CANCER AND METASTATIC FOOTNOTE
Metastasising cells are like an iceberg that breaks away from a large mass and is carried off by the ocean currents. Sooner or later the iceberg will melt and vanish: much the same happens to almost all metastatic cells: they get picked up and destroyed by scavenger cells in the circulation. But some metastasising cells are not eliminated: they stick to a target site somewhere on the lining of the circulatory systems and then manage to cross the vessel wall into the surrounding tissue. This is similar to what happens after an egg is fertilised: the cell starts dividing to form a clump of cells (a blastocyst), which implants in the endometrium and, eventually when it’s reached a suitable size, makes its entry into the world. But at least 25% of pregnancies end in miscarriages and this figure rise to 75% with increasing age. Despite metastasis being the most life-threatening facet of cancer, the odds against escaping are much worse. Only about 2% of circulating tumour cells manages to extravasate so that they can form micro-metastases, and only 1% of these manage to persist and expand into tumours. Attrition within the circulation arises from shear stress leading to mechanical damage and by a form of apoptosis activated when the cells are no longer anchored to a substrate (called anoikis). Some tumour cells can acquire a degree of protection by coating themselves with platelets, but even so, with a success rate of less than 0.02%, the overall efficiency of metastasis is very low. Given the importance of stromal cells in promoting tumour development, it is that they might also modulate the efficiency of metastasis. Direct evidence for such a role has come from mice that ubiquitously express the green fluorescent protein (GFP) with implanted tumours of cells expressing a red fluorophore. While over 80% of the collected from circulating blood were single cells, there were also tumour clumps that included host (green) cells (Fig 1).
1. Metastatic seeding of tumour cells associated with stromal fibroblasts. Left:
The aggregates, in which fibroblasts were the predominant stromal cell, are more efficient than single tumour cells at establishing metastatic growths in the lung. The suggestion that, as Rakesh Jain and his colleagues put it, metastatic cells may ‘bring their own soil’ in the form of associated stroma, has been confirmed by examining metastases of human primary tumours occurring in the brain. Because fibroblasts are not part of normal brain tissue, their identification in such metastases indicates that they must have been transferred from the primary site with the tumour cells, subsequently establishing themselves and proliferating. If this emerges as a common feature of human cancers it would raise the possibility of targeting tumour clumps as a useful anti-metastasis strategy.
There’s another, perhaps rather obvious question, you might ask about metastasis. If primary tumours shed cells into the circulation and some of these eventually become secondary tumours in a new location, what’s to stop cells from a metastasis doing the same thing in reverse? At least in mice, the answer is ‘nothing and they do’. This has given rise to the idea of what Joan Massagué of the Memorial Sloan–Kettering Cancer Center has described as ‘tumour self-seeding’ and it can be visualised by labelling metastatic cells and inoculating them into mice.
This type of metastasis may occur in at least some human cancers and, given that the cells doing the seeding have already jumped some of the major hurdles on the road to becoming a fully malignant tumour, may contribute to the aggressiveness of some cancers. This process could also occur even after a primary tumour has been surgically removed, giving a second mechanism for tumour recurrence in addition to incomplete removal of a primary tumour. There is, however, no conclusive evidence that this occurs in humans and it is a very difficult phenomenon to tackle experimentally.
Cancers are, of course, abnormal growths of cells and their damaging effects on the tissues they grow in can kill in a direct way. Thus, colon cancers, other tumours in the gastrointestinal tract and also ovarian carcinomas can obstruct the bowel, which would be fatal without surgical intervention. Similarly, lung tumours can be fatal if they block lung function (Fig. 2) and anaplastic thyroid tumours can, in effect, cause strangulation.
2. Metastatic tumour in mouse lung and the protective effect of combretastatin.
A variant on this theme is contributed by leukaemias that result in a huge excess of white cells over red cells that so increases the viscosity of blood that circulation is drastically impaired. In general, however, human beings are remarkably resilient to organ damage. We can manage with half a kidney and if we lose two-thirds of our liver it will regenerate itself. It is therefore rare for cancer fatalities to be due to organ failure. Death is usually caused by secondary effects: principally infection. Cancer patients generally become increasingly susceptible to infection due to the decreased efficiency of their immune systems. Tumours that damage the walls of tissues can also increase vulnerability to infection. The agents are common bacteria (e.g. E. coli, Pseudomonas), which can overwhelm the host even with antibiotic treatment, but fungi are also significant contributors. Tumours can also cause damage to blood vessels, leading to haemorrhage, particularly in the liver.
About 40% of cancer deaths occur from malnutrition – a general condition of starvation and debilitation called cachexia (wasting syndrome) that develops in many other chronic diseases. Cancer cachexia is not understood and there are no satisfactory therapeutic treatments. Both chemotherapy and radiation therapy can also induce cachexia, and weight loss, in turn, reduces the efficacy of chemotherapy. Metastases may also suppress the immune system, e.g. if they are in the bone marrow or if in the brain, raised intracranial pressure.
Tumour development is a complex process driven by multiple mutations and also by contributions from neighbouring host cells. Despite this complexity, the basic changes associated with the conversion from normal to tumour cell can be simply summarised (see key points below). In essence, these reduce their dependence on environmental cues, avoid mechanisms that eliminate abnormal cells and recruit host cells to enable them to survive and spread to secondary sites. These remarkable achievements derive from the type of driver mutatios. In particular, constitutively activated RTKs promote proliferation independent of extracellular signals, and loss of function of key tumour suppressors (p53 and RB1) circumvents the central defence systems that protect against cancer. As these complex interactions are gradually unveiled they offer an increasing range of targets for therapeutic intervention.