Abstract
Background: Lymphangiogenesis, the formation of new lymphatic vessels, is an essential process in normal growth and development and wound healing. It is also observed in pathological processes including tissue inflammation and malignancies, where it may contribute to tumour metastasis. Physiologic lymphangiogenesis is controlled by factors including vascular endothelial growth factor (VEGF)-C and VEGF receptor (VEGFR)3 which promote the commitment of particular venous endothelial cells (ECs) towards lymphatic differentiation. Upregulated levels of lymphatic vessel endothelial hyaluronan receptor (LYVE)-1 are observed in these committed ECs, which then in turn express a lymphatic vessel transcription factor, prospero homeobox protein 1 (Prox-1). With further VEGF-C signalling the lymphatic endothelial cells (LECs) sprout out, forming lymph sacs, and now express the transmembrane glycoprotein D2-40 (podoplanin). This is followed by further development and differentiation controlled by VEGF-C/VEGFR-3, resulting in numerous lymph sacs which eventually form the lymphatic network. However, during pathological processes such as in oral cancer, this controlled lymphangiogenic regulation is lost and new lymph vessels form within and around the tumour. This may increase the likelihood of lymphatic metastasis, thus reducing the affected patient’s prognosis.
Aim: To investigate the differences, if any, in the expression profile of lymphatic markers and lymph vessel density (LVD) in oral squamous cell carcinoma (OSCC) in relation to non-specifically inflamed connective tissue (ICT) and normal oral mucosa (NOM) using immunohistochemistry (IHC). The specificity of putative antibody markers to lymphatic endothelial cells (LECs) was investigated with double-labelling immunofluorescence (DLIF).
Methods: Archival formalin-fixed paraffin-embedded (FFPE) specimens (28 OSCC, 10 ICT and 6 NOM cases) were analysed using standard IHC techniques with antibodies against the LEC markers D2-40, LYVE-1, VEGFR3 and Prox-1. Using light microscopy qualitative and quantitative analysis of immunopositivity was undertaken. Within each specimen six hotspots were chosen and photographed at 200x magnification. The positively stained cells and vessels were identified and counted to determine the vessel density per 1.56mm2 (total combined area of six hotspots). One-way analysis of variance (ANOVA) with 5% level of significance, following natural log transformation of the raw data, was used to analyse the differences between the three groups. Qualitative analysis of the overall staining pattern and intensity was also performed. DLIF was used to qualitatively investigate the specificity of D2-40 and LYVE-1 to LECs on OSCC tissues with the highest number of D2-40+ and LYVE-1+ vessels. In addition, DLIF with the angiogenic marker CD34 and D2-40 or LYVE-1 markers was undertaken.
Results: Positive staining was observed with all lymphatic markers in all groups. There was a higher expression of D2-40 (p=0.001) and Prox-1 (p=0.001) in the OSCC group when compared with both control groups. There was a statistically significant difference between the mean LVD of the OSCC group and ICT group (p=0.019). No statistically significant differences were observed between the OSCC group and the NOM group for LYVE-1 expression and between all groups for VEGFR3 expression. A comparison between each different lymphatic marker when applied on the same tissue sample group also demonstrated a significant difference in LVD in the OSCC tissue groups (p<0.001) but this difference was not observed in either the ICT (p=0.172) or the NOM (p=0.220) samples. DLIF qualitative analysis with merged imaging demonstrated an absence of any vessels showing D2-40+/CD34+ or LYVE-1+/CD34+ co-expression. The D2-40/LYYE-1 DLIF combination showed that the majority of lymphatic vessels that were D2-40+ were also LYVE-1+. However, a number of D2/40+ LVs were LYVE-1-.
Discussion and Conclusion: These results establish that the OSCC tumour microenvironment possesses significantly more lymphatic vessels expressing D2-40 and Prox-1 than the control groups. To an extent, there was also higher expression of LYVE-1+ LVs in OSCC (compared with the ICT control tissue group). This increase in LVD may play a role in facilitating lymphatic invasion and later metastases. These molecular entities may serve as potential anti-oral cancer therapeutic targets or as potential prognostic markers. Further analysis of the relationship of D2-40, Prox-1, LYVE-1 and the development of oral cancer should be undertaken and their expression in other types of cancer should be elucidated. DLIF findings from this study indicated that D2-40 is the most specific LEC antibody marker for OSCC tissues. The most ideal LEC marker- in terms of specificity, sensitivity, best visualisation, ease of use and cost is the D2-40.