Lymphatic vessels play an important role, not only in physiological processes such as environmental stability, immunologic surveillance and fat absorption, but also in pathological processes such as lymphedema, tissue repair, inflammatory reaction, and particularly, in the metastatic spread of tumors. Metastatic spread of tumors is an important prognostic factor for cancer patients. The effect of angiogenesis on cancer cell proliferation and metastatic spread has been confirmed. However, less attention has been focused on research involving tumor lymphangiogenesis versus research on tumor angiogenesis, owing to the lack of specific markers for lymphatic vessel endothelial cells (LVECs). Recently, improvement in the isolation techniques of LVECs and the discovery of specific LVEC markers have led to advances in research involving lymphangiogenesis in carcinoma tissues. This review evaluates the advances in research regarding lymphangiogenesis in carcinoma tissues.

Vascular endothelial growth factor receptor-3 (VEGFR-3), also known as Flt-4, is the earliest and most widely used LVEC marker. It is specifically expressed on the lumen of LVECs in normal adult tissues. However, the liver, splenic blood sinuses, tissues involved in wound repair and neovascular endothelial cells of carcinomas are also known to express VEGFR-3. Thus, VEGFR-3 as an LVEC marker does not have a high specificity.

Podoplanin is an LVEC marker with a relatively high specificity and is mainly expressed in micro-LVECs. Podoplanin is an integrated cytoplasm/cell membrane mucoprotein that was initially confirmed in the glomerular podocyte. However, its biological function has yet to be determined. Schacht et al (1) reported that podoplanin plays an important role in the regulation of the process of lymphangiogenesis. Knock-out of the Podoplanin gene in mice leads to a reduction in the transport capacity of lymphatic vessels, an expansion of lymphatic vessels and congenital lymphedema. Studies have demonstrated that podoplanin is expressed in squamous cell carcinoma and in some germ cell tumors, suggesting a possible association of podoplanin with tumor progression (2).

Lymphatic vessel endothelial hyaluronan receptor-1 (LYVE-1), which is homologous with CD44, is expressed on the surface of LVECs and is considered to be the most valuable marker of LVECs. It is assumed that LYVE-1 is associated with the metabolism of hyaluronic acid in the extracellular matrix (3).

Prox1 is a homeobox gene transcriptional factor and is specifically expressed in embryos, adult normal tissues, and intratumoral LVECs. One study revealed that Prox1 is involved in the formation of the LVEC phenotype during lymphatic vascular development, indicating that it can be used as a marker of LVECs (4).

However, the aforementioned markers are not entirely specific for LVECs. Therefore, a certain amount of cross-reaction between LVECs and microvascular endothelial cells was observed during the staining procedure. The immunohistochemical method of double dyeing is used to label lymphatic vessels in order to discriminate between blood and lymphatic vessels. In addition, a recent gene analysis of lymphatic and blood vascular endothelial cells revealed certain novel LVEC-specific genes, such as macrophage mannose receptor-1, integrin-α9 and chemotactic factor CCL20.

The mechanism of tumor lymphangiogenesis is complicated and has yet to be determined. New lymphatic vessels in tumor tissues may originate from bone marrow endothelial progenitor cells, directly from preexisting lymphatic vessels, and even by cell transformation.

Bone marrow endothelial progenitor cells express CD34 and VEGFR-2 and are a cellular subgroup that originates from endothelial cells lining the circulatory system. These endothelial progenitor cells proliferate, permeate into, and fuse with vascular endothelial cells and are involved in tumor angiogenesis under pathological conditions. Salven et al (4) reported that besides CD34 and VEGFR-2, certain endothelial progenitor cells also express the LVEC marker VEGFR-3, indicating that these cells may be the progenitor cells of LVECs. Moreover, the endothelial progenitor cells are involved in tumor lymphangiogenesis in the same manner as tumor angiogenesis mentioned above.

However, it has been shown that new intratumoral lymphatic vessels originate by a ‘budding’ of preexisting lymphatic vessels. He et al (5) found that new lymphatic vessels mainly originated from preexisting lymphatic vessels of tissues undergoing lymphangiogenesis in carcinoma cases, but no or few vessels originated from bone marrow endothelial progenitor cells.

Other investigators showed that body macrophages, not only promote the proliferation of preexisting LVECs, but also migrate into the tissue stroma and are directly transformed into LVEC clusters to communicate with preexisting lymphatic vessels (6). This finding suggests that some cells are transformed into LVECs to generate new lymphatic vessels. However, whether tumor cells generate new lymphatic vessels in this manner has yet to be confirmed.

In the process of tumor lymphangiogenesis, lymphangiogenic factors, such as VEGF-C and VEGF-D, combine with the specific receptor VEGFR-3, which is distributed in LVECs, and promote LVEC proliferation to generate new lymphatic vessels. These new intratumoral lymphatic vessels have structures that are similar to new physiological lymphatic vessels, but the former have thin vessel walls and lack tight junctions, which are beneficial for the promotion of tumor metastasis (7).

Notably, whether such new intratumoral lymphatic vessels exist and play a role in tumor metastasis has yet to be determined. Animal trials have shown that new lymphatic vessels are generated by activating lymphangiogenic factors and are associated with lymph node metastasis of tumors (8). However, no similar result has been reported in humans. Human intratumoral lymphatic vessels consist of only 2 or 3 LVECs, which are small, irregular and collapsed. Consequently, these vessels are thought to provide no activities, although some of them show positive staining for proliferating cell nuclear antigens (9). A possible explanation for this finding is the rapid proliferation and tumor tissue division in human cancer cells, which increases the interstitial hydrostatic pressure and blocks the lumen of new lymphatic vessels (10). In addition, tumor invasion can disrupt lymphatic vessels (11). Although the intratumoral lymphatic vessels may not provide any activity, the increase in lymphatic vessels can increase the risk of lymph node metastasis of tumor cells, and subsequently increase the intratumoral lymphatic vascular density (ILVD) associated with lymph node metastasis and poor prognosis of cancer patients. Maula et al investigated LYVE-1-positive lymphatic vessels in 97 patients with head and neck squamous cell carcinoma. Results of these authors showed that ILVD was also associated with a risk of local recurrence in patients (12). Sipos et al investigated pancreatic endocrine cancers, which were characterized by vascular invasion and metastasis. In their study, it was shown that the levels of ILVD increased in certain types of tumors (13). This suggests that intratumoral lymphatic vessels are not only associated with lymph node metastasis, local recurrence and poor prognosis, but may also be associated with the malignant progression of tumors. Therefore, ILVD can be used to predict risk of lymph node metastasis and prognosis of cancer patients.

Peritumoral lymphatic vessels are believed to play a more important role in the process of tumor metastasis by providing more channels for lymphatic invasion and metastatic spread (14,15). Using a dye-adsorption test, Padera et al (14) confirmed that functional intratumoral lymphatic vessels were mostly located in the peritumoral zones in that the peritumoral lymphatic vessels had an expanded lumen that was disordered and distorted and had visible cancer emboli in their lumen, compared with the collapsed and flat intertumoral lymphatic vessels. However, Straume et al (15) found that an increase in peritumoral lymphatic vessels density (PLVD) in the cutaneous melanoma was often accompanied by lymphocyte invasion and was positively correlated with the prognosis and survival of patients. One possible explanation for this finding is that increased peritumoral lymphatic vessels promote the recruitment of antigen-presenting cells (APCs), thereby stimulating immune response against the tumors.

The molecular mechanism of tumor lymphangiogenesis is complicated, and numerous factors are directly or indirectly involved in this process. Currently, the VEGF-C/VEGF-D/VEGFR-3 signaling pathways are believed to be the most important mechanism underlying tumor lymphangiogenesis (16–22).

Although current research on tumor lymphangiogenesis is incomplete and a number of conflicting issues need to be resolved, we believe that suppression of tumor lymphangiogenesis and lymphatic metastasis provide novel ways of treating cancers. The treatment modalities use gene therapy based on advances in the research on tumor lymphangiogenesis and elucidate its underlying mechanism.