19.3 — How Cancer Spreads

Around 90 percent of cancer deaths are caused by metastasis, not by the original tumour.

A cancer confined to where it started is usually curable by removing it. Everything difficult about cancer follows from its ability to leave.

And metastasis is extraordinarily inefficient. Millions of cells enter the bloodstream from a tumour daily, and the vast majority die. Estimates put the proportion that successfully establish a metastasis at well under 0.01 percent.

Which is encouraging in one respect: the process has many steps, and every step is a potential point of failure — or of intervention.

The steps

1. Local invasion.

The cell loses its attachments to its neighbours, acquires the ability to migrate, and digests its way through the basement membrane (Chapter 4.2).

And that membrane is the single most important structure in cancer staging. A cancer above it is in situ — it cannot spread, and removing it is curative. A cancer that has crossed it is invasive.

Which is the entire purpose of screening: catching cancers before they cross.

2. Entering a vessel — intravasation.

3. Surviving in the circulation.

And this is where most cells die. They are not anchored, which normally triggers a specific form of apoptosis; they are subject to shear stress; and they are attacked by natural killer cells (Chapter 13.1).

Cells that travel in clusters, or coated in platelets, survive better — the platelets providing physical protection and shielding them from immune recognition.

Which is one proposed explanation for observations that aspirin use is associated with reduced metastasis, though the evidence remains debated.

4. Arrest and exit — extravasation.

5. Survival at the new site — and this is the hardest step.

Most cells that arrive somewhere new die, or enter dormancy.

And dormancy explains one of the most difficult features of cancer: recurrence years or decades later. Breast cancer can recur 20 years after apparently successful treatment, and the explanation is disseminated cells that lay dormant and then reactivated.

What wakes them is not established — inflammation, injury, immune changes and hormonal shifts are all candidates. It is an active research area with real therapeutic implications, because keeping cells dormant would be as good as eliminating them.

6. Colonisation — growing into a detectable deposit, which requires recruiting a blood supply (Chapter 19.2).

Where cancers go, and why

Two explanations, and both are partly right.

Mechanical — the first capillary bed encountered.

Colorectal cancer drains into the portal vein, so it metastasises to the liver first (Chapter 7.5). Most other cancers drain into the systemic veins, so the lung is the first filter.

"Seed and soil" — proposed by Stephen Paget in 1889 after noticing that breast cancer metastasised to specific organs more than blood flow predicted.

Some tissues are hospitable to particular tumours and some are not. The spleen receives substantial blood flow and is a rare site of metastasis; bone receives less and is common.

And the modern version is the pre-metastatic niche: the primary tumour releases factors and vesicles that prepare a distant site before any cell arrives — recruiting cells and altering the matrix to make it receptive.

Which is a genuinely striking finding: the tumour prepares its destination in advance.

The characteristic patterns are worth knowing, because they determine where staging scans look:

Cancer	Common sites
Breast	Bone, lung, liver, brain
Lung	Brain, bone, liver, adrenal
Prostate	Bone, especially spine
Colorectal	Liver, then lung
Melanoma	Almost anywhere, including brain
Sarcoma	Lung, via blood

Lymphatic spread

Most carcinomas spread through lymphatics first (Chapter 7.8).

And the reason is anatomical: lymphatic capillaries have overlapping cell edges with gaps large enough for a cell, and no basement membrane to cross.

Cells travel to the first draining node, lodge, and grow — then to the next.

That orderly progression is what makes lymph node status the single most important prognostic factor in most solid cancers, and it is why nodes are sampled during cancer surgery.

Sentinel node biopsy — injecting a tracer at the tumour and identifying the first node it reaches.

If that node is clear, the rest almost certainly are, so the remainder are left in place. It has substantially reduced lymphoedema after breast cancer surgery without compromising outcomes (Chapter 7.8) — a genuine example of surgery becoming less invasive by understanding anatomy better.

Sarcomas spread through blood rather than lymphatics, which is why nodes are not routinely removed in sarcoma surgery.

Why metastasis is hard to treat

Multiple sites, so local treatment cannot address all of them.

Heterogeneity — different metastases carry different mutations, so a drug targeting one may not affect another (Chapter 19.2).

Sanctuary sites. The brain is the clearest — the blood–brain barrier excludes most drugs (Chapter 11.10), which is why brain metastases occur in patients whose disease is otherwise controlled. Newer agents designed to cross it have made a real difference in some cancers.

And resistance evolves, because the tumour is a population under selection.

Symptoms of spread

And these are what prompt investigation.

Bone — pain that is worse at night and not relieved by rest, fracture from minimal force, and high blood calcium (Chapter 10.3).

And spinal cord compression is the emergency. Back pain with leg weakness, numbness, or bladder or bowel changes, in someone with known cancer — it needs imaging and treatment the same day, because function lost is not recovered.

Liver — pain in the right upper abdomen, jaundice, abnormal liver tests.

Lung — breathlessness, cough, coughing blood.

Brain — headache worse in the morning and on lying flat, vomiting, seizures, focal weakness, personality change.

And the general signs: unexplained weight loss, drenching night sweats, and unexplained fatigue (Chapter 16.4).

Treating metastatic disease

And the framing has changed, which matters.

For most cancers, metastatic disease is treatable rather than curable — controlled for months to years, sometimes much longer.

But there are genuine exceptions, and they are worth knowing:

Testicular cancer — cured in the majority even with widespread metastases (Chapter 15.1).

Lymphoma and childhood leukaemia — curable when disseminated.

Choriocarcinoma — curable with chemotherapy alone.

Some colorectal and other cancers with limited liver or lung metastases — oligometastatic disease, where surgical removal or precise radiotherapy of a small number of deposits produces long-term survival in a meaningful proportion.

And melanoma and some lung cancers on immunotherapy — where a subset of patients with metastatic disease remain in remission for years, and increasingly appear to be cured (Chapter 19.7).

That last one is the largest change. A decade ago, "metastatic melanoma" and "months" were the same sentence.

Paraneoplastic syndromes

Effects of a cancer at a distance, caused by hormones or immune reactions rather than by the tumour's physical presence.

And they matter because they sometimes appear before the cancer is detectable.

Endocrine: small cell lung cancer producing ADH, causing severe low sodium (Chapter 10.3); squamous lung cancer producing a parathyroid-like hormone, causing high calcium; ectopic ACTH causing Cushing's syndrome (Chapter 12.4).

Neurological: Lambert–Eaton myasthenic syndrome, associated with small cell lung cancer in around half of cases (Chapter 6.2); several encephalitis syndromes.

Skin: dermatomyositis in an adult over 40 warrants a search for cancer (Chapter 6.7); acanthosis nigricans; and a sudden eruption of multiple seborrhoeic keratoses.

Haematological: clotting abnormalities, and migratory thrombophlebitis associated with pancreatic cancer.

And an unexplained venous thrombosis in an older person is one of the more useful pointers, because 5 to 10 percent have an underlying cancer (Chapter 18.10).

Can spread be prevented?

Partly, and the honest answer includes what does not work.

Adjuvant treatment — chemotherapy, hormonal treatment or targeted therapy given after surgery, aimed at cells that have already disseminated but are undetectable.

And this is where much of the survival improvement in breast and colorectal cancer has come from. Adjuvant treatment does not treat a visible tumour; it treats cells that may or may not be there, which is why it is given to populations rather than individuals and why the number needed to treat matters.

Circulating tumour DNA testing after surgery — detecting minimal residual disease — is beginning to identify who actually needs it, which would spare a large number of people unnecessary chemotherapy. It is one of the more promising developments in the field.

Neoadjuvant treatment — given before surgery, shrinking the tumour and providing early information about whether the treatment works.

Aspirin — associated with reduced colorectal cancer incidence and metastasis in several studies, and the balance against bleeding risk means it is not recommended for general prevention.

Physical activity after a cancer diagnosis is associated with improved survival in breast and colorectal cancer, and the association is consistent enough to be part of standard advice (Chapter 24.5).

What the next page fixes

Chapter 19.4 covers how cancer is described — staging and grading, what the numbers mean, and why "stage 4" does not mean what most people assume it does.