CONVENTIONAL MODELS OF METASTATIC PROGRESSION

The somatic mutation theory has long been regarded as the “conventional” model to explain metastasis at the molecular level. Originally postulated by Nowell [1], it states that metastatic tumor cells acquire the necessary characteristics to facilitate colonization and proliferation at distant sites through a sequential accumulation of somatic mutations. Experimental support for this theory was subsequently provided by Fidler and Kripke [2], who demonstrated that clonal variants isolated from bulk tumor tissue had differing metastatic potentials. It was postulated that the origins of this differential metastatic capacity arose from the acquisition of different somatic mutations in individual clonal isolates. Later work demonstrated that these somatic mutations induce hyperactivation of metastasis-promoting genes and silencing of metastasis suppressors, and that individual tumor cells are susceptible to the accumulation of such mutations as a consequence of their inherent “genomic instability” (reviewed in [3]).

Subsequent in vivo experimentation, however, revealed that although somatic evolution is likely a critical determinant of metastatic potential, it cannot entirely explain the molecular basis of metastasis [4–6]. The advent of microarray technology to assay global patterns of gene expression has shed fresh light on the limitations of the somatic evolution theory as a comprehensive mechanism for metastasis at the molecular level (reviewed in [7]). Specifically, many studies have demonstrated that patterns of gene expression (or “signatures”) in bulk tumor tissue can be used to predict survival in breast cancer (early examples include [8–10]) as well as many other solid tumors [7].