Search

Cancer’s ability to form metastases accounts for the majority of cancer deaths. Although many years of cancer research have generated an impressive list of properties detected on aggressive metastatic cancer cells, the order of events, the causal interrelationships and the decisive contributions that are limiting in the generation of the metastatic phenotype remain ill understood. One reason for this state of ignorance may be the tremendous complexity of the metastatic process, even exceeding that of carcinogenic transformation.

The acquisition of the ability to metastasize is the ultimate and most deadly level of the tumorigenic progression. Metastatic cells are able to dissociate from the primary tumor, invade through the surrounding tissue and endothelial layer into the bloodstream, migrate through the bloodstream to the site of metastasis and then reinvade and proliferate to generate a new tumor lesion (Poste and Fidler 1980). Throughout this process the metastatic cell assumes a number of different morphological configurations varying from densely packed within the tumor mass, deformed and elongated in invasion, and free-floating in the bloodstream. During each of these stages it must acquire distinct functional capabilities requiring the expression of distinct molecular entities.

Many interesting results were presented. A very interesting observation was reported by Eisenbach et al. (96), who found that metastatic mouse tumor cells express little or no major histocompatibility complex (MHC) class I H-2K molecules. This may allow these cells to avoid the expression of a transplantation antigen that could induce the rejection of metastatic cells. In the course of studying this phenomenon, they made the remarkable observation that the fos oncogene promotes the expression of H-2K.

Despite the evidence that neoplasms begin with the oncogenic transformation of a single normal cell they virtually always become genotypically and phenotypically heterogeneous by the time they are clinically detectable (Nowell 1976). This is thought to occur through the continuous emergence of new variant subclones as a result of random or induced genetic mutations combined with selection of the more aggressive of these new subpopulations (Nowell 1976). As a result, tumor progression can be viewed as a Darwinian-like microevolutionary process. There is evidence that variant cell subpopulations having competence for metastasis or manifesting certain kinds of drug resistance arise as a result of this process (Nowell 1976). With respect to the metastatic phenotype, a prediction of this model is that cells recovered from metastases would on average manifest a greater metastatic ability than tumor cells recovered from the corresponding primary tumor. This prediction was fulfilled, as first reported by Fidler (1973). As a result of this pioneering study it has become routine to compare cells of primary tumors and their metastases as a means of determining the cellular factors which govern expression of metastatic behavior.

Altered control of cellular proliferation and altered differentiation represent two important pathways determining the phenotypic properties of tumour cells. Indeed, induction of differentiation as a means of reducing tumour proliferative capability and directing the development of benign from malignant cells has long held promise as an alternative to conventional chemotherapy as a therapeutic approach to neoplastic diseases (Pierce 1967). The induction of differentiation as a means of reducing malignancy and improving survival has been studied most extensively using haematological tumours, particularly the myeloid and erythroid leukaemias of mice, but the phenomenon can be duplicated in some common solid neoplasms (Bennett et al. 1986).