SCC-pN+) and III–IV (50% SCC-pN0, 70.6% SCC-pN+). Regarding treatment and clinical course, only the patients of the SCC-pN0 and SCC-pN+ groups underwent surgical treatment with simultaneous ipsilateral or contralateral neck dissection and received adjuvant radiotherapy. Local recurrence was detected in the 3 oral SCC groups, but regional recurrence was more prevalent in SCC-pN+. A second primary tumor occurred in 1 patient of the VC group, in 4 patients of the SCC-pN0, and in 2 patients of the SCC-pN+ group. The clinical outcome for patients with oral VC, SCC-pN0, and SCC-pN+ is shown in Table 1. The 5- and 10-year overall survival rates were, respectively, 87.5% and 75.6% for patients with oral VC; 85% and 55% for patients with SCC-pN0, and 35.2% and 29.1% for patients with SCC-pN+. As shown in Figure 1, a higher significant difference among 5- and 10-year overall survival rates was detected in VC and SCC-pN0 goups when compared with SCC-pN+ group (p = .012). The survival curves of E-cadherin expression with scores ≤36.5% and >36.5% did not show statistically differences (log-rank p-value = .7515) for disease-free survival (log-rank p-value = .7063).

thumbnail image

Figure 1. Cumulative overall survival probability by oral cancer groups. VC, verrucous carcinoma; SCC-pN0, well-differentiated squamous cell carcinoma without lymph node involvement; SCC-pN+, well-differentiated squamous cell carcinoma with lymph node involvement. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Download figure to PowerPoint

E-Cadherin Gene Hypermethylation in the Oral Squamous Cell Carcinomas

The hypermethylation of the E-cadherin gene was statistically similar (p = .975) and superior to 50% in the 3 groups of oral carcinomas (VC, SCC-pN0, and SCC-pN+), as described in the Table 2.

Table 2. E-cadherin gene methylation status in oral squamous cell carcinomas
E-cadherinNo. of patients (%) by histologic groupp*
VCSCC-pN0SCC-pN+
  • Abbreviations: VC, verrucous carcinoma; SCC-pN0, well-differentiated squamous cell carcinoma without lymph node involvement; SCC-pN+, Well-differentiated squamous cell carcinoma with lymph node involvement.

  • *

    p value obtained by chi-square test.

Methylated5 (55.5)12 (60)10 (59).975
Unmethylated4 (44.5)8 (40)7 (41) 
Total9 (100)20 (100)17 (100) 

The oral fibrous hyperplasias used as control showed the unmethylated E-cadherin gene (Figure 2). CpG hypermethylation of E-cadherin gene was detected in most patients with tobacco or alcohol consumption and oral SCCs as well. Interestingly, we found that of the 8 recurrent SCCs with lymph node involvement (SCC-pN+), 6 (75%) showed hypermethylation of the E-cadherin gene.

thumbnail image

Figure 2. Representative results of methylation-specific PCRs of E-cadherin promoter in samples of oral VC, SCC-pN0, SCC-pN+, FH. Lanes U, presence of a visible PCR product indicates the presence of unmethylated E-cadherin(97 bp); lanes M, presence of product indicates the presence of methylated E-cadherin (115 bp). DNA from MCF-7 was used as negative control for methylation; H1299 cell line was used as a positive control for methylation; MM, 100 bp molecular weight ladder; No, control without DNA. VC, verrucous carcinoma; SCC-pN0, well-differentiated squamous cell carcinoma without lymph node involvement; SCC-pN+, well-differentiated squamous cell carcinoma with lymph node involvement; FH, oral fibrous hyperplasia.

Download figure to PowerPoint

E-Cadherin Immunoreactivity in Oral Squamous Cell Carcinomas

A complete membranous staining for E-cadherin (>36.5% E-cadherin score) in oral VC group (89%) was significantly higher (p = .016) when compared with the SCC-pN0 group (50%) and SCC-pN+ group (29%). For the SCC-pN0 group, an equivalent distribution of the E-cadherin expression was detected as showed in the Table 3.

Table 3. E-cadherin imunoexpression in oral squamous cell carcinomas
E-cadherinNo. of patients (%) by histologic groupp*
VCSCC-N0SCC-pN+
  • Abbreviations: VC, verrucous carcinoma; SCC-pN0, well-differentiated squamous cell carcinoma without lymph node involvement; SCC-pN+, well-differentiated squamous cell carcinoma with lymph node involvement; CMS, complete epithelial membranous staining for E-cadherin.

  • *

    p value obtained by chi-square test.

≤36.5% CMS1 (11)10 (50)12 (71).016
>36.5% CMS8 (89)10 (50)5 (29) 
Total9 (100)20 (100)17 (100) 

Most (71%) of the well-differentiated SCCs with lymph node involvement (SCC-pN+) showed reduced E-cadherin expression, with scores ≤36.5% of complete epithelial membranous staining. Details concerning E-cadherin immunoexpression in VC and SCC are summarized in Figure 3.

thumbnail image

Figure 3. E-cadherin expression in oral VC (A) and well-differentiated SCC-pN+ (B), the last one showing reduced of the complete membranous staining for E-cadherin in some epithelial cells of the invasive front tumor. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Download figure to PowerPoint

Correlation between CpG Methylation of E-Cadherin Gene Promoter and E-Cadherin Expression in Oral Squamous Cell Carcinomas

No statistically significant correlation was obtained between the E-cadherin gene methylation status (methylated or unmethylated) and its immunohistochemical expression (positive, negative, or reduced E-cadherin expression) for the oral VC and for the oral SCC-pN0 (Table 4).

Table 4. Correlation between E-cadherin gene methylation status and its immunohistochemical expression in oral squamous cell carcinomas
IHC (CMS)E-cadherin gene hypermethylation
VCSCC-pN0SCC-pN+
+p+p*+p*
  • Abbreviations: VC, verrucous carcinoma; SCC-pN0, well-differentiated squamous cell carcinoma without lymph node involvement; SCC-pN+, well- differentiated squamous cell carcinoma with lymph node involvement; IHC, immunohistochemistry; CMS, complete epithelial membranous staining for E-cadherin; NA: not available; (+): E-cadherin gene methylated; (): E-cadherin gene unmethylated.

  • Score ≤36.5% = negative or weak complete membranous staining of E-cadherin; score >36.5% = positive complete membranous staining of E-cadherin.

  • *

    p value obtained by Fisher exact test.

≤36.5%10NA82.17057.044
>36.5%44 4650
Total54 128 107 

On the other hand, a statistically significant (p = .044) correlation was obtained between oral SCC-pN+ that showed E-cadherin gene unmethylated and loss of E-cadherin immunoexpression (≤36.5% of the complete epithelial membranous staining), as described in Table 4.

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Oral cancers are heterogeneous in their histology and clinical behavior. Reduced cell–cell adhesiveness of molecules such as E-cadherin is considered indispensable for both the early and the late oral carcinogenesis steps. Human cancers, including oral SCC, appear to possess irreversible as well as reversible mechanisms for inactivating the cell–cell adhesion system.14 Promoter hypermethylation is an important mechanism for inactivation of tumor suppressor genes as E-cadherin in cancer cells.6–9, 11, 15 In our study, the methylation frequency of the E-cadherin gene varied from 55.5% for VC, 60% for SCC-pN0, and 59% for SCC-pN+ and no statistically significant difference was found among the tumoral groups (p = .975), as summarized in Table 2. Those frequencies of E-cadherin gene hypermethylation in oral SCCs were broadly similar to previous reports (85.4%,11 78.2%,9 72%,6 64%7) but they were higher compared with the finding of Viswanathan et al10 (35% of tumors) and Hasegawa et al16 for head and neck SCC (36.5% of tumors). Previous studies6–11, 16 have examined the methylation status mainly in invasive and metastatic oral cancer, and VC has not been included in those investigations. Although the E-cadherin gene has been reported to be unmethylated in normal epithelial tissue,10 we confirmed that the methylation was specific to cancer cells because our analysis of oral fibrous hyperplasia specimens found the gene to be unmethylated (Figure 2). The evidence of hypermethylation in nonmetastatic and lowest invasive potential cancer as oral VC indicates that E-cadherin suppressor gene methylation can be a relatively early event in oral carcinogenesis, as suggested by Yeh et al.11

A strong association between smoking, especially with the increased number of smoked packs a year, and E-cadherin gene promoter methylation in oral SCC has been reported.16 In the present study, we found that E-cadherin gene hypermethylation in oral SCCs occurred mainly in patients with history of tobacco smoking or alcohol consumption, corroborating the findings of Hasegawa et al.16 In addition, Chang et al7 suggested that the occurrence of hypermethylation of the E-cadherin gene could be 1 of the factors that leads to the local recurrences and regional metastasis in oral SCCs. Our results reinforce this statement because in the metastatic oral SCC group (SCC-pN+), 6 (75%) of 8 recurrent tumors show E-cadherin gene promoter methylation.

In the present study, the immunohistochemical analysis of the E-cadherin expression was performed in the oral SCC invasive front tumor, where the important molecular events occur allowing the invasion and metastasis of cancer cells. Moreover, Kudo et al6 verified the presence of the hypermethylation of the E-cadherin gene in the most invasive areas of SCC in detriment to the noninvasive clones of malignant cells.

In our study, the complete membranous epithelial staining for E-cadherin was significantly higher (p = .016) than in the other well- differentiated SCC groups (SCC-pN0 and SCC-pN+). When the values obtained in immunohistochemical analyses were divided by score (≤36.5% and >36.5%), 89% of VCs showed a score >36.5% of complete epithelial membranous staining followed by 50% of SCC-pN0 and only 29% of SCC-pN+ (Table 3). Most of the oral well-differentiated SCCs with lymph node involvement (71%) presented reduced expression of membranous E-cadherin in malignant epithelial cells (Table 3 and Figure 3). An interesting aspect found was that the unique oral VC showing downregulation of E-cadherin expression with ≤36.5% of malignant cells complete membranous staining was a recurrent tumor (Table 3). These results reinforce the previous findings2, 3, 17–19 that the loss of the E-cadherin expression really consists of an important event associated with invasion and higher metastatic potential in oral SCC. In addition, our results corroborate those of Tang et al18 and Tian et al,19 who found significant difference of E-cadherin expression between oral VC and SCC.

When the relationship between E-cadherin methylation and its immunohistochemical expression was analyzed in each SCC group, a significant correlation (p = .044) was noted only in the metastatic oral SCC group (SCC-pN+). In nonmetastatic groups (VC and SCC-pN0), the methylation profile of the E-cadherin gene was not correlated with its imunohistochemistry expression, as shown in Table 4. Contrary to previous findings6, 7, 9, 11 that showed an association between hypermethylation of the E-cadherin gene and loss or reduction of the immunohistochemical expression of this molecule in oral SCC, in the present study, all tumors of the SCC-pN+ group unmethylated for E-cadherin gene showed reduction of its immunohistochemical expression (Table 4). The immunohistochemical analysis of the loss or reduction of E-cadherin separately can be conflicting because this molecule is part of a complex net of cellular adhesion influenced by local environmental factors,2, 11, 20 and besides the epigenetics events the performance of other transcriptional repressing, as well as the genetic alterations, can also induce the loss or the reduction of the E-cadherin expression.1, 15 Thus, probably as suggested in current results, mechanisms other than CpG methylation promoter can regulate the E-cadherin gene expression in oral SCC with or without metastasis.

Regarding the prognosis, important clinical and biological differences have been shown between invasive SCC and VC in the oral cavity.21 Despite the reduced samples of the oral VCs in the present study, the overall 5- and 10-year survival rates were significantly higher in the VC group and SCC-pN0 than in SCC-pN+ groups (Figure 1). These results likely reflect the indolent clinical behavior of the oral VC, which often shows a local invasive pattern but without any distant metastases. Moreover, 5-year survival rate of patients with VC was 87.5%, which were favorably compared with those reported previously21 and reinforces the efficacy of the surgical excision as primary treatment for these tumors.

In conclusion, the altered immunohistochemical expression of E-cadherin resulting from CpG methylation promoter is common and has also been shown to be heterogeneous in metastatic and nonmetastatic well-differentiated SCCs. The presence of E-cadherin gene methylation in tumors with the lowest invasive and metastatic potential as oral VC suggests the early involvement of this epigenetic event in the multistep progression of oral carcinogenesis.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES