Metastasis is the spread of a cancer to tissues other than the site of the primary tumour where the cancer originated. The ability to metastasize has been defined as a hallmark of cancer in Hanahan and Weinberg’s original review . Metastasis is often the cause of cancer-related deaths. As such, metastasis is used as a criterion to classify tumour stage. Metastases are often already present when a patient is first diagnosed. In order to metastasise, tumour cells need to go through a stepwise process, which involves being able to ‘break free’ from the primary site, enter into the lymph and/or blood circulation, exit the circulation, and settle and proliferate at the site of metastasis. Metastasising cancer cells need to satisfy the conditions for each step before successfully invading another tissue; this often requires a genetic adaptation. Several genetic adaptations have been associated with one or more of the steps of metastasis. For example, loss of the adhesion protein E-cadherin has been found to increase tumour invasiveness, as cells lacking E-cadherin expression are more loosely connected to other cells, increasing their motility [83, 84]. Another pathway associated with metastasis is the Epithelial to Mesenchymal Transition (EMT), where an epithelial cell gains the characteristics of a mesenchymal cell, such as loss of cell polarity and enhanced motility and invasion [85, 86]. Metastatic cancer cells have also been observed to possess invadopodia, cellular structures that aid in extravasation .
One of the main discussion points around metastasis is how it originates and at what point in tumourigenesis it arises . Metastatic cells were first thought to arise from the primary tumour in a linear fashion: a subset of cells from the primary tumour accumulate metastasis-promoting mutations over time, until a subclone with full metastatic potential arises . This metastatic subclone would thus arise late in tumourigenesis and therefore have substantial genetic similarity to the primary tumour. More recently, a different model for the generation of metastases was proposed. This parallel progression model argues that a metastatic subclone is formed and disseminates early during tumourigenesis . This clone might not be fully metastatic (i.e., not yet capable of distant organ colonisation) but acquires these characteristics separately from the primary tumour. Thus, in this model the similarity between primary tumour and metastasis is expected to be lower.
Research on primary tumours and their metastases has generated results supporting both theories [91,92,93,94,95], and modes of metastases might differ between cancer types or even between patients [96, 97]. There is as of yet no consensus on whether the mode of progression depends on the cancer type and environment. Another point that has recently come under debate is how cancers exactly spread to distant organs. Lymph node metastases are often seen before distant metastases, and this has led to the assumption that distant metastases are derived from the lymph node metastases. This paradigm is questioned by a recently published paper by Naxerona and colleagues. From their phylogenetic investigation of lymph node and distant tumours, it follows that only 35% of distant metastases are derived from lymph node metastases . The cellular characteristics needed for lymph node involvement might therefore be different from those needed for distant spread. Still, lymph node involvement is a robust prognostic factor for the development of distant metastases in many cancers, so LNM does tell something about the metastatic potential of a primary tumour.