Vascular mimicry in cultured head and neck tumour cell lines

INTRODUCTION: Vascuologenesis is the de novo establishment of blood vessels and vascular networks from mesoderm-derived endothelial cell precursors (angioblasts). Recently a novel mechanism, by which some genetically deregulated and aggressive tumour cells generate "micro-vascular" channels without the participation of endothelial cells and independent of angiogenesis, has been proposed. This has been termed "vasculogenic mimicry" and has implications beyond angiogenesis and adds another layer of complexity to the current concept for the generation of tumour micro-circulation. We suggest this is common phenomenon in head and neck squamous cell carcinoma (HNSCC) cell lines and other aggressive tumour cell lines. We present experimental evidence of vasculogenic mimicry in HNSCC cell lines and compare them with other tumours and a positive control vascular cell line. MATERIALS AND METHODS: The cell lines used were HUVEC, HN 2a, 2b (primary and metastatic tongue base squamous carcinoma cell line), HCT116 (colonic carcinoma cell line) and DU145 (prostate carcinoma cell line). Pilot experiments were undertaken to assess growth of a bank of tumour cell lines on (growth factor reduced) matrigel (Sigma) with standard media (DMEM with 10% Fetal Calf Serum). A functional growth assay was performed by preparing the appropriate cell suspension in serum free medium plated onto either bare plastic or a well pre-coated with growth factor reduced type 4 collagen analogues. Phase contrast photomicrographs were taken at 4 hours and 24 hours. Image analysis was performed; particular features of interest were two dimensional area (surrogate of growth and migration), branch points and end point measurements (surrogate of intercellular complexity). RESULTS: There were observable differences in growth of the cells on laboratory plastic and collagen matrix. Tumour cells formed capillary like networks similar to HUVEC cells. Metastatic HNSCC cells lines were found to have vasculogenic properties similar to HUVEC cell lines when compared to cell lines from their corresponding primary tumour. The endothelial growth factor antibodies used did not inhibit or stimulate cell growth when compared to control but did discourage vascular mimicry. Other tumour cell lines also displayed this property. DISCUSSION: Tumour "vasculogenic mimicry" must still be regarded as a controversial issue whose existence is not proven. The clinical importance of this phenomenon however, is that it does explain the lack of complete efficacy of current anti-angiogenic treatments due to the added layer of complexity. It provides a feasible mechanism of early tumour vascular supply which can co-exist and incorporate with later angiogenic mechanisms. We suggest that "vasculogenic mimicry" maybe a common neoplastic phenomena which appears to also be dictated by the cells micro-environment. Its existence also suggests a further process that of the development of tumour mosaic vessels as the neo-vasculature integrates with the existing endothelial lined systems.