Metastasis
A Socratic walk-through of metastasis — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why is a tumour often harmless where it began and deadly once its cells travel?
A lump in the breast can be cut out, and the person lives a normal lifespan. The same cells, having reached the liver or the lungs or the spine, are usually incurable. The cells did not become more malignant in transit — often they are recognisably the same tissue, which is why a secondary in the bone is still called breast cancer. So what is added by the journey? And a second puzzle hides behind the first: if travelling is what makes cancer lethal, and tumours shed cells into the bloodstream constantly, why does anyone with a tumour survive at all?
Reasoning it through
REASONING #Take the first question first. A tumour confined to one place is a mechanical problem with a mechanical solution: remove it, or destroy it locally. What spread removes is not the option of treatment but the option of localised treatment. Once deposits exist in several organs, with more probably present below the threshold of detection, there is nothing to draw a margin around — and the cause of death is usually the failure of an organ the tumour did not start in.
Now the second puzzle. Follow a cell that leaves. It must loosen its attachments to its neighbours, degrade the basement membrane that pens epithelial tissue in, and move. Then get into a vessel, survive the journey, get out again somewhere else, survive in a tissue whose signals were never meant for it, and finally start dividing there. Each of those is a filter.
One filter deserves naming because it is unexpectedly strict. A normal epithelial cell that loses contact with its matrix kills itself — a form of programmed cell death called anoikis, from the Greek for homelessness. Detachment is the trigger, and the point of the mechanism is precisely to stop cells setting up where they do not belong. A cell that means to travel must first switch that off. The very first requirement of metastasis is not aggression but a suppressed suicide.
How strict are the filters together? Strikingly so. In a much-cited mouse experiment, melanoma cells were injected into the vessels draining to the liver and counted at each stage: the large majority survived transit and got out of the vessels, around two percent formed small clusters, and only about one cell in five thousand grew into a tumour you could see. The bottleneck is not the travelling and not even the arrival — it is the last step, growing in a foreign tissue. Shedding cells is common; the disease is rare relative to the shedding.
Where do those rare successes happen, and why? Two answers have competed since the nineteenth century. Stephen Paget, reviewing hundreds of autopsies, noticed that secondaries were not distributed as blood flow alone would predict, and proposed that particular cells thrive only in particular tissues. James Ewing later argued the simpler thing: cells lodge in the first capillary bed they meet, and the pattern is plumbing.
Both are right about different cancers. Colorectal cancer spreads to the liver largely because its venous drainage goes there first. But breast and prostate cancers go to bone, and eye melanoma to liver, far more than circulation alone explains. Mechanisms for that preference have since been found: receptors on tumour cells that answer signals abundant in particular organs, and signals sent ahead by the primary tumour that make a distant tissue more hospitable before any cell arrives. The arriving cell is not simply lucky in its landing site; to some degree the site was prepared.
The analogy
THE ANALOGY #Paget's own image is still the right one. A plant in seed casts thousands of seeds in every direction, and they are carried wherever the wind takes them — but they grow only where they fall on congenial soil. The seeds are not rare and the wind is not selective; the soil is what decides.
A seed carries its own provisions and waits passively for good ground, whereas a tumour cell must actively suppress its own death on detachment, and — unlike any seed — the parent plant sends signals ahead that change the soil before the seed lands.
Clarifying the model
THE MODEL #Three refinements connect these steps into something usable.
First, "harmless where it began" is a convenience, not a truth. A primary tumour can kill by obstruction or by sitting somewhere unresectable — a small brain tumour that never metastasises can be fatal. What spread changes is resectability, not malignancy.
Second, the seed-and-soil and mechanical accounts are not rivals to be settled. Circulation determines which organs cells reach at all; compatibility determines which of those they can grow in. Both filters, in series, are needed to explain the observed patterns.
Third, arriving is not the same as growing. Disseminated cells can sit in a tissue for years without dividing — dormancy — and this is the best available explanation for breast cancer recurrences a decade or two after apparently successful treatment. What ends dormancy is poorly understood, as is why the final colonisation step is so much harder than the earlier ones. Be careful, too, with the popular idea that cells switch on a migratory program and switch it off on arrival; that transition is well documented in tissue culture, but how far it is required in human tumours has been genuinely contested.
A picture of it
THE PICTURE #How to readRead left to right as a cohort of ten thousand cells passing through the stages of the cascade, with the width of each ribbon proportional to how many cells take that path — the proportions follow the mouse experiment described above and are illustrative of it, not a general law. Notice where the loss occurs. Transit and arrival are survivable for most of the cohort, so the wide ribbon runs almost the full width of the picture. The collapse happens at the last two splits, where the ribbon that keeps going becomes too thin to see. That nearly invisible final strand is the entire disease.
What became clearer
WHAT CLEARED #Metastasis is not a tumour becoming more dangerous but the same tumour becoming unresectable — and it is an extraordinarily inefficient process, in which the hard step is not escaping or travelling but growing in a tissue that was built for something else. Where those rare successes occur is settled by two filters in series: circulation decides where cells arrive, and compatibility — partly intrinsic, partly prepared in advance by the primary — decides where they can live. And the first thing a cell must do to begin any of it is switch off the death program that normally punishes leaving home.
Where to go next
ONWARD #- What holds disseminated cells dormant for years, and what wakes them.
- Whether circulating tumour cells in blood samples can be used to detect spread early enough to matter.
Key terms
TERMS #| Term | What it means |
|---|---|
| Anoikis | programmed cell death triggered by loss of attachment to the extracellular matrix. |
| Seed and soil hypothesis | Paget's proposal that organ preference reflects compatibility between tumour cell and host tissue, not circulation alone. |
| Pre-metastatic niche | a distant tissue altered by signals from the primary tumour before any tumour cell arrives. |
| Dormancy | the persistence of disseminated tumour cells without growth, sometimes for many years. |
Every term the collection defines is gathered in the glossary.