Vascular endothelial growth factor (VEGF) is an angiogenic cytokine and mast cells play a role in neoangiogenesis in various malignancies. The aim of the present study was to elucidate the role of VEGF and mast cells in the early stages of tumorigenesis in oral squamous cell carcinoma (OSCC). Immunohistochemistry was conducted to study VEGF expression and microvessel density (MVD) in 49 tissue samples, 31 OSCCs, 13 leukoplakias (8 with and 5 without dysplasia) and 5 samples from normal oral tissue. Counterstaining with tolouidine blue was conducted to reveal mast cells. The number of microvessels and mast cells were counted at the same optical field. A gradually increased VEGF expression was observed from normal oral epithelium to leukoplakia and OSCC. MVD was found to increase significantly between normal oral tissue and OSCC ( p = 0.000). The number of mast cells was found to increase significantly between normal oral tissue, dysplasia ( p = 0.012) and OSCC ( p = 0.000). In the early stages of tumorigenesis in OSCC, VEGF, which is secreted by the epithelium, is gradually increased immediately affecting the population of mast cells, which are then related to the increase of microvessels.
Angiogenesis is the biological process of vascular tree formation in all living beings. Angiogenesis is extremely important both in physiological conditions such as the development of the corpus luteum, the proliferation of the endometrial cells and wound healing but also in pathological conditions such as rheumatoid arthritis, diabetic retinopathy, and the growth and metastasis of solid tumours.
Vascular endothelial growth factor (VEGF) is a powerful cytokine implicated in neoangiogenesis in a number of malignancies, among which oral squamous cell carcinoma (OSCC). VEGF is related to the proliferation and migration of endothelial cells, vascular permeability, the production of many vasoactive molecules and the chemeotaxis of mast cells and monocytes.
Mast cells are known to be the main effector cells in type I hypersensitivity reactions. They are highly granulated cells capable of producing many proinflammatory, immunoregulatory and angiogenic molecules.
Several studies have related increased VEGF expression from the epithelium, or mast cells in the connective tissue with neoangiogenesis, neoplastic progression and metastasis in a number of malignancies. The role of VEGF and mast cells in the early stages of tumorigenesis in OSCC has not been elucidated. The aim of the present study was to evaluate the relation between VEGF expression, mast cells and angiogenesis in normal oral epithelium, leukoplakia without dysplasia, leukoplakia with mild or moderate dysplasia, leukoplakia with severe dysplasia and OSCC.
Materials and methods
49 formalin-fixed, paraffin-embedded tissue specimens from biopsies of oral lesions were stained and analyzed for the evaluation of VEGF expression, mast cells and angiogenesis. The tissue specimens were 31 OSCCs and 13 leukoplakias, 5 without dysplasia and 4 with mild or moderate dysplasia and 4 with severe dysplasia. Five samples from normal oral tissue were used as a control sample.
In order to focus on the early stages of cancer development in OSCC, the inclusion criteria were: patients newly diagnosed with leukoplakia without and with dysplasia or OSCC from 1995 to 2008; and OSCCs staged as T 1 N 0 M 0 or T 2 N 0 M 0 . The patients were subjected to a complete radiographic examination to determine the tumour TNM stage.
The diagnosis was based on the clinical examination of the patients and the histopathologic examination of the tissue specimens. The diagnosis of oral leukoplakia was based on the criteria proposed by Neville as ‘a white patch or lesion that cannot be characterized as any other condition such as lichen planus, chronic cheek bite, frictional keratosis, tobacco keratosis, nicotine stomatitis, leukoedema and white sponge naevus’. The diagnosis of epithelial dysplasia was based on the criteria defined by WHO (International Collaboration Center for oral precancerous lesions).
The clinical and histopathological criteria recorded were: patients’ age and gender, site, stage and histopathologic differentiation of the lesion. The study was conducted according to the criteria defined by the Helsinki declaration.
The tissue sections had been fixed with 10% buffered formalin and embedded in paraffin. From each paraffin block, two tissue sections 4 μm thick were cut.
Immunohistochemical staining for VEGF expression
VEGF expression was determined using a mouse polyclonal anti-human VEGF antibody (clone SPM225) (Spring Bioscience, Freemont, Canada). Tissue sections from human Kaposi’s sarcoma were used as an internal positive control for VEGF expression, and a mouse anti-human IgG 1 antibody was used as an internal negative control for VEGF expression.
The tissue sections were deparaffinized with xylene and rehydrated with graded alcohols. The tissue sections were incubated with 1% peroxide (H 2 O 2 ) in methanol for 30 min to block endogenous peroxide activity. The slides were then rinsed in triethanolamine buffered saline (TBS) solution and heated in a steamer for 10 min in a preheated ethylenediamine tetraacetic acid (EDTA) solution of pH 8, to reveal the antigen epitopes. The tissue sections were left to cool at room temperature.
The anti-VEGF antibody’s dilution was 1:25 and was applied overnight on the tissue sections at a temperature of 4 °C. The immunohistochemistry technique was conducted using the Envision system (Envision HRP, Mouse/Rabbit detection system) (DakoCytomation, Glostrup, Denmark) according to the manufacturer’s instructions. Diaminobenzine (DAB) chromogen was applied to visualize VEGF expression. The tissue sections were counterstained with haematoxylin for 30 s, dehydrated with graded alcohols, covered with entalan and observed under light microscopy ( Fig. 1 ).
Evaluation of VEGF expression
Two independent researchers without any knowledge of the patients’ medical records observed the immunohistochemically stained tissue sections under light microscopy. The degree of VEGF expression was determined from 0 (no VEGF expression) to +3 (major VEGF expression), according to the positivness and the intensity of the staining.
In every tissue section, in a predefined area, 3 microscopic fields were studied at a magnification of ×200. When the staining intensity at the epithelium was similar to that of endothelial cells, VEGF expression was graded as +2. Staining weaker than that of endothelial cells was graded as +1, stronger staining was +3. Tissue sections, where no VEGF expression was detected, were graded 0.
Immunohistochemical staining for MVD
Immunohistochemical staining was conducted using a mouse monoclonal anti-human anti-CD34 Class II antigen (clone QBEnd-10). A mouse polyclonal IgG 1 antigen was used for the internal negative control (DakoCytomation, Glostrup, Denmark). To reveal antigen epitopes the tissue sections were heated in a microwave oven for 10 min bathed in citric acid buffer with pH 6.7. The primary anti-CD34 antigen was diluted in a percent of 1:50 and was applied on the tissue sections for 1 h at room temperature. To visualize CD34 staining the Envision system was used (Envision HRP, Mouse/Rabbit detection system) (DakoCytomation, Glostrup, Denmark) according to the manufacturer’s instructions.
Counterstaining with tolouidine blue
After the immunohistochemical staining with anti-CD34 antigen, the tissue sections were counterstained with toluidine blue, which reveals mast cells in the connective tissue. The tissue sections were dipped in toluidine blue solution 0.1% 15 times, then rinsed in tap water. The tissue sections, were dipped quickly, only once in a 70% alcohol and a 0.5% HCl solution and then quickly dried by firmly pressing them against absorbent paper. They were dipped 12 times in xylene, left to dry and covered with entalan. Using the counterstaining technique, mast cells and microvessels were observed and counted at the same optical field ( Fig. 2 ).
Microvessel density assessment
Observing the tissue sections under microscopy, the areas with the larger number of microvessels were marked. Using a magnification of ×400, 3 optical fields, with the larger number of microvessels were selected, with a one-field depth from the epithelial basement membrane, the tumour invasive margins, or the tumour nests. Every single stained endothelial cell or cell cluster, clearly separated from the adjacent microvessels was considered as a single countable microvessel. MVD was defined as their mean number per optical field (in the 3 selected optical fields).
Mast cell density assessment
In the three selected optical fields under ×400 magnification (those with the larger number of microvessels), the number of mast cells was also counted. Mast cell density (MCD) was defined as their mean number, per optical field.
MCD in the connective tissue
Separate tissue sections from the same paraffin blocks were stained only with toluidine blue to ensure that the same number of mast cells was stained using the simple tolouidine blue staining technique and the counterstaining technique.
Statistical analysis was performed using the statistical package SPSS 16 (Chicago, Illinois). To test the statistically significant differences in VEGF expression, the Mann–Whitney U-test was used after the Kruskall–Wallis test at the level of p < 0.05. The significance of the differences in MVD and MCD between normal oral tissue, leukoplakia without dysplasia, leukoplakia with dysplasia and OSCC, were tested using the LSD test, after ANOVA, at the level of p < 0.05. To test the relation between VEGF expression and MVD and VEGF expression and MCC, the χ 2 and Spearman’s Rho tests were used. To test the correlation between MVD and MCD regression analysis was used and the curve fit model was estimated. The three variables (VEGF expression, MVD, MCD) were classified into two groups by applying the single linkage cluster analysis with Euclidean distances on the principal components since the three variables were found to be related with each other. This classification is used to elucidate the real relation between parameters that relate with one another.
Of the 49 patients, 28 were men and 21 women, they had a mean age of 61 ± 9.81 years (range 26–87 years). The lesion sites were the lingual mucosa, the buccal mucosa, the vermilion border, the floor of the mouth and other sites within the oral cavity ( Table 1 ).
|Histological diagnosis||Lesion site|
|No cases||Vermilion border||Buccal mucosa||Lingual mucosa||Floor of the mouth||Other sites|
|Leukoplakia without dysplasia||5||1||1||2||1||0|
|Leukoplakia with mild or moderate dysplasia||4||1||1||1||1||0|
|Leukoplakia with severe dysplasia||4||1||0||1||1||1|
Red cytoplasmic staining was observed in the epithelial cells that expressed VEGF. Tissue sections from a human Kaposi sarcoma lesion expressed VEGF very strongly, serving as an internal positive control and revealing that VEGF antigen strongly stains endothelial and epithelial cells. The immunohistochemical staining of the cancer cells was heterogeneous. Normal salivary glands, hair follicles and muscular fibres were also stained very strongly. Endothelial cells in the connective tissue were also stained for VEGF. Some inflammatory cells, invading the lesion margins were also stained strongly (strongly expressed VEGF). The same staining pattern was observed in tissue sections with leukoplakia without dysplasia, leukoplakia with mild or moderate and leukoplakia with severe dysplasia. Very weak VEGF staining was observed in normal oral tissues.
VEGF expression was found to increase significantly from normal oral tissue with an average value of 1 to leukoplakia without dysplasia with an average value of 1.45 ± 0.688 ( p = 0.000, MWU = 0.000), leukoplakia with mild or moderate dysplasia with an average value of 1.46 ± 0.744 ( p = 0.000, MWU = 0.000), leukoplakia with severe dysplasia with an average value of 2.20 ± 1.014 ( p = 0.000, MWU = 0.000) and OSCC with an average value of 2.12 ± 0.849 ( p = 0.000, MWU = 0.000) ( Fig. 1 ). No correlation was found between the patient’s age and the lesion site with VEGF expression ( p > 0.05).
In normal oral epithelium, leukoplakia without dysplasia, leukoplakia with mild or moderate dysplasia, and leukoplakia with severe dysplasia microvessels were found to be located just underneath the epithelium. The stained microvessels were revealed as brown spots, strands or lumens. MVD was calculated in every tissue section. In normal oral tissues the mean MVD was 14.08 ± 6.089, in leukoplakia without dysplasia 20.00 ± 5.675, in leukoplakia with mild or moderate dysplasia 22.57 ± 8.262, in leukoplakia with severe dysplasia 22.52 ± 8.684 and in OSCC 27.78 ± 9.257 ( Table 2 ).