It has long been appreciated that a melanoma metastasis is a heterogeneous collection of cells, including melanoma cells, fibroblasts, macrophages, lymphocytes and other stroma. Until quite recently it was assumed that the melanoma cell population within a given lesion was a homogeneous one, that all the melanoma cells were the same. This was an important assumption because the greater volume of molecular research into melanoma has been based on it. For example, most studies looking at the molecular biology of melanoma stage progression have treated same-stage tissues or cell lines as being replicate representatives of the stage from which they are derived. However, it turns out that this is a false assumption to make.
Many researchers now appreciate that some melanoma cells within a lesion are quite different from others. The melanoma stem cell hypothesis rests upon just this appreciation. In fact the melanoma stem cell hypothesis is held up as an explanation both for the heterogeneity of melanoma cells and why treatments that are carefully designed to attack melanoma cells ultimately fail to cure the disease (Schatton et al., 2008). While the melanoma stem cell hypothesis seems like a good model for melanoma progression, there is strong evidence which suggests it may not be correct (Quintana et al., 2008; Taussig et al., 2008). Instead, we believe that intra-lesional melanoma cell heterogeneity is the result of something quite different. We call this alternative model the phenotype switching hypothesis.
Our work is focussed on understanding the heterogeneity of melanoma cell populations within melanoma metastases. Starting from analyses of in vitro transcription patterns we developed a new model for disease progression which rests on the capacity for melanoma cells to switch back and forth between two phenotypic states (Hoek et al., 2006). Respectively, these two phenotypes are proliferative and invasive, and we hypothesized that phenotype is determined not by the melanoma cell itself but the microenvironment. We recently demonstrated an in vivo proof-of-principle for phenotype switching in a mouse (Hoek et al., 2008a).
Melanoma is a tumor-type which develops from pigment cells, and while it is one of the rarer forms of skin cancer it accounts for most skin cancer deaths. If a primary tumor develops into a metastatic disease the patient's outlook is very poor. Survival rates for metastatic melanoma rarely exceed one or two years, making it one of the most dangerous cancers known. If you get melanoma your best chance for survival remains early detection and surgical removal of the primary lesion before it has a chance to metastasize. Despite decades of clinical and molecular study there remains no effective treatment for metastatic melanoma.
Clinical observation has documented the progression from normal melanocyte to metastatic disease, passing through several well defined stages. These include nevi formation, transformation to an in situ primary, progression to a vertical growth phase and then on to distal metastases via the vasculature. For years scientists have sought to describe the molecular changes which precipitate transformation and the subsequent metastatic cascade. These efforts have not yet enabled clinicians to effect a viable therapy.
Metastatic potential is a term which characterizes the capacity for melanoma cells to escape the original tumor and form metastases in distal tissues. Our research has led us to conclude that the activities of cell proliferation and invasion are distinct programs that are determined by microenvironmental influence on signal pathways. In switching between these programs (phenotypes) melanoma cells may also change a great many other factors, including their surface antigen complement. This could explain why experimental immunotherapies have so far failed to show any improvement in patient survival. These ideas were first discussed in a paper we published in 2006 and then tested in 2008 (Hoek et al., 2006; Hoek et al., 2008).